J1: IDENTIFYING INITIAL DIAGNOSIS IN DATABASE RESEARCH

Volume 2 • Number 1 • 1999 VALUE IN HEALTH

ABSTRACTS Contributed Podium Presentations

OUTCOMES RESEARCH ISSUES AND ECONOMIC AND OUTCOMES STUDY RESULTS FOR CANCER

CONCLUSIONS: Studies using disease-free time periods of less than 12–18 months may include 10–20% of subjects identified that are not actually initially diagnosed patients. This may introduce potential bias if researchers think they are analyzing initially diagnosed subjects.

J1

IDENTIFYING INITIAL DIAGNOSIS IN DATABASE RESEARCH Lalla DV, Kozma CM University of South Carolina, Columbia, SC, US

J2

OBJECTIVE: Studies using claims databases usually establish initial diagnosis for a disease by requiring subjects to be disease-free for some period of time before inclusion in the study. This time period should be stringent enough to minimize misclassification of cases, but liberal enough to ensure an adequate sample size. This study evaluated the change in study sample size when the time period that subjects were required to be disease-free (disease-free time period) was varied. METHODS: Patients diagnosed with hypertension (ICD9-CM 410) in the South Carolina Medicaid population in 1995 (n 5 20,829) were followed back in time (1992– 1994) to determine the change in sample size when the disease-free time period was varied from 6 months to 3 years. All patients were required to be eligible for at least 11 months in each year of interest. RESULTS: The table below shows percent decreases in sample size when disease-free time periods were varied. The disease-free sample size decreased by approximately 50% in the 12-month period following patient identification. It continued to decrease at a lower rate as the disease-free time period was increased to 18 months and tapered off thereafter.

Years of interest 1994 1993 & 1994 1992, 1993, & 1994

REVIEW OF LEGISLATION RELATING TO PROMOTION OF MEDICINAL PRODUCTS IN THE EUROPEAN UNION—ARE HEALTH ECONOMICS DATA COVERED? Duthie T1, Thomander L1, Towse A2, Kotsanos J1, Tilson H3 1 Eli Lilly & Company, Basingstoke, UK; 2Office of Health Economics, London, UK; 3University of North Carolina, Chapel Hill, NC, US Increasing demand for health economics (HE) means that HE information on pharmaceutical interventions is becoming more widely used in promotion. OBJECTIVE: To identify and assess the current legislative structure relating to the inclusion of HE data in the advertising of medicines in the European Union (EU). METHOD: A review of published literature and EU legislation. RESULTS: Advertising and promotion of medicinal products for human use is regulated by the EC directive 92/28/EEC. In addition, member states have their own codes of practice, some regulate pharmaceutical advertising by tribunals (e.g., Germany), others by self-regulating bodies (e.g., the ABPI in the UK), and some by the health authority (e.g., Agence de Medicament in France). All embody the principles and standards from 92/28/EEC, whereby advertising should encourage rational use of the

Pt # eligible for 11 mo in each year of interest

6 mo

12 mo

18 mo

24 mo

30 mo

36 mo

13202 10512 8634

51.53 52.63 52.95

61.69 63.57 63.93

NA 70.41 70.81

NA 73.77 74.18

NA NA 76.11

NA NA 77.43

© ISPOR 1098-3015/99/$10.50/5 5–50

% decrease in sample size

5

6

Abstracts

product, presenting it objectively, without exaggeration and without being misleading. EFPIA and IFPMA also publish Codes of Practice for the Promotion of Medicines. In these approaches HE data are not explicitly considered; the focus is clinical. Similarly, until recently, legislation for the separate review of HE data did not exist in the US. In addition, a recent cross-functional initiative through ISPOR to develop guidelines for healthcare economic promotional materials has now been reported. CONCLUSION: We have identified no specific guidelines for advertising using HE data in the EU. Given this, key questions are: 1) Do existing codes of practice offer enough scope to review HE data? and 2) How can we ensure quality assurance of HE information provided to decision-makers?

J3

HOW DO PATIENTS MAKE SATISFACTION JUDGMENTS? ANSWERS FROM STRUCTURAL EQUATION MODELING Trudeau E MAPI Values, Lyon, France Quality of life and patient satisfaction cohabit more and more in the same paradigm, through the notion that patient satisfaction with care and treatment can lead to positive outcomes, which in turn adds to overall clinical effectiveness. Yet patient satisfaction is often used without a clear understanding of the concept. OBJECTIVE: To test a model of patient satisfaction that would identify the antecedents of the satisfaction judgment. METHOD: Self-administered questionnaire to 532 inpatients of two acute-care hospitals in Montreal (Canada). The questionnaire consisted of six Likert scales measuring: expectations (3 items), perceived quality (17 items), conformity to expectations (3 items), needs (6 items), fairness (2 items), and satisfaction (10 items). Scales reliability and construct validity were assessed with Cronbach alpha coefficient of internal consistency and confirmatory factorial analysis. The model was tested with causal modeling (LISREL 8). RESULTS: All alpha coefficients ranged from 0.70 to 0.91. The model provides satisfactory fit to the data (x2 5 26.44, df 5 28, p 5 0.55) and explains 61% of the variation in satisfaction. Examination of total effects (sum of direct and indirect effects of one variable on another) reveals that perceived quality has the greatest influence on satisfaction (total effect 0.62), followed by conformity to expectation (total effect 0.44), fairness (total effect 0.44), needs (total effect 0.16), and expectations (total effect 0.11). In conclusion, the utilization of satisfaction information to its full potential not only requires reliable and valid instruments, but also relevant concepts, such as perceived quality and conformity to expectations, to take into account the complexity of the satisfaction judgment.

CAN1

COST-EFFECTIVENESS ANALYSIS OF HIGH-DOSE CHEMOTHERAPY WITH HEMATOPOIETIC RESCUE AS PRIMARY TREATMENT FOR METASTATIC BREAST CANCER Trippoli S, Messori A Laboratorio SIFO di Farmacoeconomia, c/o Centro Informazione Farmaci, Azienda Ospedaliera Careggi, Florence, Italy OBJECTIVES: The analysis of published survival curves can be used as the basis for conducting cost-effectiveness analyses in which two treatments are compared in terms of cost per life year saved. In patients with metastatic breast cancer, high-dose chemotherapy (HDC) with autologous bone marrow transplantation (ABMT) has been reported to improve survival in comparison with control patients who receive standard chemotherapy. METHODS: An incremental cost-effectiveness analysis was undertaken in which the Gompertz model was used to determine a lifetime estimate of patient-years gained by subjects given HDC with hematopoietic rescue in comparison with controls. Our study utilized the clinical data reported in a published clinical trial. This randomized clinical trial involved 45 patients subjected to HDC with ABMT and 45 controls given chemotherapy. RESULTS: Lifetime survival advantage for patients in the HDC with ABMT group was estimated as 72.5 discounted patient-years for every 100 patients. The use of HDC with ABMT, as opposed to standard chemotherapy alone, was associated with an incremental cost of about $55,000 per discounted life year gained. CONCLUSIONS: The cost-effectiveness ratio of highdose chemotherapy combined with autologous bone marrow transplantation in patients with metastatic breast cancer is borderline if one considers that acceptable figures of cost per life year gained are generally below $50,000.

CAN2

A NEW METHOD FOR EXPRESSING SURVIVAL AND LIFE EXPECTANCY IN CANCER PATIENTS Messori A, Trippoli S Laboratorio SIFO di Farmacoeconomia, c/o Centro Informazione Farmaci, Azienda Ospedaliera Careggi, Florence, Italy We describe a new method for expressing survival in cancer. Our method has the following characteristics: In analyzing a survival curve (first analysis), the plateau of the right portion of the curve is extrapolated to infinity by assigning a normal life expectancy to long-term survivors (“cured” patients). A second survival curve (second analysis), which represents the expected survival of a healthy patient cohort (age-matched and gender-matched with the patient cohort under study), is constructed. The second curve is used as a reference for internal comparisons. Both curves are assessed by measuring the area under the curve (AUC). The ratio of the AUC from the first analysis to the AUC from the second analysis quantifies

Abstracts

7

the impact of the disease on life expectancy. As an example of the application of our method, we have reanalyzed a survival data set reported by Spinolo et al. (1992) that refers to patients with acute leukemia who relapsed after their first allogeneic bone marrow transplant (BMT). We obtained the following results: total AUC leukemia patients (normalized to 100 patients) 5 10,590 months, mean lifetime survival per patient (MLS) 5 105.9 months or 8.8 years; total AUC healthy subjects 5 58,376 months, MLS 5 583.8 months or 48.6 years. These data indicate that the average life expectancy (or MLS) after a second BMT was 8.8 years compared with 48.6 years in an age-matched and gender-matched cohort of healthy subjects. Our survival analysis shows that, in this very serious disease condition, the absolute reduction in life expectancy was 39.8 years per patient and the percent reduction was 82%.

CAN3

PRACTICE PATTERNS OF CHEMOTHERAPY USAGE AND ASSOCIATED NEUTROPENIA AND FEBRILE NEUTROPENIA (FN) IN EUROPE Lyman GH London School of Hygiene and Tropical Medicine, London, UK; H. Lee Moffitt Cancer Center and Research Institute at the University of South Florida, Tampa, FL, US Neutropenia and associated FN are major dose-limiting toxicities of cancer chemotherapy and impact considerably on the costs of care and patient quality of life. There is little information available concerning patterns of chemotherapy treatment and their associated complications in Europe. OBJECTIVES: To study patterns of chemotherapy usage and the occurrence of neutropenia and FN in cancer patients in Europe. METHODS: Recent patterns of cancer treatment were examined from survey data provided by ISIS Research for the five major European economic markets ($bn GDP): Germany (2115), France (1394), United Kingdom (1278), Italy (1146), and Spain (533). Data were evaluated from 60 cancer specialists and 1200 cancer patients receiving cancer treatment in each country. Practice patterns examined included the use of systemic chemotherapy, the occurrence of neutropenia and FN and the type of supportive care provided. RESULTS: Ninety percent of patients were receiving treatment with chemotherapy, of which 20% was administered in an adjuvant fashion. Average patient age was 55 years and 53% were female. Major treated cancers included those of breast (23%), colon (13%), and lung (12%). Approximately one-half of patients were newly diagnosed, while 25% were described as heavily pretested. Nearly 40% of patients had experienced a prior episode of severe neutropenia and 10% had experienced a previous infection. Concurrent use of colony-stimulating factors (CSFs) was reported in 15% of patients. Major factors predicting CSF use included previous or antici-

pated severe neutropenia, FN, or infection. Considerable regional variation was evident in the treatment practices observed. CONCLUSIONS: Neutropenia and FN are frequent complications of systemic cancer chemotherapy in Europe.

ECONOMIC AND OUTCOMES STUDY RESULTS FOR NEUROLOGIC AND GASTRIC DISORDERS

ND1

AN 11-YEAR, CONTROLLED STUDY OF HEALTHCARE RESOURCE USE AND MORTALITY FOR PATIENTS WITH PARKINSON’S DISEASE Donnan P, Stubbings C, Steinke D, MacDonald TM, McDevitt DG, Morris AD, Davey P Medicines Monitoring Unit (MEMO), University of Dundee, Ninewells Hospital & Medical School, Dundee, Scotland There are few long-term data about Parkinson’s disease (PD) other than from hospital clinics, which may see a biased selection of patients. METHODS: We used two populations in Tayside: one with complete prescribing data from 1989–1995 (n 5 30,000) and the second with data from 1985–1995 (n 5 10,000). The first 6 months of data were used to identify incident cases and to exclude those who had previously taken a neuroleptic drug. Cases were matched for age, sex, and general practice to nine controls. RESULTS: The number of cases of incident PD was 108 in the 7-year cohort and 28 in the 11-year cohort. In the 7-year cohort, patients were aged 45–95 (median 76.0), and 55% were male. The most commonly prescribed drug was co-beneldopa at a mean dose of 187 mg. In comparison with control, PD patients were significantly more likely to receive additional drugs for gastrointestinal (relative risk [RR] 2.0; 95% confidence interval [CI] 1.4–3.3) and cardiovascular (RR 1.7; CI 1.2–2.4) disorders. Median costs of hospitalization were higher in PD (14,119 vs. 5,077; p 5 0.0001) because of increased numbers of episodes per patient and costs per episode (1,965 vs. 1,526; p 5 0.002). Relative risk of mortality in PD patients was very similar in the 7-year cohort (RR 5 2.50, CI 1.74–3.57) and the 11-year cohort (RR 5 2.25, CI 1.43–3.54). CONCLUSIONS: This population-based study provides an unbiased estimate of the costs and mortality of PD, hence of the potential value of improved treatment.

8

Abstracts ND3

ECONOMIC EVALUATION OF ACAMPROSATE IN MAINTAINING ABSTINENCE IN ALCOHOL-DEPENDENT PATIENTS Annemans L, Vanoverbeke N HEDM, Mechelen, Belgium OBJECTIVES: To conduct a cost analysis of acamprosate in maintaining abstinence in weaned alcoholic patients in Belgium. METHODS: The analysis was based on a state-transition approach, comparing the use of acamprosate versus “doing nothing” (no pharmaceutical maintenance treatment) in alcoholic patients after detoxification. The analysis considered the probability of additional relapses over a period of 24 months. Clinical data with regard to relapse rates were obtained from a controlled double-blind trial by Witworth et al. (1996) (n 5 448), involving 12 months treatment with acamprosate and 12 months follow-up. This trial showed that the advantage of acamprosate over placebo in preventing relapse continued during the 12 months after cessation of treatment. Treatment patterns and resource utilization in first line were retrieved from a Belgian survey among 129 GPs. Data on resource utilization in second line were available from a large, open pragmatic prospective Belgian trial (n 5 582). Costs were calculated from the perspective of health insurance. Unit costs were obtained from official listing and hospital statistics. RESULTS: Net cost savings of 21,301 BEF per patient were obtained over a 24-month period for acamprosate at a daily cost of 78.6 BEF for the health insurance compared with “doing nothing.” Most savings were obtained by avoiding acute hospitalizations for detoxification and institutionalized rehabilitation. Various sensitivity analyses showed a large robustness of the conclusions. A global anticipated net saving of 70 million BEF for the Belgian healthcare sector over 2 years was estimated. CONCLUSIONS: The results of our analysis show that acamprosate is a cost-saving intervention in maintaining abstinence in weaned alcoholic patients, if applied during a period of 12 months after detoxification.

ND4

CROSS-CULTURAL ADAPTATION OF THE MIGRAINE-SPECIFIC QUALITY OF LIFE QUESTIONNAIRE (MSQ) IN 14 LANGUAGES Conway K1, Uzun V1, Jhingran P2, Kwong J2, Duthie T3 1 MAPI Research Institute, Lyon, France, 2Glaxo Wellcome Inc., Research Triangle Park, NC, US; 3Formerly of Glaxo Wellcome, Uxbridge, UK While measurement of health-related quality of life (HRQOL) has become a vital part of assessing the effects of migraine treatment in many international studies, a measure that is well translated and cross-culturally adapted has not been available. The 16-item Migraine-Specific Quality of Life Questionnaire (MSQ) (Version 2.0), which assesses as-

pects of health primarily affected by migraines (initially developed in US English), has been culturally adapted into 14 languages from Europe and Canada. The translation process of the MSQ was coordinated by a quality of life (QoL) specialist in each target country and followed a rigorous methodology that consisted of: 1) two separate forward translations produced by professional translators, native speakers of the target language and bilingual in English; 2) comparison and reconciliation of the translations by QoL specialist and translators; 3) a backward translation by a native English speaker; 4) comparison of source with backward version; 5) comprehension test in a sample target population; and 6) international harmonization. A number of conceptual and linguistic issues relating to the concept of “migraine,” “migraine attack,” and “migraine headache,” and the meaning of the expression “daily activities,” emerged during the translation process. A literal translation of idiomatic expressions such as “let someone down” was also not always possible; most languages referred to “disappoint.” Likewise, the literal translation of “social activities,” referring to humanitarian work in certain languages, required a descriptive equivalent. Several language-specific changes were made to resolve each issue during international harmonization. The MSQ (Version 2.0) was translated into 14 languages for use in international studies. A rigorous translation methodology was performed to ensure that the translated versions were conceptually equivalent to the source version and were cross-culturally valid. Psychometric testing will be conducted in the future to ensure reliability and validity in each translation.

GI1

COST-UTILITY ANALYSIS OF LAPAROSCOPIC VERSUS OPEN SURGERY FOR COLONIC DISEASE Eypasch E, Köhler L, Tschubar F, Troidl H, Lauterbach KW St. Hildegardis Hospital, II, US; Department of Surgery, University of Cologne, Cologne, Germany; and Institute for Health Economics at the University of Cologne, Cologne, Germany OBJECTIVES: The cost of laparoscopic colonic surgery is dominated by expenses for disposable instruments. The utility of this surgery in terms of quality of life has not been well documented. The two variables of costs versus quality of life were evaluated in a multi-institutional prospective trial. METHODS: In two academic departments of surgery with the same level of skill for endoscopic surgery, patients were prospectively assigned to have either open (N 5 27) or laparoscopic (N 5 35) surgery for sigmoid diverticulitis or colon cancer. Quality of life (QoL) was assessed by three different instruments: SF-36, EuroQol, and Gastrointestinal Quality of Life Index (GIQLI). Costs were evaluated using a standardized protocol where all resources were directly assigned to the individual patient and procedure (i.e., operating room and ICU hours, working minutes of surgeons, nurses, use of suture materials, stapling devices, units of

Abstracts

9

blood, antibiotics, etc.) Economic evaluation was based on 1997 prices. Data were compared to a pool of previously collected variables in a larger number of German hospitals by an independent consulting company (GEBERA). RESULTS: Costs for endoscopic surgery were 9,428 Deutschmarks (DM) versus 10,896 DM for open surgery, Operating room costs came to 5,091 DM for endoscopy versus 3,986 DM for open surgery. A shorter hospital stay of 12.7 days in comparison to 18.9 days for open surgery contributed to the higher total costs of open surgery. After endoscopic surgery, patients had higher quality-of-life scores even 6 weeks after surgery—GIQLI: 115 points versus 95 points (p , .05). Parallel results were confirmed by the EuroQol and the SF-36 scores. CONCLUSIONS: Laparoscopic colonic surgery is a more cost-effective alternative to open procedures, offering a better quality of life, but requiring highly developed surgical skills.

GI2

DIGEST INTERNATIONAL SURVEY: IMPACT OF UPPER GASTROINTESTINAL SYMPTOMS ON RESOURCE UTILIZATION AND QUALITY OF LIFE Einarson TR University of Toronto, Toronto, Canada, and the DIGEST Investigators OBJECTIVES: In 1997, a face-to-face survey of urban adults was conducted in the Nordic countries, Italy, the Netherlands, Switzerland, Canada, the US, and Japan to determine the prevalence of upper GI symptoms and impact on resource consumption (physician visits, medication use, hospitalization days, workdays off) and quality of life (QoL). METHODS: Validation of 128 patients and 72 volunteers yielded rInternal consistency 5 0.9–1.0 and rtest–retest 5 0.6–0.9. Questions were asked about symptoms, categorized as gastroesophageal reflux disease (GERD) or other upper gastrointestinal symptoms (UGIs), and healthcare utilization during the previous 3 months. Quality of life was measured using the Psychological General Wellbeing Index (PGWBI) (score 0–110) and the pertinent SF-36 questions. RESULTS: Of 5581 responses analyzed, @500 each were from Switzerland, Japan, and the Netherlands, and @1000 were from other countries. Results were similar across countries. Average age was 44 years; 51% were females; 41% reported symptoms, the most common being heartburn (36%), postprandial fullness (31%), and abdominal distention (31%). There was no age effect, except that those aged .65 had lower prevalences. More females reported UGIs (58%); males reported GERD (52%). Symptomatic persons had 2.6 times more physician visits for any reason (1.60 versus 0.62, p , 0.05), 1.5 times more prescriptions (p , 0.001) and 1.7 times more over-the-counter medications (OTCs) (p , 0.001), 2.5 times more hospital days (0.55 versus 0.22, p , 0.05), and 2.2 times more days off work (4.20 versus

1.85, p , 0.05). Patients averaged 0.51 physician visits for GERD/UGIs. QoL was significantly lower in symptomatic than nonsymptomatic persons for all domains. Overall PGWBI mean (71.8 versus 85.6, p , 0.01) correlated with increased physician visits (p , 0.001). Projecting annually for the US, .1 billion additional workdays are lost, costing $100 billion (1.5% of GDP). CONCLUSIONS: GI symptoms constitute a major health problem with substantial health and economic impact.

GI3

THE EFFECT OF AN OPEN ACCESS ENDOSCOPY SERVICE ON PRESCRIBING COSTS OF ULCER-HEALING DRUGS Davey P1, McMahon AD1, Goudie B2, MacDonald TM1, White G1, Morris AD1, Murray F2 1 Medicines Monitoring Unit (MEMO), University of Dundee, Ninewells Hospital & Medical School, Dundee, Scotland; 2 Department of Gastroenterology, Ninewells Hospital, Dundee, Scotland Open access services allow direct referral by primary care doctors without an opinion from a gastroenterologist. Our aim was to measure this new service’s effect on prescribing costs. METHODS: The study used all patients from Tayside who had complete prescribing data for 1 year before referral and 1 year after endoscopy and had a diagnosis recorded at endoscopy. The main outcome measure was the change in prescribing costs for dispensed ulcer-healing drugs (UHDs) in the 4 months after endoscopy compared to the 4 months before referral. RESULTS: Of the 1063 patients referred, 883 were resident in Tayside and had complete data. The results of endoscopy were normal in 311 (35%), esophagitis in 323 (37%), ulcer in 55 (6%), both in 33 (4%), and other upper GI pathology in 161 (18%). Significant independent predictors of prescribing cost after endoscopy were: the cost of drugs before referral, the diagnosis at endoscopy, and age (p 5 0.0001 by both generalized linear modeling and a nonparametric ranking test). Variables that were not significantly associated with prescribing cost after endoscopy were: sex, interval between referral and endoscopy, and previous history of either UHDs, endoscopy, or hospitalization for GI event. Controlling for age and cost of drugs before referral, there was a mean decrease in prescribing costs of £11 per patient with normal endoscopy. Prescribing costs increased by £33 for esophagitis, by £25 for ulcer, by £41 for both, and by £23 for other pathology. CONCLUSIONS: Open access endoscopy was associated with increased cost of prescribing for UHDs, mainly because the procedure revealed pathology in 65% of cases.

10

Abstracts L2

ECONOMIC AND OUTCOMES STUDY RESULTS FOR PULMONARY DISORDERS AND INFECTIOUS DISEASE

L1

COMPLIANCE WITH INHALED CORTICOSTEROIDS AND HOSPITALIZATION FOR ASTHMA McMahon AD1, Lipworth B2, Morris AD1, Davey P1, MacDonald TM1 1 Medicines Monitoring Unit (MEMO), University of Dundee, Ninewells Hospital & Medical School, Dundee, Scotland; 2 Department of Clinical Pharmacology, Ninewells Hospital, Dundee, Scotland BACKGROUND: Current asthma guidelines advocate early therapy with inhaled corticosteroids to reduce bronchial inflammation. The aim of the present study was to examine the association between compliance with preventive therapy (inhaled corticosteroids, sodium cromoglycate, and nedocromil) and hospitalization for asthma. METHODS: From the Tayside MEMO record-linkage database we identified a cohort of subjects receiving prescriptions for inhaled corticosteroids. Subjects aged 12–45 years, with at least 3 months’ exposure during the period August 1992 to December 1994, were included in the study. Compliance was estimated by calculating the maximum number of days for which a subject could have taken an inhaled corticosteroid. In the 3-month exposure window, subjects with 90 days therapy were designated as “compliant,” those with 1–89 days were defined to be “partially compliant,” and those with zero days were “noncompliant.” RESULTS: Of the subjects, 4535 provided calculable compliance, with a total of 88 occurrences of hospitalization for asthma. The proportion of subjects with hospitalization was 9% for compliant subjects, 3% for partially compliant subjects, and 1% for noncompliant subjects. The odd-ratios (ORs) for partial compliance versus compliance were: 0.34 (95% confidence interval [CI] 0.19– 0.62) and 0.10 (95% CI 0.05–0.19) for noncompliance versus compliance. The association between compliance and hospitalization disappeared after adjustment for number of doses of beta-agonists and the number of types of medication taken for relief of asthma. CONCLUSION: As the degree of compliance increased, the incidence of hospitalization with asthma also increased. However, this association was probably due to worsening symptoms leading to better compliance with all medications. The results suggest that patients with asthma take their preventive medications more regularly when the condition is worsening.

EFFECTIVENESS OF ORAL CIPROFLOXACIN IN THE TREATMENT OF EXACERBATIONS OF CHRONIC AIRWAY DISEASES LeLorier J1, Blais L1, Couture J1, Bourbeau J2 1 Centre de recherche, Centre hospitalier de l’Université de Montréal, Campus Hôtel-Dieu, Canada; 2Unité d’épidémiologie respiratoire, Université McGill, Montréal, Canada OBJECTIVE: Oral ciprofloxacin is an antibiotic commonly used for ambulatory treatment of exacerbations of chronic airway diseases (ECAD). However, its clinical superiority over other antibiotics remains to be proven. Randomized clinical trials in which efficacy was measured with bacterial eradication and disappearance of symptoms provided inconclusive results. METHODS: Using the administrative databases of the Quebec Health Insurance Plan, we assessed the effectiveness of oral ciprofloxacin in the ambulatory treatment of ECAD among patients aged over 65 years. Within a cohort of 10,820 subjects (mean age 73 years) followed from 1989 to 1996, we performed a case-control analysis to determine whether oral ciprofloxacin prevents hospitalizations for ECAD. RESULTS: During the first 7 days of therapy, 465 cases of hospitalization for ECAD occurred, while 175 cases occurred after the seventh day of therapy. After adjusting for markers of disease severity, we found that users of oral ciprofloxacin were less likely to be hospitalized than users of other antibiotics during the early part of therapy (1–7 days) (rate ratio 5 0.6; 95% CI 0.4–1.0). However, after 7 days of therapy we found that users of ciprofloxacin were more likely to be hospitalized for ECAD (rate ratio 5 1.7; 95% CI 1.1–3.3). CONCLUSION: Ciprofloxacin was found to be associated with a reduction in hospitalizations for ECAD compared with other antibiotics early in the course of therapy, but with an increase in hospitalizations after 7 days of therapy. The lack of effectiveness in the later part of therapy is likely due, at least in part, to confounding by indication.

L3

QUALITY OF LIFE OF ASTHMA PATIENTS RELATIVE TO ASTHMA SEVERITY Greiner W, Schulenburg Graf von JM Center for Health Economics and Health System Research, University of Hannover, Germany OBJECTIVE: Concepts of quality of life are gaining increasing significance as measures of outcome, both in medicine as well as in economics. Therefore—particularly for chronic illnesses such as asthma—instruments are necessary which make the influence of the illness on the daily life of the patients measurable from a subjective point of view. One such questionnaire is the Fragebogen zur Asthmaqualität (FLA), a disease-specific instrument for measuring the quality of life of asthma patients. This ques-

Abstracts

11

tionnaire was supplemented with a generic instrument, the EuroQol. The aim of the study was to measure the quality of life depending upon severity level of the illness. METHODS: Adult asthma patients treated in 23 pneumologic ambulatory care units were randomly selected and asked to fill out both the FLA and the EuroQol questionnaires. Separate analyses of the results were carried out for each patient severity level; they covered the average values of the subscales, the correlation and the internal reliability of the questionnaires. RESULTS: The quality of life of asthma patients correlates strongly to the respective severity level of the illness. This applies to all the examined dimensions of healthrelated quality of life. The patients of level 3 (serious asthma) in particular achieve values far below average when compared to the total population (42 out of 100, according to EuroQol). By contrast, the subjective quality of life of asthmatics in level 1 is only slightly reduced. The correlation of the FLA subscales as well as of the EuroQol values is high (0.74–0.97). CONCLUSIONS: The study makes clear that illness progression is associated with high losses of quality of life for the patient. From a health-economic point of view, this is important for the calculation of so-called intangible costs, which are often not quantified in studies on the cost of disease.

L4

CROSS-CULTURAL ADAPTATION OF THE PEDIATRIC ASTHMA CAREGIVER QUALITY OF LIFE QUESTIONNAIRE (PACQLQ) IN 19 COUNTRIES Mear I1, Conway K1, Juniper EF2, Santanello N3 1 MAPI Research Institute, Lyon, France; 2McMaster University, Hamilton, Ontario, Canada; 3Merck & Co Inc, Blue Bell, PA, US Childhood asthma is a disease that affects not only patients but also the quality of life of their parents and caregivers. The PACQLQ is a 13-item scale that was developed in English to measure this impact. Prior to use in an international trial, the original questionnaire underwent cultural adaptation giving rise to 19 language versions. This cultural adaptation process involved the recruitment of a QoL specialist in each country as supervisor of the translation process. First, each QoL specialist discussed the concepts of the original instrument with the developer (EFJ). Native speakers then produced two independent forward translations. These were reconciled and backtranslated twice into English. Finally, the translations were tested on a sample target population through indepth interviews conducted by the senior translator and asthma clinicians. Translation problems included finding conceptual equivalents for typical English expressions and constructions. For example, “to feel frustrated,” “to feel upset,” and “irritable” either had no literal equivalent or led to misinterpretation in some languages. The item “Did you feel angry that your child has asthma?”— the content of which was found inappropriate in most

countries—had to be changed to reflect an emotion more commonly experienced by parents. Finally, the design and the mode of administration of the questionnaire was somewhat problematic in countries where people are not used to completing questionnaires and have a low level of education. Before use in an international trial, rigorous cultural adaptation was essential to produce cross-culturally valid language versions of the PACQLQ. The comparison of answers across countries will clarify whether the impact of asthma on parents or caregivers is similar in all countries. Measurement properties will also have to be confirmed once clinical data are available.

ID1

IS TRIMETHOPRIM RESISTANCE AN OUTCOME OF COMMUNITY ANTIBIOTIC PRESCRIBING? Steinke D1, Seaton R2, Phillips G3, MacDonald TM1, Davey P1 1 Medicines Monitoring Unit (MEMO), University of Dundee, Ninewells Hospital & Medical School, Dundee, Scotland; 2 Departments of Infectious Diseases; and 3Medical Microbiology, Ninewells Hospital, Dundee, Scotland Antibiotic resistance is a growing problem in all countries. However, because the precise role of community antibiotic prescribing is undefined, practitioners attribute the problem to overuse of antibiotics by others. METHODS: Case-control design with incident urine samples (no sample in previous 6 months) from the population of Tayside from 1 July 1993 to 31 December 1995. Cases were subjects from whom trimethoprim-resistant gramnegative bacteria (TR) were isolated. Controls were subjects whose samples grew trimethoprim-sensitive gramnegative bacteria (TS) or showed no bacterial growth (NG). Exposure to risk factors in each group was determined from the MEMO and DARTS record-linkage databases. Logistic regression was used to identify independent risk factors. Demographic variables were age, sex, socioeconomic status, and diabetes mellitus. Exposure variables (in the previous 6 months) were hospitalization and dispensing of the following drugs from community pharmacies: antibiotics, corticosteroids or estrogens. RESULTS: Of the 9362 incident urine samples used, 380 were TR, 1073 TS, and 7909 NG. The most important risk factors were: exposure to trimethoprim (odds ratio [OR] 3.08; 95% CI 2.48–3.81) or other antibiotics (OR 1.43; CI 1.14–1.80). Additional independent risk factors were age (OR 1.28 for every 10 years after the first decade; CI 1.23–1.35) and prior hospitalization (OR 1.38; CI 1.07–1.79). However, 35% of the TR group had not been exposed to antibiotics or been hospitalized in the previous 6 months. CONCLUSION: There was a strong link between trimethoprim resistance and community prescribing of trimethoprim or other antibiotics. However, prior hospitalization was an independent risk factor, and 35% of cases were not associated with either risk factor. These data suggest that antibiotic resistance is multifactorial.

12

Abstracts ID2

CIPROFLOXACIN IS MORE SUCCESSFUL AND COST-EFFECTIVE THAN TRIMETHOPRIM/ SULFAMETHOXAZOLE IN THE ORAL AMBULATORY TREATMENT OF WOMEN WITH UNCOMPLICATED ACUTE PYELONEPHRITIS LeLorier J, Benhaddad A, LeLorier Y, Castilloux A-M, Desgagné A Centre de recherche, Centre hospitalier de l’Université de Montréal, Campus Hôtel-Dieu, Québec, Canada BACKGROUND: To compare the success rate and costeffectiveness (from the perspective of the healthcare provider) of ciprofloxacin (C) and trimethoprim/sulfamethoxazole (TMP/SMX) in the ambulatory treatment of women with acute uncomplicated pyelonephritis (AUPN). METHODS: A randomized double-blind study was performed on women presenting to the emergency room with symptoms of AUPN. Treatment regimens consisted of either: C (500 mg) twice daily or TMP/SMX (160 mg/ 800 mg) twice daily. Duration of treatment was 14 days. The data necessary to calculate success rates and direct medical costs were collected. RESULTS: A total of 89 patients were randomized and included in the cost-effectiveness analysis. Thirty-six patients (C:19; TMP/SMX:17) were excluded from the success rate analysis for the following reasons: 2 due to renal abnormalities (C:11; TMP/SMX:1), 3 due to failed urine cultures (C:12; TMP/SMX:1), 9 due to resistance to TMP/SMX (C:5; TMP/SMX:4), 16 due to negative urine cultures (C:7; TMP/SMX:9), and 6 due to loss to follow-up (C:4; TMP/ SMX:2). Fifty-three evaluable patients remained (C:26; TMP/SMX:27). Treatment success was defined as a negative urine culture and absence of symptoms at the time of the follow-up visit. Overall success rate: C: 22 of 26; TMP/SMX: 16 of 27 (p 5 0.0405). Direct medical costs: C: $278.39; TMP/SMX: $546.15 Canadian (p 5 0.0048). CONCLUSION: Ciprofloxacin produces fewer side effects and has a higher cure rate than trimethoprim/sulfamethoxazole. As a result, and despite its higher acquisition cost, ciprofloxacin is more cost-effective than trimethoprim/sulfamethoxazole in the ambulatory treatment of AUPN in women.

ECONOMIC AND OUTCOMES STUDY RESULTS FOR CARDIOVASCULAR DISEASE

CV1

PRELIMINARY QUALITY-OF-LIFE EVALUATION OF THE GERMAN SAT TRIAL: COMPARISON OF SF-36, EUROQOL, TIME TRADE-OFF METHOD, AND WILLINGNESS TO PAY METHOD Kurscheid T, Hunsche E, Wendland G, Lauterbach KW Institute for Health Economics at the University of Cologne, Cologne, Germany

OBJECTIVE: To assess the effectiveness of an evidencebased treatment of obesity in a primary care setting, we conducted the SAT trial (Sibutramine in Adiposity Trial), a multicenter, double-blind, placebo-controlled, randomized parallel group comparison of a 54 weeks continuous therapy with Sibutramine 15 mg or placebo, in addition to a standardized nonpharmacological therapy consisting of a calorie-restricted diet, modification of eating behavior and lifestyle, and an activity program. Patients with a body mass index (BMI) of greater than 30 and less than 40 were included. The primary endpoint in this study is weight loss in kilograms between the last measurement of body weight (visit 11) and the measurement at the first visit. Secondary endpoints are, amongst other clinical parameters, health economic and quality of life data. METHODS: Quality of life is measured by four different instruments: the Short Form 36 (SF-36), the EuroQol, the time trade-off method (TTO), and the willingness to pay method (WTP). The SF-36 is a general, standardized and validated quality-of-life instrument, consisting of 36 questions that can be aggregated into eight health dimensions. The EuroQol is also a general, standardized and validated quality-of-life questionnaire, having the advantage of aggregating the five health dimensions into one qualityof-life index. Therefore, the EuroQol is frequently used in health economic evaluations to calculate quality-adjusted life-years saved (QALYs). In the TTO interview, patients are asked questions so that we may calculate a quality-oflife index for each patient (stratified by age). In the willingness to pay questionnaire, obese patients are asked how much they would be prepared to pay per month for an immediate and sustained weight loss of 10 kg from their current body weight. The answers are stratified by net income classes. RESULTS: Quality of life measured by the SF-36 has improved in all eight health dimensions at visit 2 compared to visit 1. Social functioning is especially impaired in obese patients. The quality-of-life index measured by the EuroQol has slightly improved from visit 1 to visit 2 over all age groups. The results of the quality-of-life index of the TTO interview and the EuroQol questionnaire are very similar except in the age group ,35 years. Willingness to pay correlates positively with net income. DISCUSSION: The results reveal the massive burden of obese patients.

CV2

COST-EFFECTIVENESS OF TREATING HIGHRISK INDIVIDUALS, AGED 45–65, WITH STATINS IN GERMANY FOR PRIMARY AND SECONDARY PREVENTION OF CORONARY HEART DISEASE Lauterbach KW, Hunsche E, Klewer U Institute for Health Economics and Clinical Epidemiology, University of Cologne, Cologne, Germany Cost-effectiveness of primary and secondary prevention of coronary heart disease in Germany was analyzed re-

Abstracts

13

garding the potential effects of treating all patients with hypercholesteremia age 45 to 65. METHODS: The current analysis is based on a model that approximates the discounted lifetime costs of all patients who presented themselves in 1996 to a German hospital for either primary or secondary events of coronary heart disease. The costs of this cohort in 1996 and up to their projected end of life were estimated, including all direct and indirect costs. Data for direct costs were generated from the federal bureau of statistics, sickness funds, retirement funds, hospital admission data sources, primary care panel physician data, and other sources. In calculating indirect costs, gender- and age-specific unemployment rates were considered. Additionally, the effects on revenues and expenses for sickness funds and other parts of the social security system were estimated. Thus, the model allows cost-effectiveness studies from the perspective of society as well as from the perspective of the social security system and parts of it, including the effects on retirement payments. RESULTS: Direct costs are approximately 31 billion US$ discounted at 4%. Indirect costs total 40 billion US$ discounted at 4%. The model was used to calculate the costeffectiveness of treating all patients for secondary prevention and all patients with a total cholesterol of 250 mg/dl irrespective of further risk factors in the age group 45–65 from the perspective the social security system and the perspective of society. For this age group, there is little difference between these two perspectives and a combined cost-effectiveness of 33,000 US$ per life-year saved. The results are highly sensitive for the price of the prescribed statins. CONCLUSION: Primary and secondary prevention can be cost-effective in the age group 45–65 for both genders, from both the perspective of society as well as the social security system.

CV3

UTILIZATION OF ACE INHIBITORS IN CONGESTIVE HEART FAILURE AMONG THE ELDERLY IN QUEBEC LeLorier J, Couture J, Blais L, Castilloux A-M Centre de recherche, Centre hospitalier de l’Université de Montréal, Campus Hôtel-Dieu, Québec, Canada BACKGROUND: In 1991, SOLVD demonstrated a decrease in hospitalization and mortality rates by using angiotensin-converting enzyme inhibitors (ACEI) in chronic congestive heart failure (CHF). METHODS: We conducted a study on a 10% random sample of subjects aged 65 years and older from the universal health insurance plan in Quebec. We studied all the elderly who had a dispensation of digoxin (0.125 mg/d) and furosemide (40 mg/d) (DF) from September 1991 to December 1994. RESULTS: Among 3,565 subjects, 61.7% used an ACEI at least once after the first dispensation of DF. Visiting a cardiologist was the strongest predictor of an ACEI dispensa-

tion (OR 1.58; 95% CI 1.44–1.74). The use of ACEI remained stable from 1991 to 1994 (p 5 0.92). In a subgroup of subjects free of ACEI and of any combination of digoxin/furosemide 1 year prior to the first DF dispensation (incident cases), 51.7% had at least one dispensation of an ACEI following the first dispensation of DF. Among these incident users of DF, 79.3% began the ACEI in the 3 months following the first dispensation of DF. CONCLUSION: In the elderly taking digoxin and furosemide, the use of ACEI is widespread. Since the publication of SOLVD in 1991, there has been no change in the utilization of ACEI over time.

CV4

IMPROVED MYOCARDIAL PROTECTION DURING CORONARY BYPASS SURGERY SHORTENS HOSPITAL STAY AND SAVES COSTS Mehlhorn U, Fattah M, Kuhn-Regnier F, Geissler HJ, Südkamp M, Horst M, Hekmat K, Dapunt O, de Vivie ER Clinic for Cardiothoracic Surgery, University of Cologne, Germany OBJECTIVES: We have recently shown that continuous coronary perfusion with warm blood enriched with the ultra–short acting b-blocker Esmolol (ES) improves functional and structural myocardial protection during coronary artery bypass grafting (CABG) as compared with conventional cardioplegia (CP). The purpose of the present study was to compare the myocardial protection techniques in terms of patient outcome and economical aspects. METHODS: We retrospectively analyzed the charts of 150 consecutive patients who received CABG using the ES technique; 150 patients matched for age, gender, preoperative left ventricular function, history of renal failure, and history of neurological symptoms who underwent CABG using the conventional CP-technique served as the control group. RESULTS: There were no significant differences between the groups with respect to perioperative myocardial infarction rate, need for positive inotropic medication, need for mechanical circulatory support, duration of mechanical ventilation, duration of intensive care unit stay, time of mobilization, postoperative renal failure, cardiac arrhythmias, neurological symptoms, infections or in-hospital mortality. However, ES patients were less frequently re-admitted to the intensive care unit (ES: 3 of 150 [2%] vs. CP: 13 of 150 [9%]; p 5 0.015) and discharged earlier from the hospital (ES: 12.3 6 4.8 days vs. CP: 13.5 6 3.8 days; p 5 0.009). In addition, the ES technique itself was more economical (drug costs for ES technique: about DM 120 per patient vs. CP technique: about DM 260 per patient). CONCLUSIONS: Our data suggest that improved myocardial protection during CABG using the ES technique may be a cost-saving alternative to the conventional CP technique due to both shorter hospital stay and lower procedural costs.

14

Abstracts CV5

ECONOMIC ASSESSMENT OF A PROPHYLACTIC TREATMENT OF RECURRENCE OF ISCHEMIC ACCIDENTS BY LYSINE ACETYLSALICYLATE Marissal JP, Selke B, Lebrun T CRESGE, Lille, France Despite its low cost and its proven efficacy in prevention of recurrence of ischemic events, acetylsalicylic acid is a strategy not used in France as frequently as it should be. OBJECTIVE: In order to convince practitioners and public authorities, we provide a cost-effectiveness analysis (CEA) of lysine acetylsalicylate (LA) 300 mg as compared to “do nothing.” METHODS: We used Garber and Phelps’s specification of CEA. Efficacy data were derived from the Antiplatelet Trialists’ Collaboration meta-analysis. We used French data sources to estimate the cost of the prophylactic treatment, the cost of treatment-related side effects, and the cost of nonfatal events to be prevented (ischemic stroke, myocardial infarction). The perspective taken into account for this CEA is that of the French Social Security system. Cost-effectiveness ratios (1996 US$) Endpoint Qualifying event Unstable angina Prior MI Prior ischemic stroke Stable angina

MI

Ischemic stroke

25,703 215 2610 4,375

190,711* 237 2176 15,749*

METHODS: MITRA (Maximal Individual TheRapy in Acute myocardial infarction) is a AMI registry of 54 hospitals in southwest Germany. The aim is a more frequent use of therapies improving prognosis. In the MITRA-1 study (6/94 to 1/97) the use of thrombolysis or PTCA, aspirin, beta blocker, or ACE-inhibitor was increased. The ongoing MITRA-2 study (6/98 to 6/99) is aimed at improving the use of CSE inhibitors to lower cholesterol after AMI, and includes a cost-utility analysis. RESULTS: The use of all four therapy forms increased during the study, especially with beta blocker and ACE inhibitor. This was associated with a significant decrease in hospital mortality (220%).

MITRA-1

Pilot phase

Study phase 1

Study phase 2

P

n 5 6068 Lysis and/or PTCA Aspirin Beta blocker ACE inhibitor Hospital mortality

6 mo 1298 55% 88% 33% 16% 16.3%

1 yr 2678 58% 94% 55% 58% 16.1%

1 yr 2062 60% 96% 62% 63% 13.1%

0.003 ,0.001 ,0.001 ,0.001 0.005

CONCLUSION: In clinical practice, many AMI patients do not adequately receive prognostic therapy. In the MITRA-1 study, the use of prognostic therapy was markedly reduced, as was the outcome. In the ongoing MITRA-2 study, the additional effect of an increased use of cholesterol reduction (CSE inhibitor) is being evaluated, together with a cost-utility analysis.

*Insignificant difference in efficacy between LA and placebo.

RESULTS: The results presented in the above table provide economic evidence of the interest of lysine acetylsalicylate in preventing ischemic accidents since the benefits produced by the prophylactic treatment almost exceed the costs induced by this strategy. CONCLUSION: As a consequence, such negative net costs induce savings for the Social Security system. Therefore, the prescription of acetylsalicylic acid and, in particular, lysine acetylsalicylate is medically effective and economically efficient.

CV6

OUTCOME OF ACUTE MYOCARDIAL INFARCTION PATIENTS WITH MAXIMAL INDIVIDUAL THERAPY (MITRA-STUDY) Dommke C, Schiele R, Fischer F, Wienbergen H, Voigtländer T, Glunz HG, Berg G, Gieseler U, Lauterbach KW, Senges J Herzzentrum Ludwigshafen; Universität Cologne, for the MITRA Study Group, Cologne, Germany Drugs improving prognosis in acute myocardial infarction (AMI) are underused in clinical practice, despite established therapy from randomized trials.

ECONOMIC AND OUTCOMES STUDY RESULTS FOR MENTAL HEALTH

MH1

ROAD TRAFFIC ACCIDENTS ARE ASSOCIATED WITH BENZODIAZEPINE USE Barbone F2, McMahon AD1, Davey P1, Morris AD, McDevitt DG, MacDonald TM1 1 Medicines Monitoring Unit (MEMO), University of Dundee, Ninewells Hospital & Medical School, Dundee, Scotland, 2 Chair of Hygiene and Epidemiology, DPMSC, University of Udine, Italy BACKGROUND: Psychomotor studies suggest that psychoactive drugs impair driving skills. We have examined the association between the use of psychoactive drugs and road traffic accidents (RTA). METHODS: We carried out a case-crossover study of drivers .18 years, resident in Tayside, Scotland, and who experienced a first road traffic accident recorded by Tayside Police between 1 August 1992 and 30 June 1994. Each driver had used one of the following drugs: tricyclic antidepressant (TCA), selective serotonin reuptake inhibitor (SSRI), benzodiazepines, or other psychotropics, be-

Abstracts

15

tween 1 August 1992 and the date of RTA. For each driver, the probability of having the RTA while exposed to a drug was compared with the probability while unexposed. RESULTS: Of 19,386 drivers with a first RTA during the study period, 1,731 were users of any study drug. On the days of the accidents, 189 subjects were taking TCAs, 84 were taking selective SSRI, 235 were taking benzodiazepines, and 47 were taking other psychoactive drugs. The treatment-group exposure relative risks were 0.93 (95% confidence interval [CI] 0.72–1.21) for TCAs, 0.85 (0.55–1.33) for SSRIs, 1.62 (1.24–2.12) for benzodiazepines, and 0.88 (0.62–1.225) for other psychotropics. The benzodiazepine risk decreased with increasing driver age and increased in subjects with positive breath tests for alcohol. A dose-response relationship was evident with benzodiazepines, the relative risk increasing with dose. CONCLUSION: Users of benzodiazepines were at increased risks of experiencing a road traffic accident; however, users of TCAs and SSRIs were not. Users of benzodiazepines should not drive.

MH2

NEW BENZODIAZEPINE USE IS ASSOCIATED WITH AN AGE-RELATED RISK OF HOSPITALIZATION WITH FRACTURE AND OTHER TRAUMA: A POPULATIONBASED STUDY Steinke D1, Davey P1, Elliott L2, Dowell J3, Morris AD1, MacDonald TM1 1 Medicines Monitoring Unit (MEMO), University of Dundee, Ninewells Hospital & Medical School, Dundee, Scotland; 2 Tayside School of Nursing & Midwifery; 3Tayside Centre for General Practice, Ninewells Hospital, Dundee, Scotland OBJECTIVE: There are limited data linking benzodiazepine prescribing to road traffic accidents (RTA) and hip fractures but “epidemiological evidence for increased risks of several kinds of accidents with benzodiazepine use remains inconclusive.” METHODS: We used a case control study. Between June 1993 and October 1995, 4019 people were admitted with fractures and 3186 people were admitted with trauma other than fracture from a fixed Tayside population of 319,465. Of the population as a whole, 19,304 people were dispensed benzodiazepines between June 1993 and October 1995 but had not been exposed in the previous 6 months. RESULTS: Benzodiazepine users had an increased risk of admission for fractures (odds ratio [OR] 1.9; 95% CI 1.7–2.1) and other trauma (OR 1.6; CI 1.4–1.8). The risk of fracture was markedly age-related. The numbers needed to treat (NNT) for one associated admission with fracture was 95 (CI 79–120) for the whole population, but the NNT fell from 199 (CI 111–918) at age 30–39 to 9 (CI 6–17) at age 90–99. The risk of other trauma was also age-related but showed a bimodal distribution. The NNT for one admission with other trauma were 163 (CI

126–234) for the whole population. At age 20–29 the NNT was 72 (CI 47–154), rising to 376 at age 40–49 and falling again to 61 (CI 38–167) at age 80–89. CONCLUSION: These results challenge the widely held belief that benzodiazepine drugs are safe with few serious side effects.

MH3

ASSESSING THE SASKATCHEWAN DATABASE FOR OUTCOMES STUDIES OF DEPRESSION AND ITS TREATMENT West S, Richter A, Nennstiel M, Mauskopf JA Research Triangle Institute, Research Triangle Park, NC, US OBJECTIVE: Because antidepressants are prescribed for numerous conditions, users of administrative databases for conducting research on depression must be cautious. We examined the Saskatchewan Health databases, containing three-digit ICD-9 codes, concurrent medications, prescriber information, comorbidities, and fee-for-service codes, to determine their suitability for conducting depression research. METHODS: We identified a cohort of individuals who began a “new” period of antidepressant use (no use 180 days prior) with at least one additional dispensation of the antidepressant. A stratified random sample of 600 individuals was chosen for medical record abstraction. The medical record abstraction classification of depression was the gold standard against which our definitions of depression were judged. We developed a primary definition of depression based on ICD-9 codes and tried to further refine it with various algorithms. Since so few patients were false-positive and/or false-negative in the primary classification, it was not feasible to split the data and use pattern recognition techniques. Instead, we developed medically probable scenarios using the explanatory variables and evaluated them for their sensitivity, specificity, and predictive values. RESULTS: Defining depression using the ICD-9 codes 296 (affective disorders), 309 (adjustment reaction), and 311 (depressive disorders) had a sensitivity of 71%, a specificity of 85%, a positive predictive value (PV1) of 86%, and a negative predictive value (PV2) of 70%. Using other definitions gave different results: limiting the number of false-negatives raised the sensitivity to 84% and the PV2 to 77%; limiting false-positives raised the specificity 90% and PV1 to 86%. The false-negatives had very similar characteristics to the true-positives. CONCLUSION: The ICD-9 definition will permit future studies of depression to be conducted using the Saskatchewan Health databases.

16

Abstracts MH4

ECONOMIC IMPACT OF AGITATED AND AGGRESSIVE BEHAVIORS OF ELDERLY INDIVIDUALS DWELLING IN THE COMMUNITY IN FRANCE: 1. EPIDEMIOLOGY Cal M-L1, Dartigues J-F2, Gérard D3, Léger J-M4 1 Groupe de Recherche en Économie de la Santé, Université Victor Segalen, Bordeaux, France; 2INSERM U330, Université Victor Segalen, Bordeaux, France; 3Laboratoires Synthélabo, Meudon-La-Forêt, France; 4Centre Hospitalier Spécialisé Esquirol, Limoges, France OBJECTIVE: The aim of this study was to establish the epidemiological features of agitation and aggressiveness in patients aged 65 years or older and living at home and what the etiological factors might be, to begin a cost analysis in the absence of published data. METHODS: Prevalence was estimated, first, in a random sample of 3777 elderly subjects (the PAQUID cohort), using DSM-III criteria and psychologists’ reports, at different timepoints; second, in the patients of a nationally representative sample from 159 general practitioners, from diagnoses based on behaviors and clinical scrutiny. A national retrospective study including 410 patients was conducted to investigate the consequences of symptoms such as somatic troubles, social disability and dependence. Disability was also evaluated in the PAQUID cohort with the AGGIR scale used to identify patients eligible to a dependence-specific benefit. Risk factors (age, gender, medical variables, etc.) were examined. RESULTS: In 1997, the prevalence of agitation and aggressiveness ranged between 0.19% and 0.34% of the French population. Of concerned subjects, 55% presented two to three somatic troubles, 97% a restriction of social life, and 85% a restriction of daily activities. Prevalence depended positively on age. Men exhibited more symptoms than women (p , 0.0001). Agitation and aggressiveness were mainly associated with dementia and psychiatric diseases, but demented subjects showed more physical agitation (p , 0.02), social disability, and dependence (p , 0.0001). CONCLUSION: The economic impact of agitated and aggressive behaviors of elderly individuals should be great due to the high degree of associated disability, and it will likely increase in the future.

MH5

ECONOMIC IMPACT OF AGITATED AND AGGRESSIVE BEHAVIORS OF ELDERLY INDIVIDUALS DWELLING IN THE COMMUNITY IN FRANCE: 2. COST ANALYSIS Cal M-L1, Prudhomme M-L2 1 Groupe de Recherche en Économie de la Santé, Université Victor Segalen Bordeaux 2, France; 2Synthélabo Groupe, Le Plessis-Robinson, France OBJECTIVE: This study was designed to estimate the cost incurred by French Social Security due to agitated

and aggressive behaviors of patients aged 65 years or older, living at home. METHODS: Economic appraisal was performed on a representative sample of 410 patients included by 212 general practitioners in a national specific retrospective survey. The bottom-up approach to calculating cost of those psychotic symptoms for total population was based on our preliminary estimation of their prevalence. The whole direct medical costs were taken into account so as to discover the medical costs of dependence associated with behavior troubles. The main determinants of costs were examined through the various profiles of patients we pointed out. RESULTS: In 1997, the average cost to the French Social Security of an agitated and aggressive old patient dwelling in the community was estimated at 23,513 6 1,665 FF. It varied with the age of patients, from 18,392 6 2,642 FF in the 65–78 year age bracket to 21,920 6 3,447 FF in the 88–99 year bracket, with a peak in the group of patients aged 79–87 (29,724 6 2,675 FF) (p , 0.01). The average cost of demented patients (27,791 6 2,239 FF) appeared to be significantly higher than that of nondemented (18,410 6 2,445 FF) (p , 0.005). Healthcare services at home for dependent patients were the most costly category of service. CONCLUSION: Our results highlight the economic interest of better controlling behavior disorders with special emphasis on demented patients in order to prevent them from becoming dependent.

MH6

COST-UTILITY ANALYSIS OF RISPERIDONE IN CHRONIC SCHIZOPHRENIA Iskedjian M, Oh P, Einarson TR, Lanctôt K, Addis A University of Toronto, Toronto, Canada BACKGROUND: Schizophrenia imposes a heavy economic burden due to its early onset, high morbidity and mortality, chronicity, and limited response to traditional pharmacotherapy. Newer antipsychotics such as risperidone improve both negative and positive symptoms, and impact on economic and quality-of-life measures. However, the acquisition cost of risperidone is higher than that of traditional agents. An economic evaluation was performed to compare risperidone with oral haloperidol, depot haloperidol, and fluphenazine, to assess relative differences in costs and consequences. METHODS: A cost-utility analysis was performed using a decision analytic model with the perspective of the Ontario government (Canada). Probabilities including success rates, dropouts due to adverse events, and EPS were determined from a systematic review of the literature. Costs were obtained from the provincial formulary and fee schedules, standard lists for community resources, and hospital case costing. Utilities were measured in 25 stable patients with schizophrenia, using the Standard Gamble technique, and combined with the clinical outcomes into a measure of quality-adjusted life-years (QALYs).

Abstracts

17

RESULTS: Risperidone dominated all comparators, having the highest clinical success rate (67%), greatest number of QALYs (0.89), and lowest expected costs (CDN $69,885) over a 1-year period. Fluphenazine had the highest expected cost (CDN $82,264) and lowest number of QALYs (0.85). Both oral and depot haloperidol were associated with higher total expected costs and lower number of QALYs than risperidone. CONCLUSIONS: The use of risperidone in place of haloperidol in Canada would be associated with annual savings of CDN $832 million in hospital expenditures, CDN $113 million in incremental drug expenditures, and CDN $180 million in annual incremental community care expenses. Prospective validation of our findings, as well as comparisons with the US and Europe, may be warranted.

ECONOMIC AND OUTCOMES STUDY RESULTS FOR DIABETES AND INFECTIOUS DISEASE

D1

SELF-MONITORING OF BLOOD GLUCOSE IN DIABETES: IS IT WORTH IT? Evans JMM, Boyle DIR, Davey P, Morris AD The DARTS/MEMO Collaboration, Dundee University, Dundee, Scotland OBJECTIVES: To investigate patterns and costs of selfmonitoring of blood glucose, and its effect on glycemic control, among insulin-using diabetic patients in Tayside, Scotland. METHODS: The population-based DARTS diabetes database was used to identify all insulin-using patients diagnosed with type 1 or type 2 diabetes prior to 1993. The numbers of glucose-monitoring test strips dispensed in 1993–1995 were determined from the MEMO-dispensed prescribing database, and average costs per patient calculated. The effect of self-monitoring on glycemic control was investigated in a linear regression model (with HbA1c as the outcome variable). Independent variables were numbers of strips dispensed in the 6 months prior to the HbA1c reading, age, sex, and deprivation. RESULTS: Among 807 patients with type 1 diabetes, 15% obtained no test strips, 20% obtained enough strips to test glucose daily, and 1% for four times daily. The corresponding figures for the 1,240 insulin-using patients with type 2 diabetes were 30% (no strips) and 12% (daily tests). Age and deprivation influenced strip uptake. Average costs/patient/year were £64.40 for type 1 diabetes and £44.80 for type 2 diabetes, compared with £409 for self-monitoring four times daily (recommended in the DCCT). For 258 patients with type 1 diabetes who had HbA1c values recorded, there was an association between strip uptake and glycemic control (p , 0.001), with an estimated decrease in HbA1c of 0.7% for every 180 test strips dispensed. There was no such association for 529 insulin-using patients with type 2 diabetes. CONCLUSIONS: Self-monitoring of blood glucose im-

proves glycemic control in type 1 diabetes, but a minority of patients self-monitor regularly. Costs are moderate compared with other costs of diabetes care.

D2

PATIENT-HELD INSTRUMENTS FOR RECORDING AMBULATORY CARE RESOURCE USE: EXPERIENCE IN PATIENTS WITH DIABETIC FOOT ULCERS Constenla D1, Chambers M1, Sigler C2 1 MEDTAP International Inc., London, UK; 2Parke-Davis Pharmaceutical Research, Ann Arbor, MI, US BACKGROUND: Patients with diabetic foot ulcers require long-term management by community-based healthcare providers. To determine the amount of health resources used by patients with foot ulcers, we constructed a patient-held booklet for recording ambulatory care resource use alongside a multinational trial (contacts, dressings, topical medications). OBJECTIVE: To evaluate the acceptability and completeness of the booklet. METHOD: Thirty-seven patients in five study centers (in the UK and Spain) were asked to use the booklet for 4 weeks. Local center coordinators interviewed patients and clinicians and transferred data from the booklet onto the trial economic case report form (CRF). RESULTS: Thirty-four (92%) patients returned the booklet to local coordinators; two had been admitted to hospital; one was lost to follow-up. Of patients, 91% found the booklet easy to use, and 94% remembered to carry the booklet with them most of the time. Thirty (91%) participating clinicians reported that the booklet was easy to use. Comments about the booklet included: (a) size of the text: too small; (b) volume: too bulky; (c) look-up list of dressings and topical medications: difficult to use. Of 277 contacts with clinicians reported in the booklets, 256 (92%) were transferred to the CRF. Participants recorded 1,004 dressing changes and 1,065 applications of topical medications in the booklets, 95% and 98% of which were transferred to the CRF. The number of contacts recorded in the booklet was found to be consistent with an independent report from participating clinicians in 18 out of 19 cases for which information was provided. CONCLUSION: The patient-held booklet is an acceptable data collection instrument, which with minor modifications is suitable for use in ambulatory care settings.

18

Abstracts D3

ANTITHYROID DRUGS OR RADIOIODINE AS FIRST-LINE THERAPY OF GRAVES’ HYPERTHYROIDISM IN A COUNTRY WITH STRICT REGULATIONS FOR RADIATION PROTECTION? Dietlein M1, Moka D1, Hunsche E2, Dederichs B1, Lauterbach KW2, Schicha H1 1 Department of Nuclear Medicine, University of Cologne, Cologne, Germany; 2Institute for Health Economics, University of Cologne, Cologne, Germany OBJECTIVES: Radioiodine therapy has become more economic in Germany since the new recommendations by the Federal German Radiation Protection Committee for patient discharge. METHODS: Costing models for primary conservative or primary radioiodine therapy of Graves’ hyperthyroidism included the overall costs from first diagnosis to regular follow-up care over 30 years. The costs of hospitalization for radioiodine therapy were calculated for 300 patients, discharged with 250 MBq I-131 residual activity. Sensitivity analyses took into account the relapse rate of conservative or radioiodine therapy, use of diagnostic tests, level of health insurance, drops in productivity, and a discount factor. RESULTS: Cost estimates based on life expectancy ranged from 6682 DM to 41,216 DM for primary conservative treatment and from 7811 DM to 35,430 DM for radioiodine therapy in Germany, depending on the costing model used. The maximum cost savings amounted to 1129 DM for conservative treatment and 5786 DM for radioiodine therapy. Antithyroid drugs were considered cost-effective when they achieved relapse rates of 50% or less, a cut in the number of tests needed and reduced working hours. Failure to meet any one of these conditions makes radioiodine therapy most cost-effective in spite of hospitalization (mean 4.6 days). If the patient received long-term antithryoid drugs for repeated relapses, the costs can run to 7849 DM more than radioiodine therapy. CONCLUSIONS: Radioiodine is a cost-effective, firstline therapy in patients with a special risk of relapse after primary conservative therapy (goiter, younger patient, persistent elevated TSH receptor antibodies or Tc-uptake) even in a country with strict regulations for radiation protection.

D4

MEASURING THE IMPACT OF DIABETIC FOOT ULCERS ON QUALITY OF LIFE FROM THE PATIENT’S PERSPECTIVE Ward J1, McNulty P2, Abetz L3, Sutton M1, Brady L2 1 Johnson & Johnson, Raritan, NJ, US; 2Royal Hallamshire Hospital, Cheshire, UK; 3MAPI Values, Cheshire, UK OBJECTIVE: Foot ulcers are a common complication of diabetes and are estimated to affect 15% of patients with

diabetes during their lifetime. Lower extremity ulcer care represents a significant proportion of the cost of diabetes; around 25% of the cost of inpatient care for these patients is estimated to be due to foot ulcers, primarily due to infections and amputations. To examine the relative burden of foot ulcers over and above diabetes alone, this study utilized a comprehensive measurement strategy. METHODS: First, patient interviews and focus group discussions aided in the development of a quality of life assessment tool, the Diabetic Foot Ulcer Scale (DFS). RESULTS: Foot ulcer patients reported that loss of mobility caused by the non-weight-bearing treatment regimens placed many restrictions on their daily activities, leisure pursuits and social interactions, resulting in a substantial amount of frustration and anger. The uncertainty of whether lower extremity ulcers will ever heal or will result in amputation is also a source of worry to patients. Patients also often feel isolated and a burden to others. Thus, the interviews yielded 11 domains which were found to be impacted by foot ulcers: Leisure, Physical Health, Daily Activities, Emotions, Non-compliance, Family, Friends, Treatment, Satisfaction, Positive Attitude, and Financial. The 58 item DFS was developed and assessed for its face and content validity by patients and clinicians. Finally, the DFS was administered to 173 patients with diabetes to assess the psychometric properties of the DFS. The DFS met or exceeded the minimum standard criteria for reliability, construct validity and clinical validity. CONCLUSIONS: The wider application of the DFS in foot ulcer management will provide clinicians with useful information on the total burden of lower extremity ulcers on patients, as well as giving them specific areas to target with support and counseling. The impact of different treatment regimens can also be compared from the patient’s perspective, thus allowing more efficient decisions on the design and application of different treatment strategies.

IM1

ASSESSING THE VALUE OF EPOETIN-a VERSUS AUTOLOGOUS BLOOD DONATION IN ORTHOPEDIC SURGERY: A FEASIBILITY STUDY OF CONTINGENT VALUATION Persson U1, Martens L2 1 IHE, Swedish Institute for Health Economics, Lund, Sweden; 2 ICOM Health Economics, Johnson & Johnson, Raritan, NJ, US OBJECTIVE: This study assessed the feasibility of the contingent valuation method for measuring the monetary value to patients of avoiding allogeneic transfusions during orthopedic surgery. METHODS: A survey was administered to 400 patients randomized to receiving Epoetin-a preoperatively or undergoing autologous blood donation (ABD) in a US trial in 1996/97. The survey was administered after patients received Epoetin-a or underwent ABD. The chance of avoiding allogeneic transfusion was stated as 85% and 75% in the Epoetin-a and ABD groups, respectively. Participants were asked whether they would pay a particular

Abstracts

19

price out of their own pocket in order to obtain treatment with Epoetin-a rather than undergo ABD. One of six prices was randomly assigned to each participant ($100, 500, 1000, 1500, 2000, 2500). Internal consistency was examined by correlating willingness to pay (WTP) values with risk perceptions and patient characteristics. RESULTS: Mean WTP for Epoetin-a was $1845 in the Epoetin-a group and $1309 per patient in the ABD group. In multivariate analysis, WTP varied with fear of transfusion and education level, as expected. The probability of purchasing Epoetin-a continuously decreased with price in 10 of 12 observations. CONCLUSIONS: Our findings indicate high internal consistency of the survey. Patients indicated considerable WTP for Epoetin-a to reduce the risk of allogeneic transfusion relative to ABD. As indicated by the higher willingness to pay in the Epoetin-a group, patients may perceive autologous blood donation to be more inconvenient than they rate it once they have experienced the procedure.

IM2

EPOETIN-a REDUCES POSTOPERATIVE LENGTH OF STAY IN PATIENTS UNDERGOING KNEE REPLACEMENT Martens L, Gagnon D, Parasuraman B ICOM Health Economics, Johnson & Johnson, Raritan, New Jersey, US OBJECTIVES: A US multicenter study of Epoetin-a versus pre-operative autologous donation (PAD) in patients undergoing major orthopedic surgery showed that Epoetin-a reduced the risk of allogeneic transfusion. In addi-

tion, patients receiving Epoetin-a had higher postoperative hemoglobin levels than PAD patients. We examined 1) whether higher postoperative hemoglobin levels are associated with a reduced postoperative length of stay; and 2) the relationship between treatment assignment and postoperative stay. METHODS: Analyses were stratified by type of surgery (hip or knee). Multivariate analysis was used to explore the relationship between postoperative hemoglobin levels and postoperative length of stay. The relation between treatment assignment and postoperative stay was examined using both univariate and multivariate analysis. RESULTS: We found a relationship between postoperative hemoglobin levels and postoperative length of stay in the patients undergoing knee replacement, but not among hip patients. In the univariate analysis, there was no difference in postoperative stay between the Epoetin-a group and the PAD group among hip patients 4.91 (2.07 days) versus 4.90 (2.66 days). Among knee patients, the postoperative length of stay in the Epoetin-a group was 4.27 (1.60 days, significantly lower than in the PAD group, 4.70 (1.80 days [p 5 .044]). Multivariate analysis confirmed these findings. CONCLUSIONS: In addition to reducing the exposure to allogeneic blood, Epoetin-a significantly reduced postoperative length of stay when compared to pre-operative autologous donation among patients undergoing knee replacement surgery. The absence of a similar effect among patients undergoing hip replacement may be explained by the finding that postoperative hemoglobin level is not a predictor of postoperative length of stay among these patients.

Volume 2 • Number 1 • 1999 VALUE IN HEALTH

Contributed Poster Presentations

PCA2

CANCER AND AIDS

PCA1

PHARMACOECONOMIC EFFICIENCY OF THE USE OF PROTEASE INHIBITORS FOR HOSPITAL INPATIENTS WITH AIDS Rahmouni S, Peyron F, Flori A, Moreau J, Charbit JJ, BuèsCharbit M, Balansard G North Hospital, Marseilles, France OBJECTIVE: Assessment of the impact of introduction of protease inhibitors (April 1996) on the global expenditure in an infectious unit providing classical hospitalization. METHOD: A prospective study has been in progress since 1990 in an AIDS unit (20 beds) of F. Houphouët Boigny Hospital. All information was entered into a database software (Microsoft Access 2.0) in order to facilitate the data acquisition and the statistical analysis. Admissions were classified according to 31 different diagnoses in order to provide an average cost for each. 1558 patients (2717 admissions) were enrolled for whom drug costs and details of hospitalization were analyzed. RESULTS: In 1997, when the majority of inpatients were receiving combination therapy using three antiretroviral agents, we noticed a huge decrease in the number of patients and stays (41% and 45%, respectively, between 1995 and 1997). Correlating with these results, the total drug cost dropped by a factor of 3, linked to a lower use of antibiotics and oncology/immunology drugs in the same period. However, this decrease must be interpreted including costs of the antiretroviral therapy that is mainly distributed in the daycare unit (multiplied by 5 between 1995 and 1997). The mean annual cost per stay decreased significantly (p 5 0.0376). During the same period, the cost per prescription was relatively stable (296 FF in 1995 and 241 FF in 1997). All these findings are associated with a lower mean cost for some opportunistic infections such as multiple infections (12,658 FF per stay in 1992 and 2915 FF in 1997). CONCLUSION: Our results highlight several interesting points concerning treatment of HIV infection. Further studies are needed taking into account the overall cost of hospitalization in relation to improvement in the quality of life.

© ISPOR 1098-3015/99/$10.50/20 20–50

THE COST-EFFECTIVENESS OF ITRACONAZOLE SOLUTION IN THE TREATMENT OF ORAL CANDIDIASIS IN HIV/AIDS PATIENTS: AN INTERNATIONAL COMPARISON McGann ME1, Halpern M1, Rance L2, Ajax MJ1 1 MEDTAP International, Seattle, WA, US; 2Janssen-Ortho, Inc, North York, Ontario, Canada OBJECTIVES: To compare the costs and outcomes of therapy with itraconazole solution (IS) versus alternative therapies in treating oral candidiasis (OC) in HIV/AIDS patients in the US and Canada. METHODS: Three clinical decision-analytic models were developed for each country to evaluate the cost-effectiveness of treating mild, moderate, and severe OC in HIV/ AIDS patients. Data for the models were derived from published studies and a survey of American and Canadian clinical experts. Outcomes included adverse events, lack of response to therapy, relapse, and cure. RESULTS: First-line treatments of mild OC in Canada were nystatin suspension or nystatin suppository. Fluconazole replaced nystatin suspension in treating moderate OC. Mild and moderate OC were treated with fluconazole and clotrimazole in the US. First-line treatments for severe OC were fluconazole and ketoconazole in both countries. The expected cost of treating each severity level of OC was lower in Canada. In both countries, therapy with IS yielded greater costs and cure rates than clotrimazole or nystatin, respectively, resulting in cost-effectiveness values for IS of $125 to $1891 (all costs in 1998 US dollars). IS was dominant over fluconazole in Canada, and had a cost-effectiveness ratio relative to fluconazole of $300 to $364 in the US. Sensitivity analyses indicate the above results were robust with respect to changes in model parameters. CONCLUSIONS: Alternative therapies in both countries for OC were different but model results were similar. Treatment with itraconazole solution led to greater cure rates in all scenarios, was cost-effective in the US, and was cost-effective or cost-saving in Canada. As such, itraconazole solution is likely an appropriate choice for treating oral candidiasis in this population.

20

Abstracts

21 PCA3

ECONOMIC CONSEQUENCES OF MONOCLONAL ANTIBODY 17-1A ADJUVANT THERAPY OF RESECTED DUKES’ C COLORECTAL CARCINOMA IN GERMANY Schädlich PK1, Lützelberger U2, Brecht JG1, Volmer T2 1 InForMed GmbH, Hamburg, Germany; 2Glaxo Wellcome GmbH, Hamburg, Germany OBJECTIVES: The purpose of our analysis was to quantify the cost-effectiveness of monoclonal antibody 17-1A (MAB) adjuvant therapy versus observation only in patients with resected Dukes’ C colorectal carcinoma (CRC), from the perspective of a German third-party payer. METHODS: This retrospective analysis used the costeffectiveness ratios “additional cost for MAB per life-year gained (LYG)” and “per disease-free year gained (DFYG)” as target variables and was based on a modeling approach. Resource use in adjuvant therapy and in treatment of recurrence was collected from a survey among specialist health care providers. Costs were determined by multiplying utilized resource items by the price or tariff of each item. LYG and DFYG after 5 years’ follow-up were obtained from the Kaplan-Meier curves of the original patient data in the clinical trial, representing a mix of 1.4/1 for colon cancer/rectal cancer. Costs and effectiveness were discounted by 5% annually. RESULTS: In the base-case analysis, a scenario of 30% inpatient and 70% outpatient MAB adjuvant therapy was applied. There were additional costs for MAB of 11,300 DM per LYG and 11,000 DM per DFYG. In the sensitivity analysis, the impact of different treatment modalities was investigated. Exclusive inpatient administration revealed additional costs for MAB of 470 DM per LYG and 450 DM per DFYG, whereas exclusive outpatient administration revealed additional costs of 15,940 DM per LYG and 15,520 DM per DFYG. CONCLUSIONS: Costs of MAB adjuvant therapy in patients with CRC are counterbalanced by avoided treatment costs of recurrence to great extents. The additional cost for MAB per LYG and per DFYG vary considerably depending on where MAB is administered.

PCA4

OUTCOME OF MONOCLONAL ANTIBODY 17-1A ADJUVANT THERAPY OF RESECTED COLON CANCER DUKES’ C IN GERMANY: A SUBGROUP ANALYSIS Brecht JG1, Volmer T2, Schädlich PK1, Lützelberger U2 1 InForMed, GmbH, Hamburg, Germany; 2Glaxo Wellcome GmbH, Hamburg, Germany OBJECTIVES: It has been argued that monoclonal antibody 17-1A (MAB) is more effective in colon cancer (CC) than in colorectal cancer (CRC) Dukes’ C. The purpose of this analysis was to estimate the cost-effectiveness of MAB in CC versus observation only from the perspective of a third-party payer in Germany.

METHODS: The investigation was based on a subgroup analysis of original data from patients with CC within the clinical trial in patients with CRC. The modeling approach used the cost-effectiveness ratios “additional cost for MAB per life-year gained (LYG)” and “per disease-free year gained (DFYG).” LYG and DFYG after 5 years were obtained from the corresponding Kaplan-Meier curves. Relapse prevention with MAB was nearly twice as effective in CC as in CRC. Economic consequences of resource use, collected from a survey, were determined by multiplying utilized items by the prices or tariff of each item. Costs and effectiveness were discounted by 5% per year. RESULTS: The base-case analysis revealed additional cost for MAB of 5,380 DM per LYG and 4,860 DM per DFYG, when a scenario of 30% inpatient and 70% outpatient administration was applied. In the sensitivity analysis, there were net savings of 3,820 DM per LYG and 3,450 per DFYG in the case of exclusive inpatient administration and additional cost for MAB up to 9,320 DM per LYG and 8,420 DM per DFYG, when MAB was administered on an outpatient basis exclusively. CONCLUSIONS: MAB is very cost-effective in patients with CC. In inpatient administration, it results in net savings for the third-party payer while improving patients’ prognosis at the same time.

PCA5

QUALITY OF LIFE IN HODGKIN’S DISEASE: RESULTS FROM THE EORTC AND GHSG TRIALS Flechtner H1, Rufer JU2 1 Clinic for Child and Adolescent Psychiatry; 2Medical Department I, University of Cologne, Cologne, Germany OBJECTIVES: 1) Investigating and comparing the quality of life of patients with Hodgkin’s Disease (HD) on different dimensions during active treatment and follow-up; 2) identifying longitudinal patterns of QoL dimensions during re-adaptation to normal life; 3) obtaining crosscultural comparisons between the participating countries and study groups (EORTC and GHSG). BACKGROUND: Available trials addressed the negative long-term sequelae in HD survivors, but longitudinal data—relating the outcome to various treatment-related variables and to the process of re-adaptation into normal life after end of treatment—are lacking. METHODS: Within the randomized EORTC trial H8 for clinical stage I-II HD, which started in September 1993, patients receive a QoL questionnaire for completion at each follow-up visit during the first 5 years after the end of active therapy. The corresponding HD8: Study of the GHSG employs the assessment of QoL during and after active treatment periods. Within both studies the EORTC QLQ C 30 is used for quality of life assessment incorporated in the H8/HD8 questionnaire for quality of life in HD, which addresses, in addition, the aspects of fatigue/ malaise, sexuality, specific side effects, and retrospective evaluation of treatment. RESULTS: To date, over 2500 questionnaires of 900 pa-

22

Abstracts

tients from both cooperative groups and 10 feasibility problems occurred in the multicenter setting. Patients reported mixed patterns of responses regarding the different domains of QoL. Treatment-related effects could be observed. In general, severe limitations in perceived QoL during the first 3 years of follow-up were reported. In particular, levels of emotional strain and fatigue remained high after the end of active treatment. Women in general reported a lower QoL than men. CONCLUSION: QoL assessment within international multicenter trials in HD proved feasible within the two differently organized study groups of EORTC and GHSG. The applied H8/HD8 QoL questionnaire was able to detect distinct patterns of QoL in subgroups of patients.

PCA6

COST-EFFECTIVENESS OF GEMCITABINE AS FIRST-LINE THERAPY FOR PATIENTS WITH ADVANCED PANCREATIC CANCER Trippoli S, Messori A Laboratorio SIFO di Farmacoeconomia, c/o Centro Informazione Farmaci, Azienda Ospedaliera Careggi, Florence, Italy OBJECTIVES: Gemcitabine is a new anticancer drug that has recently been proposed for the treatment of advanced pancreatic cancer. The therapy with gemcitabine has been reported to confer a survival advantage in comparison with fluorouracil. Because gemcitabine is expensive, we estimated the cost per life-year gained using gemcitabine for this clinical indication. METHODS: The clinical material utilized in our analysis was derived from a randomized clinical trial in which the survival of patients receiving gemcitabine was compared with that of patients receiving fluorouracil. To obtain an estimate of effectiveness, the survival curves published in the trial were analyzed using the Gompertz methodology. Gemcitabine acquisition cost was based on wholesale price in Italy, in the UK and in the US. The overall cost of treatment is presently being estimated by collecting individual data on the use of resources and morbidity costs. RESULTS: The analysis of the survival curves showed that the mean survival of patients treated with gemcitabine was 6.29 months, while the corresponding value for patients receiving standard treatment was 3.20 months (both values include discounting at an annual rate of 3%). The survival gain for the gemcitabine group was 2.9 months per patient. Our analysis of cost data for these two treatments is in progress, but preliminary data show that the incremental cost is less than $20,000 per patient. CONCLUSIONS: In pancreatic cancer, gemcitabine determines a survival gain of approximately 3 months. Its cost-effectiveness profile seems to be within acceptable values.

PCA7

COST ESTIMATION OF SEVERE NEUTROPENIA IN BELGIUM Annemans L1, Eynatten C1, De Smedt B2, Alwan A3, Vanschoubroek K4, Standaert B4 1 HEDM, Mechelen, Belgium; 2BDS Clinical Services, Mechelen, Belgium; 3MEDSTAT, Antwerp, Belgium; 4Amgen, Brussels, Belgium OBJECTIVE: To collect resource utilization and unit costs in Belgium for treatment of severe neutropenia of five chemotreated cancer types for which no reimbursement of prophylactic Neupogen use is available. These cancer types include multiple myeloma (MM), metastatic breast cancer, small cell lung cancer, non–small cell lung cancer, and bladder cancer. Another goal is to identify factors that may explain cost variation in treatment of severe neutropenia. METHODS: Retrospective data collection of bills from patients treated for severe neutropenia from nine hospital centers spread all over Belgium over a period of 1 year (1/96–1/97). Severe neutropenia was defined as a neutropenic episode requiring hospitalization. Only direct medical costs were considered from the perspective of reimbursement. Items collected included: hospital stay (duration and type), diagnostic procedures, drugs, transfusions, interventions; and medical consults. RESULTS: Seventy-nine patient bills were collected. More than 75% of the registered neutropenic events occurred during first three chemo cycles. Total mean cost (35 Bfr 5 1 US$) was $4.918 (Med.: $4.529; 95% CI: $4.303–$5.533). DISCUSSION: Multiple regression analysis demonstrates that two factors (hospital duration and institution type) explain 85% of the cost variation. The institution factor may, however, mask a bias, as all the hematological cancers (17%) with their more costly treatment procedures for neutropenia were treated in university centers. 85% of treatment cost is composed of only three factors: hospital stay (60%); drug regimen (15%); and lab tests performed (10%). CONCLUSION: The treatment cost of neutropenia varies. Its spread is skewed to the right. It is appropriate to calculate geometric mean cost and to introduce specific cost estimates per institution type or per cancer type (hematological/nonhematological).

PCA8

COST-MINIMIZATION IN THE USE OF COLONYSTIMULATING FACTORS (CSFs) BASED ON CLINICAL PREDICTION MODELS (CPMs) Lyman GH1,2, Kuderer NM3, Balducci L2 1 London School of Hygiene and Tropical Medicine, London, UK; 2H. Lee Moffitt Cancer Center and Research Center, Tampa, FL, US; 3Albert-Ludwigs-Universität, Freiburg, Germany Febrile neutropenia (FN) represents one of the leading dose-limiting toxicities of cancer chemotherapy. CSFs have been shown to reduce the incidence of FN in a variety of settings. Cost-minimization models have established FN

Abstracts

23

risk thresholds (FNRTs) for primary prophylaxis with CSFs. OBJECTIVES: To determine the impact of CPMs on FNRT estimates with the CSFs. METHODS: Cost-minimization models based on standard decision theory were generated, incorporating 1) probabilities and resource utilization from randomized clinical trials and 2) unit cost information from local institutional sources. Sensitivity analyses were performed, varying resource use, costs, and test performance characteristics. FNRTs were derived at which cost equivalence was reached in each model. The test performance of possible CPMs was assessed utilizing the likelihood ratio (LR) for discriminating high-risk patients with probability (prob[HR]). RESULTS: Baseline cost-minimization models were equated with CPMs with a LR 5 1.0. The FNRT for CSF use decreases as hospitalization length of stay (LOS) or cost/day increase. Any CPM associated with a LR1.1 generates lower FNRTs than under baseline conditions. Test efficiency of the CPMs improves with increasing LR1, with optimal performance observed between 2.0 and 4.0. In balanced CPMs with LR1 5 3.0, FNRTs are reached for costs/day of $1000, $1500, and $2000 at 0.21, 0.16, and 0.13 respectively. In unbalanced CPMs, FNRTs are ,10% at prob(HR) 5 0.4, 0.3, 0.2, and 0.1 at costs/day of $1700, $1000, $600, and $300, respectively. CONCLUSIONS: Efficient CPMs to identify high-risk patients should be able to further reduce costs and FNRTs for CSF use based on cost-minimization.

PCA9

QUALITY OF LIFE (QOL) OF PATIENTS WITH LOW-GRADE NON-HODGKIN’S LYMPHOMA (IGNHL) TREATED WITH FLUDARABINE (F) OR CYCLOPHOSPHAMIDE-VINCRISTINEPREDNISONE (CVP) Bérubé; S1, Meyer R2, Klasa R3, Shustik C4 1 Berlex Canada Inc, Québec, Canada; 2Hamilton Regional Cancer Centre, Hamilton, Canada; 3BC Cancer Agency, Vancouver, Canada; 4McGill Oncology Group, Montréal, Canada OBJECTIVE: To evaluate the QoL of Ig-NHL patients enrolled in a trial comparing the efficacy and safety of F and CVP. METHODS: The EORTC-QLQ-C30 (v1) questionnaire was completed by all French- and English-literate patients at baseline, during treatment, and after treatment. QoL was expressed as a global health status (QL) and five functional scores (FS); physical, emotional, social (SF), cognitive (CF) and role (RF), for which high values indicate better QoL. It was also expressed as eight symptom/single item (SSI) scores: fatigue, nausea, and vomiting (NV), pain, dyspnea (DY), insomnia, appetite loss, constipation (CO), diarrhea (DI), and financial difficulties (FI) for which lower values mean better QoL.

RESULTS: Of the 91 patients recruited (F 5 47, CVP 5 44), 84 completed the questionnaire at baseline, 74 during treatment, and 41 after treatment. All FS were high at baseline (70–90) and varied from 1 to 12 points during the study. All FS improved during treatment except CF and QL in the F group, which remained unchanged; and RF and SF in the CVP group, which deteriorated. The difference between groups reached statistical significance for SF (p 5 0.0076) but not for RF (p 5 0.4597). All SSI scores were low (10 to 40) and varied only slightly during the trial. Most scores improved except DY and CO in the F group and NV, CO, DI, and FI in the CVP group. CONCLUSIONS: QoL was high in these patients. The higher incidence of nausea, vomiting, and alopecia observed in the CVP group may explain the difference observed for SF, and, to a lesser extent, RF.

PCA10

THE RELATIONSHIP BETWEEN PROFILE-BASED QUALITY OF LIFE SCORES AND EUROQOL-5D SCORES IN BREAST CANCER SURGERY PATIENTS Namjoshi M, Taylor T, Hohl R The University of Iowa, Iowa City, IA, US OBJECTIVE: The objective of this study was to determine the relationship between profile-based cancer-specific quality of life dimension scores, and EuroQol-5D scores in a breast cancer surgery sample. METHODS: Seventy-five breast cancer surgery patients participated in the study. The patients completed the Functional Living Index-Cancer (FLIC), the Functional Assessment of Cancer Therapy-General Version (FACT-G), and the EuroQol-5D (EQ-5D) at the same visit, and the order of instruments was randomized. Scores on both the EQ-5D classification system, as well as on the thermometer, were obtained, and both scores were related to dimension scores on the FLIC and the FACT-G. Information on the age of the patient, the cancer stage, and the number of comorbidities was obtained through chart review. RESULTS: The mean age of patients was 60 (SD 5 10.7). The majority of patients (96%) were diagnosed with either stage I or stage II breast cancer. Over 90% of patients had up to three comorbid conditions, with hypertension being the most frequent. Mean scores on the EQ-5D classification system and the thermometer were 87.4 (SD 5 13.6), and 86.6 (SD 5 12.6), respectively. Ceiling effects were observed on the classification system as well as the thermometer. Both linear and double log regression results showed that quality of life dimension scores were significantly related to scores in both the EQ-5D classification systems and the thermometer. FLIC dimension scores made statistically significant contributions to the EQ-5D classification score, but not to the score obtained on the thermometer. FACT-G dimension scores did not contribute significantly to either the EQ-5D classification score or to the thermometer score. CONCLUSION: The results of this study show that can-

24

Abstracts

cer-specific quality of life dimension scores are significantly related to EuroQol-5D scores.

CARDIOVASCULAR DISEASE

PCV1

COST-EFFECTIVENESS OF LIPID MODULATOR AGENTS LeLorier J, Lacour A, Derderian F Centre de recherche, Centre hospitalier de l’Université de Montréal, Campus Hôtel-Dieu, Québec, Canada OBJECTIVE: In an era of cost-conscientiousness and budgetary restraints, the cost-effectiveness of lipid modulator agents must be considered. The study objective was to evaluate the cost-effectiveness ratio of lipid modulator agents, based on the agent’s effects on total cholesterol/ HDL-cholesterol ratio (TC/HDL-c) and on the cost of acquisition. METHODS: The choice of TC/HDL-c for the efficacy criterion was dictated by the overwhelming evidence that it is the most predictive parameter in assessing CHD risk when compared to other lipid/lipoprotein levels or ratios. Also, the evidence supporting the crucial role played by HDL is demonstrated in several studies, among which are the Framingham Study, LRC-CPPT, MRFIT, and PROCAM. The cost of acquisition of each medication was obtained from the Quebec Provincial Formulary (Liste de Medicament—July 1996). The meta-analysis of each dose of each medication’s effects on TC/HDL was based on a search generating 456 publications, from which 124 met the appropriate criteria. RESULTS: From these publications, on average, the population is 50.5 years old, with 62.5% males; the baseline TC/HDL is 7.3. The various doses of each medication were ranked into three categories of efficacy based on the percent decrease of the TC/HDL ratio: less than 20%, 20–30%, and greater than 30%. CONCLUSION: In the most effective category, the medications which showed the best efficacy to cost ratio were fluvastatin 60 mg (20% decrease per $ per day) and micronized fenofibrate 200 mg (19% decrease per $ per day), and then simvastatin 20 mg (15% decrease per $ per day).

PCV2

ADAPTING ECONOMIC ANALYSES FROM ONE COUNTRY TO ANOTHER: DO THE WOSCOPS UK FINDINGS HOLD IN BELGIUM? Huybrechts KF, Caro JJ The WOSCOPS Study Group, Caro Research, Boston, MA, US In a previous analysis, based on the West of Scotland Coronary Prevention Study (WOSCOPS), pravastatin was demonstrated to lower the risk of cardiovascular disease (CVD) in hypercholesterolemic men, at an economic

efficiency that is not prohibitive overall from the perspective of the National Health Service in the UK. For these results to provide guidance to those who set policy in other jurisdictions, they must account for differences among healthcare systems. OBJECTIVE: The applicability of these findings to the Belgian setting was investigated by integrating the WOSCOPS results and Belgian epidemiological, resource use, and cost data. METHODS: The Belgian Interuniversity Research on Nutrition and Health data were used to evaluate the baseline risk factor distribution and to partially validate the CVD exponential regression model derived from WOSCOPS. The rates at which subjects move for the first time from health to CVD were calculated, assuming the relative efficacy for pravastatin observed in WOSCOPS. The number of such transitions avoided was valued in terms of cost savings to the Belgian healthcare system and life-years gained. Resource use and costs were primarily derived from diagnosis-related expenditure in a subset of Belgian hospitals. Belgian life table data were obtained and substituted for the Scottish data to account for differences in life expectancy. RESULTS: The cost-effectiveness ratio remains well below USD 25,000 (BEF 37 5 USD 1) per life-year gained, the bound typically considered as “moderate to strong evidence for adoption and appropriate utilization.” CONCLUSION: Although the precise estimate depends on the specifics of the country, the variation does not have an impact on the treatment decision the initial study supports.

PCV3

EFFICACY OF CARDIAC REHABILITATION THERAPY: A PROSPECTIVE MULTICENTER COHORT STUDY Willich SN, Völler H, Binting S, Gohlke H, Hahmann H, Kleber F, Klein G, Krobot K, Bestehorn K The PIN Study Group, Institute for Social Medicine and Epidemiology, Charité University Hospital, Humboldt University of Berlin, Berlin, Germany To determine the efficacy of cardiac inpatient rehabilitation therapy (RT), 2442 consecutive patients (78% male, age z60 (610 years), and 22% female, age z65 (610 years) in 18 rehabilitation centers were examined in the period from January to June 1997. The primary indications for RT were myocardial infarction (56%), PTCA (6%), and CABG (38%). RT was initiated 34 days (median) after the acute event. During hospitalization (26 6 5 days), patients received standard health training and medical therapy. Diagnostic tests and standardized surveys were prospectively documented. The prevalence of conventionally defined cardiovascular risk factors at the start and end of the rehabilitation therapy is shown in the table below.

Abstracts

25 Admission (%)

Arterial BP .140/90 mmHg Smoking Cholesterol .250 mg/dl LDL cholesterol .100 mg/dl LDL/HDL cholesterol .2.5 Triglycerides .200 mg/dl Glucose .140 mg/dl BMI .30 kg/m2

24 39 20 87 86 22 14 18

Discharge (%) ** ** ** ** ** * ns *

8 5 5 67 70 15 11 15

* p , 0.01, ** p , 0.001, ns 5 not significant.

Subjective well-being (v. Zerssen scale) improved during RT from 15.7 6 8 to 11.8 6 8 (p , 0.001). CONCLUSION: Alterable cardiovascular risk factors and subjective quality of life are significantly improved by health training and treatment during standardized cardiac inpatient rehabilitation therapy. Follow-up studies are needed to assess the long-term course and to improve preventive strategies.

PCV4

THE ECONOMIC IMPACT OF HYPERTENSION STUDY FROM THE PISTOIA PROJECT Potenza A1, Berto P2, Feroci P1, Lattari P1, Pacini R1, Mundula V1, Papini D3, Reggio S4 1 ASL 3, Pistoia, Italy; 2Pbe consultants, Verona, Italy; 3School of Medicine, Firenze, Italy; 4Pfizer Italiana, Roma, Italy OBJECTIVE: Within the Pistoia Project (PP), a large screening study whose objective is the identification of hypertensive disease, organ damage, and related risk factors to activate actions of primary and secondary prevention, we conducted a preliminary cost of illness analysis to investigate if age, gender, and/or severity of hypertension could influence the use of healthcare services and therefore have a direct impact on costs. METHODS: Forty-five patients’ records were randomly selected from the GP’s screening forms of the PP: patients were either hypertensive or normotensive under treatment. Using healthcare codes, we computed hospital data (DRG code and cost), while pharmaceutical consumption (units and cost) was retrieved from the pharmaceutical regional database. The local Hypertension Centre retrieved number and typology of specialist consultations and procedures. Indirect costs were partly calculated from social security system certificates, and partly estimated under the hypothesis that each hospital stay and each specialist visits imply a loss of working days. RESULTS: In 1997, 45 patients generated a total cost of Lit. 29.9 million, with an average cost per patient per year of Lit. 664,282. Pharmaceutical care accounted for 45% of total cost, followed by hospital admissions (34%), productivity losses (16.4%), and specialist visits and procedures (4.3%). CONCLUSIONS: This preliminary analysis has demonstrated the potential of such an instrument to evaluate the cost of managing hypertension in a large area as the Pis-

toia province. Further research is needed on larger samples to identify patterns of resource use and to evaluate the influence of illness severity on the total cost of this disease.

PCV5

POTENTIAL COST SAVINGS OF THE 4 FRENCH INFINITI CATHETER IN DIAGNOSTIC CORONARY ANGIOGRAPHY Neighbors D1, Buckingham T1, Greiner W2, Schulenburg J2 1 Center for Economics Research, Research Triangle Institute, Research Triangle Park, NC, US; 2Centre for Health Economics and Health System Research, University of Hannover, Hannover, Germany OBJECTIVE: To estimate the economic value of the 4F catheter relative to the 6F in diagnostic coronary angiography. METHODS: A decision tree model was developed, based on German practice patterns and costs. Expected total costs of diagnostic coronary angiography using 4F and 6F catheters in both the ambulatory and inpatient settings were calculated and compared, together with measures of procedure success. Costs include the costs of catheters, contrast media, time in clinic, professional services, and costs to treat complications. Measures of procedure success, as observed in trials of 4F versus 6F, include fluoroscopy, procedure, compression, and immobilization times; complication rates; and catheter exchange rates. Complication rates are those observed in two large German observational studies. RESULTS: In the base-case scenario, the 4F catheter saves an average of 22 DM per procedure, despite a higher list price. These savings derive primarily from reduced need for contrast media. The 4F is associated with fewer complications; however, because of the low overall complication rate, the associated savings are minor. Given a lower likelihood of bleeding complications, the 4F may enable more procedures to be performed in the ambulatory setting. Assuming a modest increase in the proportion of ambulatory procedures results in an additional savings of 10 DM per procedure. CONCLUSION: Compared to the 6F catheter, the 4F is estimated to save an average of 22 DM per diagnostic coronary angiography because of a reduced need for contrast media and a lower complication rate. If the 4F catheter enables more procedures to be performed on an outpatient basis, the potential cost savings increase.

PCV6

MEASURING DISEASE-SPECIFIC SYMPTOM IMPACT: DEVELOPMENT OF THE ATRIAL FIBRILLATION IMPACT SCALE (AFIS) Turner R, Kocal M, Testa M PHASE V Technologies Inc, Wellesley Hills, MA, US; Harvard School of Public Health, Boston, MA, US

26

Abstracts

OBJECTIVE: Quality of life is not well studied in atrial fibrillation (AF), the most prevalent rhythm disturbance with significant impact on morbidity and mortality. We report the development of a brief disease-specific QoL instrument for use in clinical trials, surveillance studies, and clinical practice. METHODS: Fifty-one items, based on literature, patient focus groups, and a cardiology expert panel were administered via telephone interview to 73 chronic AF and 23 congestive heart failure (CHF), along with 25 angina control patients. RESULTS: Item reduction techniques produced the 20-item AFIS comprising four symptom scales (dyspnea, fatigue, palpitations syncope: 11 items) and three QoL scores measuring functional, social and emotional impact (9 items). Internal consistencies were acceptable for all symptom scales (Cronbach ~ 0.73–0.77) and the QoL scales (Cronbach ~ 0.74–0.92). AFIS Total Symptom Impact (SI) score explained 50% of the variation in each QoL impact score (R2 5 0.49 [functional], 0.44 [emotional], 0.55 [social]) and 75% (R2 5 0.74) of the total QoL impact (QI) score. We identified two symptom impact (SI) threshold scores: moderate AF impact (SI 5 40; sensitivity 0.86, 1-specificity 0.92) and major AF impact (SI 5 60; sensitivity 0.96, 1-specificity 0.98). Mild symptomatic CAF patients (SI , 40) reported little QoL interference. Moderate symptomatic and major symptomatic CAF patients had QI scores 200% and 665% worse than the mild CAF patients. SI scores accounted for 50% (R2 5 0.50) and nearly 25% (R2 5 0.22) of total QI variation in CHF and angina patients, respectively. CONCLUSION: Mild AF symptoms are well tolerated. However, moderate symptom impact (over 50% of patients) doubles the interference with QoL; while major impact (33% of patients) approached a seven-fold worsening of QoL. The Atrial Fibrillation Impact Scale appears to be a useful tool for predicting worsening QoL among atrial fibrillation patients and for informing treatment decisions.

cialist in each country. Native speakers produced two independent forward translations. These were reconciled and back-translated into English. The translations were tested for comprehension in a sample target population, compared and internationally harmonized. The developer of the SAQ clarified concepts underlying each item. RESULTS: Translation problems involved finding conceptual equivalents of typical US expressions and constructions and translating the original response categories. Examples for which a rationale of translation choice and modification to the original will be discussed, are the following: although activities such as “showering” and “swimming” are less common outside the US, satisfactory equivalents were found. The translation of “how satisfied are you that everything is being done. . .,” however, had to be rendered as “how convinced are you. . . .” It was impossible to reproduce the original response scale referring to “very, moderately, somewhat, a little, not at all.” Equivalent answers in the translation had to refer to “extremely, quite a bit, moderately, slightly, not at all.” For reasons of comparability, the original was changed accordingly. CONCLUSIONS: Before use in an international trial, rigorous cultural adaptation was essential to producing cross-culturally valid language versions of the SAQ. This process, entailing modifications to an already validated original, might make further validation necessary. Further responsiveness data will be generated by the clinical trial. Ideally, initial instrument design should include international input.

PCV8

IMPACT OF CLINICAL SENSITIVITY ANALYSIS IN AN EVALUATION OF PRIMARY PREVENTION OF CORONARY HEART DISEASE IN GERMANY Munro V1, Hutton J1, Glick H2, Schöffski O3 1 MEDTAP International Inc., London, UK; 2University of Pennsylvania, Philadelphia, PA, US; 3University of Hanover, Hanover, Germany

PCV7

CROSS-CULTURAL ADAPTATION OF THE SEATTLE ANGINA QUESTIONNAIRE (SAQ) IN 11 COUNTRIES Conway K1, Uzun V1, Spertus J2, Fletcher A3, Sagnier P4 1 MAPI Research Institute, Lyon, France; 2University of Missouri, Kansas City, MO, US; 3London School of Hygiene and Tropical Medicine, London, UK; 4Bayer AG, Pharma Europe, Leverkusen, Germany OBJECTIVE: With the increase of QoL assessment in clinical trials, it is necessary to have cross-culturally valid instruments to pool data across countries. METHODS: The SAQ, a 19-item instrument designed in US English, assesses five domains of coronary artery disease. Prior to use in an international trial, the original had to undergo cultural adaptation, giving rise to 11 language versions and a modified original. The cultural adaptation process involved the recruitment of a QoL spe-

OBJECTIVES: An analysis is being carried out to assess the impact of clinical parameters, over and above the treatment effect, in an evaluation of coronary heart disease (CHD) in German males. METHODS: This analysis is part of a study which uses the Coronary Heart Disease Risk Assessment Model (CHD RAM), a decision analytical model, to evaluate the costeffectiveness of cholesterol-lowering agents for the primary prevention of CHD in German males. Although the wider study considers data for men in eight 5-year age groups ranging from 35–39 to 70–74 years, the clinical sensitivity analysis focuses on the 50–55 age group for three doses of cerivastatin, a cholesterol-lowering agent. Baseline data were primarily obtained from the published literature and expert opinion. The clinical parameters—compliance rate profiles, smoking behaviour, mean body mass index and systolic blood pressure—will be assessed in terms of their impact on CHD risk and life years gained. This will be esti-

Abstracts

27

mated by initial low density lipoprotein (LDL) cholesterol and the initial ratio of total cholesterol (TC) to high density lipoprotein (HDL) cholesterol. RESULTS: Analyses will be conducted for each clinical parameter varying the parameters by appropriate amounts. This will assess the robustness of the model and the impact of key clinical parameters on baseline results. CONCLUSIONS: By separating behavioral factors from treatment effect, the study will determine if clinical parameters play a major role in primary prevention of CHD with cholesterol-lowering agents.

PCV9

COMPARISON OF THE EPIDEMIOLOGY AND COSTS ASSOCIATED WITH PRIMARY PREVENTION OF CORONARY HEART DISEASE IN FIVE EUROPEAN COUNTRIES Munro V1, Hutton J1, Mast O2, Sagnier P2 1 MEDTAP International Inc., London, UK; 2Bayer AG, Leverkusen, Germany OBJECTIVES: A study was undertaken to compare the epidemiology and costs associated with primary prevention of coronary heart disease (CHD) in males in France, Germany, Italy, Spain, and the United Kingdom. METHODS: Country-specific epidemiological and cost data were collected as part of an economic evaluation of CHD and analyzed for differences between countries. Epidemiological and economic data were collected for men in eight 5-year age groups (35–39 to 70–74 years). These data were primarily obtained from the published literature and complemented by expert opinion. RESULTS: Mean systolic blood pressure ranged from 125 to 132 for the 35–39 year olds and from 143.9 to 155.5 for the 70–74 year olds. Mean body mass index ranged from 24.7 to 26.9 for the 35–39 year olds and from 25.5 to 27.75 for the 70–74 year olds. The proportion of smokers ranged from 0.38 to 0.56 for the 35–39 year olds and from 0.2 to 0.29 for the 70–74 year olds. With adjustment by the respective health purchasing power parities, direct costs of CHD ranged from £11,515 to £20,431 for acute myocardial infarction, £8,484 to £17,898 for complicated angina, £606 to £10,227 for uncomplicated angina, and £305 to £9,734 for sudden and nonsudden death. CONCLUSIONS: The epidemiological data showed similar trends within the five countries with some observed differences for certain variables and/or age groups. Resource utilization data differed significantly between countries—a not surprising finding considering the profound differences in healthcare systems and treatment patterns across Europe. In addition, differences in currency values and unit costs make direct comparison of costs difficult.

PCV10

A HEALTH ECONOMIC EVALUATION OF HORMONE REPLACEMENT THERAPY IN BELGIUM Arikian SR1,2, Doyle JJ1,2, Casciano R1, Tarride JE1, Schoonveld E3, Casciano J1 1 The Analytica Group, New York, NY, US; 2Columbia University, School of Public Health, New York, NY, US; 3 Wyeth-Ayerst Laboratories, St. Davids, PA, US OBJECTIVE: We conducted a health economic and outcomes study of post-menopausal women in Belgium to quantify the benefits and risks associated with HRT. Using data provided by l’Université Catholique de Louvain, the analysis was designed to predict the long-term clinical and economic effects related to the conjugated estrogens/ medroxyprogesterone HRT (E/M HRT) regimen. METHODS: Hypothetical cohorts of uterine-intact postmenopausal women who were fully compliant with E/M HRT were compared with similar cohorts of post-menopausal women who did not use any form of HRT. Markov cohort simulation analysis, guided by a Delphi panel of local clinical and health economic experts, was employed to model the short- and long-term implications of compliance with E/M HRT. Eight distinct health states were considered in the analysis: healthy, osteoporosis, coronary heart disease, cerebrovascular disease, Alzheimer’s Disease, breast cancer, colorectal cancer, and death. RESULTS: Results of this analysis suggest that patients using E/M HRT therapy are expected to experience improved health outcomes and an increased life expectancy of more than 5 months per patient when compared those not on HRT. These results are achieved with a modest increase in cost of 8,415 BF, or 18,300 BF per life-year gained.

GASTROINTESTINAL AND URINARY TRACT DISEASE

PGIU1

WILLINGNESS TO PAY OF PATIENTS WITH PAINFUL-IRRITABLE BLADDER SYNDROME Budhiarso I1, Bushnell D1, Ramsey S2 1 Health Research Associates, Inc., Seattle, WA, US; 2University of Washington, Seattle, WA, US OBJECTIVE: To address willingness to pay (WTP) of patients with painful-irritable bladder syndrome (PIBS) using a regression model. METHODS: A sample of 60 patients with PIBS from Seattle and Philadelphia were interviewed. The survey questionnaire contained a WTP question with 11 bids ranging from $1 to $200. Two versions of the questionnaire were administered, one for “improvement” of the condition and one for “cure.” Respondents recorded whether they would accept or reject each bid at the stated dollar value.

28

Abstracts

Data about patients’ self-perceived severity, symptomatology, and demographic variables such as education and income were collected. RESULTS: Results were analyzed with ordinary least square regression (OLS). The dependent variable is maximum willingness to pay (WTP), which is defined as the largest dollar value of the “yes, definitely” or “yes, probably” response to the bids. Independent variables included income, education, self-perceived severity, symptom bothersomeness, number of medical appointments for their condition in the past year, and number of years having the condition. CONCLUSION: The OLS analysis produced a higher expected WTP for cure than for improvement. The R-square and F statistics for each model showed that the variables in the model strongly explain the variation of WTP at a significant level. The regression results showed that education and number of medical appointments in the past year had a positive impact on willingness to pay for improvement, while education and self-perceived severity demonstrated a positive impact on WTP for cure.

PGIU2

EVALUATING CLINICAL OUTCOMES OF GERD TREATMENT OPTIONS USING DECISION ANALYSIS McGhan WF1, Smith MD2, Crawley JA3 1 University of the Sciences at Philadelphia, Philadelphia, PA, US; 2 Health Decision Strategies, Princeton, NJ, US; 3Astra Pharmaceuticals, Wayne, PA, US Long-term maintenance therapies with proton pump inhibitor (PPI), H2 receptor-antagonist (H2RA), or prokinetic agents are often utilized in patients with symptoms of gastroesophageal reflux disease (GERD) or erosive esophagitis. Maintenance therapy can be short-term symptomatic drug treatment repeated when symptoms reoccur (intermittent) or long-term continuous treatment (continuous). OBJECTIVE: To assess which maintenance treatment results in the most asymptomatic patients, and for continuous treatment, to assess if the dose should be reduced (step-down therapy) or the patient switched to another drug. METHODS: A decision tree analytic model with a 1-year time frame comparing clinical outcomes of maintenance drug treatment options with omeprazole, ranitidine, or cisapride was used. Healing rates used in the model were determined by meta-analysis of randomized clinical studies. The decision tree was constructed based on published treatment algorithms. RESULTS: For continuous treatment: at 4 months, 94% of patients taking omeprazole 20 mg daily were asymptomatic vs 77% for ranitidine 300 mg daily. Stepping down from omeprazole 20 mg to 10 mg resulted in 91% asymptomatic, and switching to ranitidine 300 mg daily resulted in 77% of patients asymptomatic; at 12 months, 99% of patients were asymptomatic for all continuous treatment options. However, of these, 63% had no treat-

ment failures throughout the year for omeprazole 20 mg, versus 51% for patients stepped down to omeprazole 10 mg daily, versus 32% for patients switched to ranitidine 300 mg daily, and versus 38% for patients taking ranitidine 300 mg daily initially. For intermittent treatment: at 12 months, 79%, 76%, and 80% of patients were asymptomatic for omeprazole 20 mg, ranitidine 300 mg, and cisapride 40 mg daily, respectively. CONCLUSIONS: Continuous maintenance treatment options resulted in more patients asymptomatic than intermittent treatment options. Continuous treatment with PPI resulted in more patients asymptomatic with fewer treatment failures throughout the year vs H2RA, stepdown, or switch therapy.

PGIU3

THE EFFECTS OF ULCER-HEALING DRUG COPRESCIPTION ON NSAID-INDUCED GASTROINTESTINAL EVENTS: A RECORD LINKAGE COHORT STUDY Morant SV1,2, Shield MJ2, MacDonald TM 1 Medicines Monitoring Unit (MEMO), University of Dundee, Ninewells Hospital & Medical School, Dundee, Scotland; 2 Searle, High Wycombe, UK In the UK, certain anti-ulcer agents are used for prophylaxis against non-steroidal anti-inflammatory drug (NSAID) toxicity. OBJECTIVE: The objective of this study was to compare the incidence of hospitalizations for gastrointestinal events during exposure to these ulcer-healing drugs coprescribed with NSAIDs. METHODS: This was a population-based, observational, record linkage study that included 427,786 patients registered with general practitioners in Tayside, Scotland between 1989 and 1994. Redeemed prescriptions for NSAIDs and anti-ulcer drugs, and hospital discharge diagnoses of patients to whom they were given, were obtained. RESULTS: The risks of hospitalization for a gastrointestinal event during concurrent exposure to NSAIDs and either misoprostol or proton-pump inhibitors, adjusted for recorded risk factors, were no greater than during exposure to NSAIDs alone in patients with and without a prior history of such events. By contrast, adjusted event rates were twice as high during combined exposure to H2 receptor antagonist and NSAID compared with NSAIDs alone in patients with a prior history (p , 0.001), and 20% higher in patients without (p , 0.001). Risks of hospitalizations for gastroduodenal ulcers, erosions, gastritis, and duodenitis during exposure to misoprostol were 23% less than those during exposure to H2 receptor antagonists (p , 0.001). No such difference was found with proton-pump inhibitors. Despite the apparent greater efficacy of misoprostol and proton-pump inhibitors relative to H2 receptor antagonists in patients with a prior history, their use was less common in these high-risk patients. CONCLUSIONS: The observed preferential use of H2

Abstracts

29

receptor antagonist co-therapy in patients at a high risk of NSAID-associated gastrointestinal complications appears not to be justified. Despite substantial confounding between the use of ulcer-healing drugs and recorded risk factors; in the setting of primary care, misoprostol and, to a lesser extent, proton-pump inhibitors appeared to be effective in reducing hospitalization for gastrointestinal events associated with NSAID use.

PGIU4

CROSS-CULTURAL ADAPTATION OF THE PATIENT ASSESSMENT OF CONSTIPATIONSYMPTOM QUESTIONNAIRE (PAC-SYM) Conway K1, Uzun V1, Frank L2, Marquis P3, Farup C4, Dubois D5 1 MAPI Research Institute, Lyon, France; 2MEDTAP International Inc., Bethesda, MD, US; 3MAPI Values, Lyon, France; 4Janssen Research Foundation, Titusville, NJ, US; 5Janssen Research Foundation, Beerse, Belgium OBJECTIVE: With the increase of QoL assessment in clinical trials, it is necessary to have cross-culturally valid instruments to pool data across countries. METHODS: The initial version of the PAC-SYM was designed in US English to assess frequency and intensity of constipation symptoms and patients’ evaluation of their condition in relation to treatment. Prior to use in an international trial, it had to undergo linguistic validation in 12 languages. This involved the recruitment of a QoL specialist in each country. Two independent forward translations in the target languages were reconciled and back-translated into English. Translations were tested for comprehension in a sample target population, compared and internationally harmonized. RESULTS: The developer clarified concepts underlying each item, but translation problems involved finding conceptual equivalents of typical US expressions and response choices. For example, use of similar terms designating one concept could not systematically be retained in all languages. Expressions such as “loose stools” and “watery stools” had only one equivalent in most languages. In conjunction with psychometric results, this item was deleted. The term “abdomen” presented two translation alternatives in most languages, designating either the top or lower part of this area. To maximize comprehension, the most common term was retained. Finally, it was impossible to produce a direct equivalent of “extremely well” in the Scandinavian languages. The option “very well” was therefore chosen. CONCLUSION: Prior to use in an international trial, rigorous cultural adaptation is essential to producing crossculturally valid language versions of the PAC-SYM. Results from linguistic and psychometric validation facilitated item reduction and establishment of a final questionnaire. Further psychometric testing will be important to ensure similar relationships among scales across countries. This work highlights the value of international input in instrument design.

PGIU5

RATE OF GASTROPROTECTIVE AGENT USE, DIAGNOSTIC TESTS, AND HOSPITALIZATIONS AMONG ELDERLY PATIENTS WHO STARTED ARTHROTEC, A FIXED COMBINATION OF MISOPROSTOL AND DICLOFENAC Rahme E1, Joseph L2, Kong SX3, Watson DJ3, LeLorier J 1 Centre de recherche, Centre hospitalier de l’Université de Montréal, Québec, Canada, 2Department of Epidemiology and Biostatistics, McGill University, Montréal, Québec, Canada; 3 Merck & Co., Inc., Whitehouse Station, NJ, US Non-steroidal anti-inflammatory drugs (NSAIDs) cause both minor and severe gastrointestinal (GI) side effects. Significant amounts of healthcare resources are used to prevent/manage these side effects. Arthrotec is a combination of an NSAID (diclofenac) and a gastroprotective agent (GPA, misoprostol) that is thought to decrease NSAID-induced side effects. OBJECTIVE: To investigate whether there is a reduction in GPA use, diagnostic tests, and hospitalizations attributed to Arthrotec versus other NSAIDs among an elderly population. METHODS: We analyzed a 10% random sample (n 5 73,850 patients) of the Régie de l’assurance-maladie du Québec (RAMQ) database collected between January 1993 and December 1997. Patients were eligible for the study if they were free of NSAIDs, GPAs, and upper gastrointestinal diagnostic tests during the year preceding the index date (the date of first NSAID dispensation during the study period). The patients were followed for 2 years. RESULTS: Of 9,618 patients included in the study, 1,153 were started on Arthrotec and 8,465 on any other NSAID. Comorbidity markers were not significantly different in the two groups. Of the Arthrotec patients 17 (1.5%) were hospitalized for GI-related problems, compared to 144 (1.7%) of the other group; 3 (0.3%) of the hospitalized Arthrotec patients died during the hospitalization compared to 15 (0.2%) of the other group; 292 (25.3%) Arthrotec patients used GPAs compared to 2,117 (25.0%) of the other group; 131 (11.4%) Arthrotec patients had GI diagnostic tests compared to 808 (9.5%) of the other group. None of these outcome differences was statistically significant. The cost of the GI-related events was $172/patient-year in the Arthrotec group compared to $103/patient-year in the other group. CONCLUSIONS: Our data showed no significant difference in GI-related events attributable to Arthrotec compared to other NSAIDs. Moreover, the cost of managing these events tends to be higher in the Arthrotec group.

30

Abstracts INFECTIOUS DISEASE

DIABETES

PD1

PID1

COMPLIANCE WITH AMERICAN DIABETIC ASSOCIATION (ADA) RECOMMENDATIONS IN A TYPE 2 DIABETIC POPULATION IN AN ACADEMIC FAMILY MEDICINE CENTER Taylor AT1, Longe RL1, Wagner PJ2 1 University of Georgia, Athens, GA, US; 2Medical College of Georgia, Augusta, GA, US

COST-EFFECTIVENESS OF LEVOFLOXACIN VERSUS CLARITHROMYCIN IN PATIENTS EVALUABLE FOR CLINICAL EFFICACY IN THE TREATMENT OF ACUTE BACTERIAL SINUSITIS Rance L1, Dornsief B2 1 Janssen-Ortho Inc., Toronto, Canada; 2Advanced Biologics, Lambertville, NJ, US

OBJECTIVES: To characterize the population, measure compliance with ADA recommendations, and categorize medication usage. METHODS: All type 2 diabetics (N 5 116) visiting a family medicine center during January 1997 as recorded in an electronic medical record were included. Recommendations for “continuing care visits” were the key data elements retrospectively reviewed for a 1-year period (January–December 1997). Demographic, health insurance, routine office visit (e.g., weight, blood pressure, eye and foot exams), dietary and exercise planning, disease monitoring (e.g., self blood glucose monitoring, serum glycated hemoglobin, serum cholesterol, urinalysis), and detailed drug usage data were collected. RESULTS: The sample had a mean age of 61.1 years, was 53.4% African American and 44.8% Caucasian, and was 69.0% female. A government-financed health insurance plan insured most patients (71.5%). On the average, a patient had about eight office visits annually. Agreement with ADA recommendations of weight and blood pressure measurements at each visit was observed in 56.0% and 78.4% of the patients respectively. Most patients did not have annual primary care eye (57.8%) or dilated ophthalmic (42.2%) or foot (73.3%) examinations documented. Agreement with laboratory recommendations was as follows: annual total cholesterol: 59.5%; semiannual glycated hemoglobin: 14.7%; and annual urinalysis: 66.4%. Most patients (92) were treated with a single drug; 23 of 116 patients used two drugs. Glyburide and glypizide were prescribed most frequently. Compliance with recommendations increased with age and in patients receiving multiple versus single or no drug therapy. CONCLUSION: Documentation of current practice shows wide variation from ADA recommendations for type 2 diabetics. Interventions to modify documentation are planned.

OBJECTIVE: Resource utilization and cost-effectiveness data were compared in patients taking 500 mg od (AM) levofloxacin (levo) with that of 500 mg bid clarithromycin for up to 14 days. METHODS: Patients were eligible for this multicenter, double-blind study if they had signs and symptoms of acute sinusitis, including X-ray evidence. Clinical response was evaluated 2–5 days after completion of therapy and at 1 month post therapy. Utilization parameters were determined from the cost of drug, cost of the physician’s office or emergency room visit, and costs of concomitant medications. The total cost of resources and average cost per patient was calculated. The cost-effectiveness ratio was determined by taking the average cost of product divided by the proportion of clinical success as assessed by investigator. The incremental cost-effectiveness ratio is the mean cost per levo patient minus the mean cost per clarithromycin patient, divided by the proportion of clinical success of levo minus the proportion of clinical success of clarithromycin. Cost-minimization was calculated from the mean cost per levo patient minus the mean cost per clarithromycin patient. RESULTS: 191 patients had resource utilization data, 98 in the levo group and 93 in the clarithromycin group. The total cost of resources was $11,566.24 for levo and $12,635.14 for clarithromycin. Average cost per patient was $119.24 for levo and $135.86 for clarithromycin. The cost-effectiveness ratio for levo was 1.3, and for clarithromycin 1.5. Because clinical success rates were similar a cost minimization analysis was performed. The result was 16.6. The incremental cost-effectiveness ratio was 250.5. CONCLUSION: In patients with similar clinical success rates, this study shows that levo 500 mg od is more costeffective than 500 mg bid clarithromycin.

PID2

COST-MINIMIZATION ANALYSIS OF CEFAMANDOLE AND TOBRAMYCIN VERSUS STANDARD THERAPY IN THE TREATMENT OF PATIENTS WITH LOCAL APPENDICULAR PERITONITIS Rozenson O1, Bazganov N2, Belkov A1, Bikov A3, Stratchounski L1 1 Smolensk State Medical Academy, Smolensk, Russia; 2 Yaroslavlskya State Medical Academy, Yaroslavl, Russia; 3 Volgogradskaya State Medical Academy, Volgograd, Russia

Abstracts Cost $ (1997 year) Level I Level II Level III

31 CT

ST

P value

103.0 6 41.0 275.3 6 118.1 172.2 6 96.9

14.9 6 13.2 189.9 6 99.1 175.5 6 99.6

0.001 0.001 0.904

OBJECTIVES: Pharmacoeconomic assessment of cefamandole/tobramycin (CT) versus standard therapy (ST) in the treatment of local appendicular peritonitis (LAP). Standard therapy was antibacterial drugs routinely used for treatment of LAP in centers of study. It was allowed to add metronidazole to both regimes. METHODS: Clinical outcomes and health resource utilization data were collected from 92 patients enrolled in an open, comparative, randomized, prospective, multicenter trial comparing the efficacy and safety of CT versus ST in the treatment of LAP. A cost-minimization analysis was performed from the institutional perspective. Costs of study antibiotics, treatment of failures and adverse events (AE), and hospitalization were calculated. RESULTS: The most common regimes in ST were ampicillin (27%), ampicillin 1 oxacillin (11%), penicillin 1 gentamicin (9%), penicillin 1 kanamycin (9%). Metronidazole was used in five patients in CT and in four patients in ST. Cure rates were 42 of 44 (95.5%) for CT and 41 of 45 (91.1%) for ST (p 5 1.000). Adverse events (AE) were 12 of 47 (25.5%) for CT and 10 of 45 (22.2%) for ST (p 5 0.547). Three levels of cost were analyzed: level I 5 study drug acquisition cost; level II 5 level I 1 preparation/administration 1 treatment of AE/ failures 1 hospital bed costs; level III 5 level II 2 level I. CONCLUSIONS: ST was cost-effective compared to CT in the treatment of LAP. The pharmacoeconomic advantage of ST was due to the significantly lower acquisition cost of ST antibiotics.

MENTAL HEALTH AND NEUROLOGIC DISORDERS

nothiazines, butrophenones, and atypical antipsychotics); benzodiazepines and sedatives. The annual direct costs of the medications were ascertained from the recent monthly index for medical specialties (MIMS), from standard chemists, and from pharmaceutical firms. The statistical mode or most frequent cost was considered for further computations. RESULTS: Regarding antidepressants, tricyclics were noted to have the following annual costs (1 US$ 5 Rs. 43): Imipramine Rs. 1130 (US $26.3), amitriptyline Rs. 1460 ($34), dothiepin Rs. 2410 ($56.1), trimipramine Rs. 2430 ($56.5) and clomipramine Rs. 6130 ($142.6). Fluoxetine has an annual cost of Rs. 840 ($19.5), mianserin Rs. 2010 ($46.7), amoxapine Rs. 3835 ($89.1), amineptine Rs. 10,700 ($248.8), and the most expensive, tianeptine, Rs. 21,000 ($488). The recently introduced antidepressant sertraline has an annual cost of Rs. 2740 ($63.7). Prophylaxis for affective disorders with lithium therapy would have an annual cost of Rs. 1200 ($27.9). Regarding antipsychotics, the annual costs of oral medications of chlorpromazine and haloperidol were the same, Rs. 2000 ($46.5), long-acting oral penfluridol Rs. 500 ($11.6), and depot long-acting injectable fluphenazine decanoate, haloperidol decanoate, and flupenthixol Rs. 750 ($17.4), Rs. 1500 ($34.9), and Rs. 2100 ($48.8), respectively. The newer atypical antipsychotics costs were: resperidone, Rs. 4500 ($104.7) and clozapine Rs. 5500 ($127.9). Benzodiazepines have an annual cost of $7.5 for diazepam, $3.5 for lorazepam and chlordiazepoxide, $2.8 for oxazepam, and $17.6 for alprazolam. Sleep medications nitrazepam, flurazepam and zopiclone have an annual cost of $3.3, $8, and $18.8 respectively. CONCLUSION: The annual costs are amazingly less than those in Western nations, but nevertheless produce a significant burden on Indian families, and are a common reason for poor compliance. Another trend noted above is that newer antidepressants such as amineptine and tianeptine, antipsychotics such as resperidone and clozapine, and hypnotics such as zopiclone are much more expensive than the older medications. Thus, the annual costs of medications are only likely to increase with time, enhancing the financial burden of treating these psychiatric disorders.

PMH1

ANNUAL DIRECT COSTS OF PSYCHOTROPIC MEDICATIONS FOR THE TREATMENT OF PSYCHIATRIC DISORDERS IN INDIA Chaturvedi SK National Institute of Mental Health & Neurosciences, Bangalore, India OBJECTIVES: The present study ascertains the annual direct costs of common psychotropic medications used in the management of common psychiatric disorders in a country with limited financial resources. METHODS: The direct costs of the psychopharmacological agents evaluated were antidepressants (tricyclic, SSRI, and newer antidepressants available); antipsychotics (phe-

PMH2

CLINICAL AND ECONOMIC CHARACTERIZATION OF FAMILY MEDICINE PATIENTS RECEIVING ANTIDEPRESSANT DRUGS Taylor AT1, Wagner PJ2, Ahmed A1 1 University of Georgia College of Pharmacy, Athens, GA, US; 2 Medical College of Georgia School of Medicine, Augusta, GA, US Many antidepressant drugs are prescribed by generalist physicians. However, few details are available concerning the circumstances in which these drugs are utilized in primary care settings.

32

Abstracts

OBJECTIVE: The objective of this study was to document the clinical and economic characteristics of ambulatory family medicine patients receiving antidepressant medications. METHODS: The medication profiles of the electronic medical records (EMR) of all Family Medicine patients at the Medical College of Georgia were evaluated. Only patients prescribed an antidepressant drug within the previous 6 months were reviewed. Patient records were reviewed in the order in which they were entered into the EMR system. RESULTS: The electronic records of 116 patients were reviewed by the same physician researcher. Demographic and clinical data for 89 (76.7%) female and 27 (23.3%) male patients were evaluated. The mean age for these patients was 46.5 years. Sixty-two (53.4%) patients were white, 38 (32.8%) were black. Although usually considered risk factors for the development of depression, marital status and the use of alcohol were not noted in 47 (40.5%) and 62 (53.4%) patients, respectively. The most common comorbid conditions were hypertension (47.4%), gastroesophageal reflux disease (GERD) (20.7%), arthritis (20.7%), anxiety (18.1%), diabetes mellitus (17.2%), and pain syndromes (45.7%). Thirty-one (26.7%) patients, received amitriptyline. Twenty-two (71%) received less than 50 mg per day. Fourteen (12%) patients received two antidepressants, most often an SSRI in the morning with a sedating agent at bedtime. Of the 26 patients with available financial data, 25 patients paid a co-payment of $5.00 or less. CONCLUSIONS: Antidepressant drugs are frequently prescribed in primary care patients with multiple comorbid conditions (mean 3.9).

PMH3

COST-EFFECTIVENESS OF FLUVOXAMINE IN THE TREATMENT OF RECURRENT DEPRESSION IN FRANCE Nuijten MJC1, Piercy J2 1 The Lewin Group, Hoofddorp, The Netherlands; 2Solvay Pharmaceuticals, Hannover, Germany OBJECTIVE: The objective of this study was to examine the cost-effectiveness of fluvoxamine compared to treatment with tricyclic antidepressants (TCAs) in patients who suffer depressive episodes. METHODS: The setting for this study was France. A Markov process model was constructed to model the effectiveness, as measured by time without depression, and the costs of both treatments. The model examined a period of 18 months in order to capture the influence of both relapses and recurrences on the outcomes. Data for the construction of the model came from the published literature, an expert panel, and a large multicenter randomized clinical trial. Costs were obtained from published sources. RESULTS: The results of the baseline analysis showed that the use of fluvoxamine in the maintenance treatment

(recurrence prevention) of depressive disorders was less costly than TCAs with total costs (direct and indirect costs) of 40,232.40 FF versus 52,257.53 FF, respectively. In addition, due to the prevention of relapse and recurrence, effectiveness fluvoxamine was favored, as it was associated with a longer period of time without depression when compared to therapy with TCAs: 79% of the study period compared to 71% for TCAs. Sensitivity analyses confirmed the robustness of these findings. CONCLUSION: In conclusion, based on the assumptions used in the model, the use of fluovoxamine as maintenance therapy is clinically and economically justified in patients with depressive disorders.

PMH4

THE BUDGET IMPACT PHARMACOTHERAPY SELECTION FOR MAJOR DEPRESSIVE DISORDER: A MULTINATIONAL STUDY Casciano JP1, Einarson TR3, Arikian SA1,2, Doyle JJ1,2 1 The Analytica Group, Ltd., New York, NY, US; 2Columbia School of Public Health, New York, NY, US; 3Faculty of Pharmacy, University of Toronto, Ontario, Canada OBJECTIVE: We conducted a multinational pharmacoeconomic evaluation of treatment for major depressive disorder (MDD) to assess the potential budget impact of changes in antidepressant prescribing patterns in Canada, Germany, Italy, the Netherlands, Poland, Spain, Sweden, Switzerland, the United Kingdom, the United States, and Venezuela. METHODS: The analysis compared the serotonin-norepinephrine reuptake-inhibitor (SNRI), extended-release venlafaxine, to selective serotonin reuptake-inhibitors (SSRIs) and tricyclic antidepressants (TCAs). A metaanalysis was performed to determine the clinical rates of success, defined as a 50% reduction in depression scores on the Hamilton Depression Scale (HAM-D) or the Montgomery-Asberg Depression Rating Scale (MADRS). Treatment regimen costs were determined from standard lists, fee schedules, and communication with local health economists in each country. We included data from 48 studies on 4,690 patients in the meta-analysis. The metaanalytic rates were applied to decision analytic models to calculate expected cost and expected outcomes for each comparator. Cost-effectiveness was determined using expected values for both a successful outcome (as defined by the meta-analysis), and a composite measure of outcome termed “Symptom-Free Days.” A policy analysis was conducted to examine the budgetary impact on the health system, in each country, of increasing the utilization of the most effective comparator. RESULTS: Treatment of MDD with extended-release venlafaxine yielded the highest overall efficacy rates for outpatients (73.7%) and inpatients (62.3%). Extendedrelease venlafaxine had the highest failure rate due to lack of efficacy (5.3%) and adverse drug reactions (11.3%). Initiating treatment of MDD with extended-release venlafaxine yielded the lowest expected cost for outpatients

Abstracts

33

in 9 of the 11 countries studied, and for inpatients in 10 of the 11 countries studied. The weighted average expected cost per patient for MDD treatment varied from US $642 (Poland) to US $5,825 (US). The estimated budgetary impact to the primary payer for each 1% of utilization of extended-release venlafaxine ranged from US $61,640 (Switzerland) to US $1,565,900 (Germany) for each 1 million population. CONCLUSIONS: Policy-makers should implement measures to increase the utilization of cost-effective antidepressant treatment such as extended-release venlafaxine.

PMH5

COST OF ALZHEIMER’S DISEASE CARE IN THE UNITED KINGDOM Thwaites RMA1, Yeardley HL2 1 Glaxo Wellcome Research & Development, Greenford, UK; 2 Quintiles, Neu-Isenburg, Germany OBJECTIVE: To assess the relationship between illness severity and costs of care in noninstitutionalized patients with Alzheimer’s Disease in the United Kingdom. METHODS: In this cross-sectional, multicenter analysis, Alzheimer’s Disease patients cared for in the community (128), their caregivers (128), and 56 matched controls were recruited chronologically and interviewed once to establish resource use over the previous 3 months. Patients were stratified into three severity groups according to their cognitive function as assessed by their Mini Mental State Exam score. Costs were calculated from the perspective of society as a whole. RESULTS: Over the 3-month period, total mean cost per control subject (£387) was minor compared with the mean cost incurred by patients with mild (£6,616), moderate (£10,250), and severe (£13,593) Alzheimer’s Disease. Indirect cost, mainly time spent by caregivers, was the main cost component in all groups (68.6%), followed by direct medical costs (24.7%). CONCLUSION: The cost of care for an Alzheimer’s Disease patient is substantial and is directly related to the severity of the patient’s illness.

PMH6

DEVELOPMENT OF A NEW DISEASE-SPECIFIC QUALITY OF LIFE SCALE FOR USE IN SCHIZOPHRENIA Wild D1, Jenkinson C1, Fitzpatrick R1, Wilkinson G2, Hesdon B2, Farina C3 1 Oxford Outcomes, Institute of Health Sciences, Oxford, UK; 2 Department of Psychiatry, University of Liverpool, Liverpool, UK; 3Janssen Cilag Ltd., Saunderton, High Wycombe, UK OBJECTIVE: To develop a health-related quality of life questionnaire to measure the effects of treatment on the quality of life of people with schizophrenia, the Schizophrenia Quality of Life Scale (SQLS).

METHODS: A standardized state-of-the-art methodology has been used for the development of the SQLS. Items were generated from in-depth interviews with 20 patients and eight health professionals. From a list of 440 statements, a pilot questionnaire containing 80 items was developed by a panel of experts. The pilot questionnaire was tested on 20 patients, which resulted in some wording changes. A sample of 161 patients completed the 80item questionnaire. A factor analysis using Varimax rotation was then conducted. This identified 31 questions tapping three factors: psychological, motivational, and physical symptoms. The psychometric properties of the SQLS were then assessed using the following methodology. Sixty-five patients were given the SQLS to complete, along with the following measures: the SF-36, the Hospital Anxiety and Depression Scale (HADS), and the General Health Questionnaire (GHQ-12). A clinical assessment using the Clinical Global Impressions and the Health of the Nation Outcome Scales (HONOS) was made by a psychiatrist. A subset of the patients was asked to complete the SQLS for a second time to assess the reproducibility of the measure. RESULTS: The validity and reliability of the scale are good. CONCLUSIONS: The SQLS is now available for use, although further studies of its psychometric properties are ongoing.

PMH7

CROSS-CULTURAL ADAPTATION OF THE INTERNATIONAL INDEX OF ERECTILE FUNCTION (IIEF) IN 31 COUNTRIES Conway K1, Mishra A2, Giroudet C1, The Global Translation Group (31 QoL Experts) 1 MAPI Research Institute, Lyon, France; 2Pfizer, Groton, CT, US With the increase of QoL assessment in clinical trials, it is necessary to have cross-culturally valid instruments to make comparisons of health status outcomes and pool data across countries. The IIEF is a 15-item instrument designed in English to assess erectile dysfunction. Prior to use in an international trial, it underwent cross-cultural adaptation in 31 languages. This involved the recruitment of a QoL specialist in each target country. Two independent forward translations were produced by native speakers, bilingual in English. These were reconciled and backtranslated into English. The translations were tested for comprehension in a sample target population, compared and internationally harmonized. The developer clarified concepts underlying each item. Translation was difficult, especially when selecting equivalent response categories and appropriate equivalents of common English structures and expressions. The difficulty of the translation and its acceptability depended on the comparability of the original and the target language and culture. Because English is flexible, it was possible to accept “Most of the times (much more than half of the time)” as an answer to the question: “How often have you felt sexual desire in

34

Abstracts

the last 4 weeks?” In many languages a literal translation would have been unclear and so the question was replaced by “how much of the time. . . .” Patient testing showed different reactions to the instrument across countries and made it clear that terms such as “intercourse,” “erection,” or “ejaculation” were not always understood and required explanation. Prior to use in an international trial, rigorous cultural adaptation was essential to produce cross-culturally valid language versions. Psychometric testing will be important to ensure similar relationships among scales across countries. The comparison of answer patterns across countries will clarify whether people with similar health give equivalent answers.

PMH8

EFFECT OF RISPERIDONE ON THE COST OF NURSING HOME CARE FOR RESIDENTS WITH BEHAVIORAL AND PSYCHOLOGICAL SIGNS AND SYMPTOMS Caro JJ1, O’Brien JA1, Shomphe LS1, Torfs K2 1 Caro Research, Concord, MA, US; 2Janssen Research Foundation, Beerse, Belgium OBJECTIVE: To evaluate the economic effect of using risperidone in nursing home residents with documented behavioral and psychological signs and symptoms of dementia (BPSSD). METHODS: Based on 1995 data from a large nursing home database, the average annual incremental cost of care due to BPSSD was calculated for three diagnosis groups (All Diagnoses, All Dementia, and Non-Dementia). These costs, in 1997 US dollars, included increased reimbursement resulting from transition to a higher level of care and more frequent physician visits due to BPSSD. These were compared with the cost of 1 mg per day risperidone therapy. RESULTS: BPSSD was documented in 13.5% of all nursing home residents—in 20% of those with dementia and in 11% in those without dementia. The annual cost of risperidone was estimated at $101, $148, and $85, respectively, per patient in each group. The BD-related costs considering all patients, regardless of behavioral status, were $245, $383, and $195, respectively, per resident in each group. Risperidone was estimated to break even if 42% of BPSSD-related transitions to a higher level of care were prevented among all patients, 39% if only demented patients are considered, and 44% for those without dementia. With higher effectiveness, treatment with risperidone would result in cost-savings. CONCLUSIONS: The use of risperidone in institutionalized patients with BPSSD should provide not only substantial relief of what can be a difficult-to-manage situation but also a favorable economic effect.

PN1

A UK PRIMARY CARE DATABASE (UKPCD)-UK MEDIPLUS AS A READILY AVAILABLE RESEARCH TOOL FOR DETERMINING THE INCIDENCE AND PREVALENCE OF MULTIPLE SCLEROSIS, ITS TREATMENT, PRESCRIPTION COSTS, AND REFERRALS IN UK PRACTICE Grigas E, Carmichael R, Harrison P, McDermott J Disease Treatment and Outcomes Group, IMS Health, Middlesex, UK OBJECTIVES: The annual prevalence and annual incidence of the diagnosis of multiple sclerosis (MS) in the UK were determined, together with the therapies employed, the associated costs, and the referrals to secondary care. METHODS: This study used UKPCD, which contains longitudinal records of nearly 2 million patients. Data are captured daily from general practitioners who use the AAH Meditel System 5 software as part of their standard record keeping. Patients were selected according to the primary care coding system, “Read,” (entered as a problem or problem-linked note code), from June 1997 to June 1998. Therapies, prescription costs, and referrals were extracted from these patients’ records. RESULTS: The annual prevalence of MS was 62/100,000 (95% CI 58–66); the annual incidence was 7/100,000 (95% CI 6–8). A comparison of all therapies used to manage MS patients showed apparent differences between prevalent and incident patients, except in antidepressants (36% both groups). Non-narcotic analgesics were prescribed for 45% of prevalent patients and 32% of incident patients. When therapy for prevalent patients was linked to the Read code for MS, 33% of patients were prescribed central muscle relaxants and 24% antidepressants. The cost to the community for all prescribed products in the study year was £159,508. The majority of note referrals for both prevalent and incident patients was for neurologist care (20% and 40%) and physiotherapy (12% and 14%), respectively. CONCLUSION: This study is an example of the use of an automated, observational, primary care database such as MediPlus in pharmacoeconomic and outcomes research.

PN2

RESULTS OF DISEASE MANAGEMENT IN HEADACHE TREATMENT Behrens M2, Göbel H1 1 Neurological Clinic, University of Kiel and Neurological Pain Clinic, Kiel, Germany; 2Glaxo Wellcome GmbH & Co., Germany OBJECTIVE: Patients report in general to be dissatisfied with headache treatment. A disease management program was developed to improve the quality of headache treatment of a general sick fund–insured population. METHOD: Employees at a ship-building site (n 5 1996) and their families were asked to fill in questionnaires if they suffered from headache: the Short-Form 36, the Kiel

Abstracts

35

Headache Questionnaire, and some additional items to document medication and satisfaction with general therapy. General information about management and prevention of headache was offered to the 341 respondents suffering from headache, as well as to physicians and pharmacists. In a second survey, questionnaires were filled out by the participants to record the impact of the program. RESULTS: Results from the first survey were reported by Göbel et al. (Cephalalgia 17[1997]325). Of the 183 traceable initial respondents, 68.9% took part in the final survey: 29.1% of the participants reported an improvement in health status; 43.1% stated fewer headaches; 38.9% reported shorter headache attacks; 22.6% had better knowledge to identify trigger factors; 21.0% reduced their consumption of medication; and 41.9% learned an active relaxation technique. Together with other extensive healthcare measures initiated through the company, there was a reduction of lost labor time due to illness. CONCLUSION: It is possible to create a network between employees, physicians, pharmacists, the sick fund, and the pharmaceutical industry to successfully address the question of headache management within a work-site environment. Unmet medical needs can be identified, and awareness for intervention programs can be raised through studies using general headache questionnaires. The information of patients as well as physicians and pharmacists reduced the burden of headache and improved the outcomes of headache treatment. More collaborative efforts to improve the unsatisfactory situation of headache patients and migraine patients are needed.

PN3

A FOUNDATION STUDY FOR COSTEFFECTIVENESS ANALYSIS: THE DEFINITION OF ECONOMIC PARAMETERS AND CLINICAL ENDPOINTS IN PARKINSON’S DISEASE Saal G1, Adesina K1, Shuster S2 1 MAPI Values, Cheshire, UK; 2Medtronic Neurological, Minneapolis, MN, US The identification of cost components deriving from direct and indirect consequences of disease is a fundamental step in conducting cost-effectiveness (CE) analysis. In time-constrained efforts to collect CE data for inclusion in a product’s commercial development dossier, little attention is likely to be paid to fundamental research. Coupled with other early research components, such as expert and/or patient interviews, a prior comprehensive review of the literature can ensure a strong foundation on which robust economic arguments could subsequently be formulated. OBJECTIVE: The objective of this research was to explore and define 1) cost drivers and 2) clinical endpoints representing global treatment effectiveness in Parkinson’s disease (PD). METHODS: A review of the published literature in PD was conducted, surveying elements including:

• Contemporary treatment options and costs; • Clinical markers measuring progression; • Indirect morbidity and related costs. The geographic foci of the research were Germany and the UK. RESULTS: A global indicator of symptoms and complications of PD treatment is the Unified Parkinson’s Disease Rating Scale (UPDRS); significant events representing complications and co-morbidity include prevalence and severity of motor complications—dyskinesias, depression, dementia, and postural imbalance—leading to falls and other accidents. Hospitalization accounts for 50% to 90% of costs of PD care in Germany and the UK, followed by medication costs. The main determinants of hospitalization rates are disease stage, akinetic type of disease and presence of motor fluctuations in spite of, as well as in relation to, therapy. CONCLUSIONS: In collecting economic data for PD treatment, the definitive demonstration of CE must include improved long-term efficacy, safety, and reduced rates of direct and indirect resource utilization—in particular, hospitalization.

PN4

QUALITY OF LIFE AND FUNCTIONAL STATUS MEASURES IN ADVANCED PARKINSON’S PATIENTS TREATED WITH PRAMIPEXOLE, BROMOCRIPTINE, OR PLACEBO Greer M1, Guttman M2, Meier D3, Leonard J3, Koester J3 1 Boehringer Ingelheim Pharmaceuticals, Inc., Ridgefield, CT, US; 2 Boehringer Ingelheim Pharmaceuticals, Inc., Markham, Ontario, Canada; 3Boehringer Ingelheim Pharmaceuticals, Inc., Cologne, Germany OBJECTIVE: Patient outcome measures of health utility and functional status were used to assess sensitivity to drug treatment and its relationship to PD severity. BACKGROUND: A multicenter trial in seven countries compared the safety, tolerance, and efficacy of pramipexole (to 4.5 mg) and bromocriptine (to 30 mg) in advanced PD. Patients (N 5 247, 80 pramipexole, 84 bromocriptine, 83 placebo) were male and female, in- and outpatients, 30 years and older, with advanced idiopathic PD (stages II-IV Hoehn and Yahr [H&Y]), with motor fluctuations characterized as end-of-dose phenomena. Each received individually optimized dosages of levodopa. (Guttman et al., “Double-blind comparison of pramipexole and bromocriptine treatment with placebo in advanced Parkinson’s disease,” [Neurology 49:Oct 1997, 1060–1065]). METHODS: Patients were treated for up to 9 months, 11 days: 1) ascending dose up to 3 months, 2) maintenance dose for approximately 6 months, and 3) decreasing dose up to 11 days. Primary efficacy endpoints were: UPDRS II—Activities of Daily Living (ADLs), averaged over “on” and “off” periods, and UPDRS III—Motor Examination during “on” periods. Secondary endpoints included the EuroQoL visual analog scale (VAS) and Functional Status

36

Abstracts

Questionnaire (FSQ). Analyses were based on differences between baseline and final measures. RESULTS: Significant differences (p , 0.05) in FSQ subscores were found in Basic ADLs, Intermediate ADLs, and Mental Health, for both pramipexole and bromocriptine compared with placebo. Pramipexole patients scored significantly higher than placebo patients, while bromocriptine patients did not, on FSQ Intermediate ADLs, Usual Activity Days Reduced, and EuroQol VAS. FSQ and EuroQoL scores declined with increasing PD severity, as assessed by H&Y on FSQ domains—Basic ADLs, Intermediate ADLs, Social Activities, Quality of Interaction—and on FSQ single items—Days Cut Down on Usual Activities Due to Illness/Injury and Satisfaction with Sexual Relationship, as well as the EuroQoL VAS. CONCLUSIONS: The EuroQoL VAS and some domains/items of the FSQ were sensitive to drug treatment. An inverse relationship was revealed between EuroQol VAS scores and some FSQ scores and PD severity.

PN5

CROSS-CULTURAL ADAPTATION OF THE FUNCTIONAL ASSESSMENT OF MULTIPLE SCLEROSIS (FAMS) QUESTIONNAIRE IN NINE COUNTRIES Conway K1, Ojanen M1, Cella D2, Lent L2, Eremenco S2, Sagnier P3 1 MAPI Research Institute, Lyon, France; 2Center on Outcomes, Research and Education, Evanston, IL, US; 3Bayer AG, Pharma Europe, Leverkusen, Germany OBJECTIVE: With the increase of QoL assessment in clinical trials, it is necessary to have cross-culturally valid instruments to pool data across countries. The FAMS, a 59-item instrument designed in US English, assesses six domains of multiple sclerosis. METHODS: Prior to use in an international trial, the FAMS needed to undergo cross-cultural adaptation in nine languages. This involved the recruitment of a QoL specialist in each target country. Two independent forward translations were produced by native speakers, reconciled and then back-translated into English. The translations were tested for comprehension in a sample target population, compared and internationally harmonized. The developer clarified concepts underlying each item. Translation problems involved finding conceptual equivalents of typical US expressions and the response categories. For example, a literal translation of the term “social activity,” referring to humanitarian work in certain languages, required a descriptive equivalent. Translating “getting around in public places” necessitated lengthy descriptions in certain languages to retain the original concept. The neutral term “condition” could not be retained in certain translations, where a reference to “health state” or even “illness” was the best compromise. Patient testing showed that a response option “not applicable” was lacking and that answers should refer more clearly to the introductory sentence. To harmonize with the vali-

dated original, however, no changes were made in the response categories. CONCLUSIONS: Before use in an international trial, rigorous cultural adaptation was essential in producing crossculturally valid language versions. The integration of changes to the responses in an updated form of the instrument might require further validation. Further responsiveness data will be generated by the clinical trial. Ideally, initial instrument design should include international input.

PN6

IMPACT OF A COST-SHARING DRUG INSURANCE PLAN ON DRUG UTILIZATION AMONG INDIVIDUALS RECEIVING SOCIAL ASSISTANCE LeLorier J, Blais L, Couture J, Rahme E Centre de recherche, Centre hospitalier de l’Université de Montréal, Campus Hôtel-Dieu, Québec, Canada OBJECTIVE: In August 1996, the Régie de l’assurancemaladie du Québec (RAMQ) introduced a new drug insurance plan for all residents of the Canadian province of Québec. Before this insurance plan, individuals receiving social assistance had all their medications paid for by the RAMQ. Now, they have to pay up to $50 per quarter for their medications in terms of deductible and co-insurance. Concerned that this financial constraint could have an impact on the use of medications, we investigated whether the consumption of three classes of medications (neuroleptics, anticonvulsants, and inhaled corticosteroids) had been modified after the new drug plan was introduced. METHODS: Using time series analysis and the computerized databases of the RAMQ, we analyzed the variation in the number of prescriptions dispensed per month for each drug class from August 1992 to July 1997. RESULTS: Users of inhaled corticosteroids were the most affected, with a statistically significant decrease of 28% in the number of dispensed prescriptions for the 11 months following the inception of the drug plan (August 1996 to June 1997) as compared with the same period 1 year before. For the same period, we expected an increase of 9% according to the trend seen from August 1992 to July 1996. For users of neuroleptics and anti-convulsants we found a 1% decrease for each, expecting increases of 7% and 8%, respectively. CONCLUSIONS: Following the introduction of the Québec drug plan, individuals receiving social assistance significantly decreased their consumption of inhaled corticosteroids, medications that are considered to be effective and essential.

Abstracts

37 PN7

COST-EFFECTIVENESS OF PRAMIPEXOLE IN PARKINSON’S DISEASE IN THREE COUNTRIES Hoerger T1, Bala M2, Greer M3, Rowland C4 1 Research Triangle Institute, Research Triangle Park, NC, US; 2 Centocor, Inc., Malvern, PA, US; 3Boehringer Ingelheim Pharmaceuticals, Inc., Ridgefield, CT, US; 4Pharmacia & Upjohn, Kalamazoo, MI, US OBJECTIVES: A pharmacoeconomic model is used to estimate the costs and cost-effectiveness of treating Parkinson’s disease patients with pramipexole in Canada, Germany, and Sweden. METHODS: We developed a cost-effectiveness model that linked Unified Parkinson’s Disease Rating Scale (UPDRS) Part II (ADL) and III (Motor) scores to disease progression, costs, and quality-adjusted life years (QALYs). Data were obtained from clinical trials, literature review, and a survey of 193 patients’ resource use and health utility. Country-specific data on treatment patterns were collected using panels of physician experts in each country. Cost and QALY estimates from the model were used to estimate the incremental cost-effectiveness of pramipexole relative to baseline treatment. Separate analyses were performed for early and advanced Parkinson’s patients. Extensive sensitivity analyses were performed. RESULTS: In Canada and Germany, treatment with pramipexole has higher costs but is more effective than baseline treatment. In Sweden, treatment with pramipexole has lower costs and is more effective than baseline treatment. The incremental cost-effectiveness ratios for early patients in Canada, Germany, and Sweden are $CDN 28,097, DM 26,780, and -SEK 41,161, respectively (in 1997 currency). For advanced patients, the corresponding ratios are $CDN 25,090, DM 49,168, and -SEK 27,386. These ratios are lower than the cost-effectiveness ratios of many widely used medical treatments. CONCLUSIONS: Subject to the inherent limitations of modeling chronic disease progression and subsequent health utility and costs, the results indicate that pramipexole is a cost-effective treatment for early and advanced Parkinson’s disease patients in Canada, Germany, and Sweden.

PULMONARY AND ARTHRITIC DISORDERS

PP1

CHARACTERISTICS OF CHRONIC AIRWAY DISEASE PATIENTS STARTED ON NEBULIZED IPRATROPIUM SOLUTION LeLorier J1, Blais L1, Castilloux A-M1, Renzi P1, Miller B2 1 Centre de recherche, Centre hospitalier de l’Université de Montréal, Campus Hôtel-Dieu, Canada; 2Boehringer Ingelheim Ltd., Burlington, Ontario, Canada BACKGROUND: There has been an important increase in the use of nebulized ipratropium solution (NIS) in the

ambulatory treatment of elderly patients with chronic airway disease. Since these are relatively expensive medications, it is important to define the characteristics and the natural history of the patients who are started on this therapy and compare them with those of the patients started on ipratropium metered dose inhalers (IMDI) or oral theophylline (OT). METHODS: The Quebec health insurance program (RAMQ) database was used for the study. The index date was the first dispensation of NIS (n 5 761), IMDI (n 5 3870), or OT (n 5 10355). The consumption of drugs and medical services was measured during the 12 months before the index date. RESULTS: At the time therapy was initiated, patients started on NIS had had more hospitalization days for chronic airway disease (NIS: 2.97, IMDI: 0.92, OT: 0.12). They were also much heavier users of beta-agonists (NIS: 13.53, IMDI: 4.9, OT: 0.11) and inhaled corticosteroids (NIS: 4.63, IMDI: 2.1, OT: 0.37). The fact that patients started on NIS had a more advanced form of chronic airway disease was confirmed by their much higher mortality rates (percent and 95% CI) during the 12 months following their index date (NIS: 17.2 [14.3– 20.1], IMDI: 8.3 [7.5–9.1], OT: 4.9 [4.6–5.3]). CONCLUSION: Patients are started on NIS when they have reached a very advanced form of their disease, as shown by their high mortality rates and extremely heavy use of beta-agonists.

PP2

THE EFFECT OF AGE ON COST-EFFECTIVENESS IN THE TREATMENT OF MODERATE ASTHMA Kielhorn A, Behrens M Glaxo Wellcome GmbH & Co., Hamburg, Germany OBJECTIVE: The aim of this work is to identify the effect of age on cost-effectiveness in the treatment of moderate asthma patients with inhaled corticosteroids. METHODS: The cost-effectiveness analysis was conducted as a retrospective analysis of clinical trial data. The data of one open and one double-blind multicenter, parallel group study comparing the efficacy of Fluticasone and Flunisolide were pooled. The pooled intent-totreat (ITT) group consisted of 308 patients. Because of the retrospective nature of the study, only the costs of treatment medication and of concomitant medication were included. No hospital admissions occurred. The adverse event costs were identified as negligible and therefore omitted. Because physicians’ visits were protocol driven, they were also excluded. Effectiveness (successfully treated patients) was defined by an increase in peak flow of 5% or more. Total treatment costs for each age group were calculated and divided over the number of patients fulfilling the effectiveness criterion. RESULTS: Total average daily treatment costs per patient were DM 4.58 (sd 5 DM 1.61). The average costeffectiveness ratio per successfully treated patient was DM 7.84 (sd 5 DM 1.72). The cost-effectiveness ratios

38

Abstracts

were positively correlated to age (p , 0.05). This effect was caused by slightly rising total treatment costs over age and a declining number of patients fulfilling the effectiveness criterion. CONCLUSION: Cost-effectiveness ratios are positively correlated to age in the treatment of moderate asthmatics.

PP3

INCREMENTAL COST OF DISEASE IN NOSOCOMIAL PNEUMONIA FROM A HOSPITAL’S PERSPECTIVE: A FEASIBILITY STUDY Dietrich ES1, Hug S1, Mast O2, Schulgen G3, Daschner F1 1 Institute of Environmental Medicine and Hospital Epidemiology, University Hospital, Freiburg, Germany; 2Health Economics and Outcomes Research, Bayer Vital, Leverkusen, Germany; 3Institute of Biometry and Medical Informatics, University Hospital, Freiburg, Germany OBJECTIVE: To investigate the feasibility of different methodologies for a prospective matched-pair design study on the incremental daily cost of disease for patients with nosocomial pneumonia (NP) from a hospital’s perspective. METHOD: Adult patients newly admitted to three ICUs with ,3 days of hospitalization and without preexisting pneumonia were followed up daily until discharge (maximum 30 days). Any new onset of NP was registered. Gender, age, diagnosis at admission, underlying disease, surgery, ward, further diagnosis, Apache II-Score, and KropecPneumonia-Risc-Score were preselected as matching criteria for controls. Consumed resources were documented daily. Costs were calculated by two methods: a combined bottom-up approach (resource consumption of drugs, blood products, diagnostics, other medical materials) and top-down approach (proportionate cost based on accounting figures by ward for all other costs). Alternatively, all costs were calculated top-down based on administrative data. RESULTS: In 25 days 112 patients were registered; 50 met all study criteria, of whom 3 developed nosocomial pneumonia. Matching was possible only on gender, age, reason for admission (eight categories), ward, hospital days until onset, immunosuppressive and respiratory status. The average incremental daily cost for patients with NP was DM 440 6 140%. Both costing methods led to similar results (2% difference). The first method required significantly more time for data collection. The sample size for a study, given that the 95% confidence range of cost equals the mean of the incremental cost (33%), is 40 patients with NP, plus an adequate number of controls. CONCLUSION: The approach involving some bottomup data is not feasible for a full study. The top-down only approach more efficiently leads to quite similar results. For reconfirmation, a larger sample size would be necessary. However, time saved accompanies a loss in precision. In the future, more wards, with preferably long average stays, should be included to efficiently document more cases of nosocomial pneumonia.

PP4

ECONOMIC IMPACT OF SMOKING IN GERMANY Ruff L1, Meyer A1, Volmer T2, Nowak D3 1 Medizinische Klinik, Universitäts-Krankenhaus Eppendorf, Hamburg, Germany; 2GlaxoWellcome, Hamburg, Germany; 3 Institut und Poliklinik für Arbeits und Umweltmedizin, Ludwig Maximilians-Universität, Munich, Germany Smoking is a high-risk behavior affecting the health and economic welfare of a society. Thus it is important to quantify the economic burden of the negative impact due to smoking in Germany. Approximately 33.4% of the male and 20.4% of the female German population are current smokers. OBJECTIVE: This study investigates the healthcare costs of smoking on the six most frequent diseases associated with the inhalation of tobacco smoke: lung cancer (ICD 162; attributable to smoking, 89%), cancer of the mouth and larynx (ICD 140–149, 161; 85%), chronic obstructive pulmonary disease (COPD, ICD 490–491; 73%), coronary artery disease (ICD 410–414; 35%), stroke (ICD 434–438; 28%), and atherosclerotic occlusive disease (ICD 440; 28%). METHODS: A data search was carried out using several literature databases including MedLine and DIMDI, as well as health insurance and Federal Institute of Statistics databases. RESULTS: Total smoking-related costs for these diseases in Germany are 31.4 3 109 DM, equivalent to 49% of the overall costs. In detail (1995 figures), smoking-related lung cancer contributes 4.8 3 109 DM, cancer of the mouth and larynx 2.5 3 109 DM, COPD 11.3 3 109 DM, coronary artery disease 8.6 3 109 DM, stroke 3.1 3 109 DM, and atherosclerotic occlusive disease 1.1 3 109 DM. CONCLUSION: The conservative calculation of the economic impact based on only 6 of more than 25 diseases known to be associated with smoking results already in a burden of more than 31 billion DM. Further detailed analyses are needed to better quantify the health economic impact.

PIM1

COST-EFFECTIVENESS ANALYSIS OF A FIXED COMBINATION (DICLOFENAC 1 MISOPROSTOL) IN THE PREVENTION OF GASTROPATHY IN ARTHRITIS PATIENTS IN BELGIUM Annemans L, Vanoverbeke N, Van Rompay W HEDM, Mechelen, Belgium OBJECTIVES: To calculate the costs and cost-effectiveness of a fixed combination of diclofenac 75 mg 1 misoprostol 200 mg bid compared with diclofenac 75 mg bid in preventing gastropathy in Belgium. METHODS: We developed a new decision-analytical model comparing diclofenac 1 misoprostol (DM) to diclofenac (D), taking into account clinical data on gastrop-

Abstracts

39

athy and data on medical practice in case of gastropathy. The time horizon was 3 months. Dyspepsia rates, ulcer rates in patients with symptoms, and silent ulcer rates were obtained from a large prospective trial (n 5 514) and recalculated by using Bayesian analysis. Current medical management of NSAID-related gastropathy was obtained through a review of 102 patient records in primary care and through expert interviews (2 rounds Delphi method). Costs were taken from the perspective of the health insurance. Unit costs were obtained from official listings; hospitalization costs for peptic ulcer were collected from a database of 58 hospitals. RESULTS: Effectiveness—expressed as number of patients free of symptomatic ulcer and as saved lives—was greater for DM: 4 patients out of 100 developed symptomatic ulcer, compared to 12 with D. In 100,000 patients, there would be 42 fatal cases with DM, compared to 106 with D. At a daily cost of 45.57 BEF for DM and 33.46 for D, the results showed an average cost savings of 2,834 BEF per patient if treated with DM, mainly because of a reduced hospitalization cost. Sensitivity analysis around the confidence intervals of the key variables did not change the conclusions. CONCLUSIONS: Diclofenac 1 misoprostol is a dominating strategy compared to diclofenac alone. It prevents gastropathy and is cost-saving for patients needing medium- to long-term treatment with NSAIDs (3 months).

PIM2

COST OF ILLNESS IN EARLY RHEUMATOID ARTHRITIS STRATIFIED FOR SOCIAL STATUS Merkesdal S, Ruof J, Mau W, Zeidler H Department of Rheumatology, Medical School of Hannover, Hannover, Germany OBJECTIVE: To assess the relationship between different components of cost of illness and social status in an early rheumatoid arthritis (RA) cohort. METHODS: 133 consecutive outpatients (age: 47 6 10 years, 63% female) with early RA (mean disease duration: 7 6 4 months) fulfilling the diagnostic criteria for RA were enrolled. All patients were gainfully employed at onset of symptoms (T1). Data collection was performed using standardized interviews at T1 and 12 months later (T2). The mean observation period was 19 6 4 months. Three major cost components were analyzed: 1) costs for application and monitoring of disease modifying antirheumatic drugs (DMARD; direct costs), 2) loss of productivity due to sick leave, work disability, and other work loss (indirect costs), and 3) patient-related non-reimbursed costs. Social status was measured by the patient education level (low or no degree vs higher degrees). RESULTS: The main results are shown in the table. Indirect costs are significantly higher in patients with lower levels of education and expenses concerning non-reimbursed treatment and medication tend to be less. No differences can be seen in the use of DMARD.

Major cost components dichotomized by education level Education level Low degree (n 5 72) High degree (n 5 61) Student’s t-test

DMARD costs

Indirect costs

Patient-related costs

115 552 DM 90 627 DM p , 0.1

40 095 DM 22 935 DM p , 0.005

690 DM 1252 DM p , 0.1

CONCLUSION: Indirect costs were found to be especially related to social strata in patients with early RA. Further research is needed to identify indicators that may explain these findings. Distinct therapeutic measures must be taken for these patients already in early stages of the disease.

PIM3

QUALITY OF LIFE SCORES AS PREDICTORS OF FUTURE HEALTHCARE RESOURCE USE IN PATIENTS WITH ARTHRITIS Kong SX2, Wolfe F1, Kahler K2 1 Arthritis Research Center, University of Kansas, Wichita, KS, US; 2Merck & Co., Inc., Whitehouse Station, NJ, US

PCS MCS HAQ FT ANX FAT GIS Pain SAT COM Global HCV GIV HOSP NSAID GI

4.70 5.59 1.14 2.59 1.51

7.28 7.36 3.85 0.29 4.69

7.12 5.61 1.42 2.63 1.99

4.83 6.44 2.24 2.34 4.45

5.40 5.16 3.09 0.60 4.23

6.35 3.28 5.38 8.75 6.12 5.50 5.98 7.39 3.45 2.69 2.07 2.54 2.06 1.78 1.67 2.07 3.97 8.96 3.63 2.09

4.78 6.89 4.39 0.24 6.87

6.58 6.82 4.81 1.10 3.12

Z-scores: .1.96 5 p , 0.05; .2.58 5 p: 0.01; .3.28 5 p , 0.01.

OBJECTIVES: Quality of life (QoL) and related measures are increasingly used in the evaluation of patients with arthritis. But a relatively uninvestigated area is the relationship between QoL, healthcare resource utilization and costs. We studied the ability of QoL measures to predict future healthcare resource use. METHODS: 412 patients with osteoarthritis of the knee or hip and 796 patients with rheumatoid arthritis completed QoL assessments. Their healthcare resource use was measured 6 months later. QoL assessments included SF-36 physical and mental component scores (PCS, MCS), health assessment questionnaire (HAQ), feeling thermometer (FT), pain, fatigue (FAT), anxiety (ANX), satisfaction with health (SAT), and comorbidity (COM). Healthcare resource use included visits to all healthcare workers, GP/internal medicine visits, hospitalizations, NSAID usage, and gastrointestinal (GI) drug use. RESULTS: In general, after controlling for the effect of demographic variables, the best predictors for physician visits were QoL mental components, global health measures, and patient global assessment of health. Comorbidity was the best predictor for hospitalization. Use of

40

Abstracts

NSAIDs was not influenced by any QoL variables. Comorbidity, severity of GI symptoms (GIS), and the MCS all had a significant impact on GI medication use. In multivariate analyses, MCS, HAQ, satisfaction with health and comorbidity were independent predictors of healthcare resource use. Results have clinical relevance, as, for example, a 1-unit change in HAQ predicts one additional visit per 6 months. CONCLUSIONS: QoL measures and health rating scales both can be used to predict healthcare resource use. Mental health is important in predicting healthcare resource use among patients with arthritis.

PHARMACOECONOMIC METHODS AND COST OF CARE

PMC1

MEDICATION COMPLIANCE AND EXPENDITURE IMPACT ASSOCIATED WITH PHARMACIST’S COGNITIVE SERVICE INTERVENTIONS Smith D1,2, Stergachis A1, Christensen D1,3, Sullivan S1 1 School of Pharmacy, University of Washington, Seattle, WA, US; 2Centre for Health Economics, University of York, York, UK; 3School of Pharmacy, University of North Carolina, Chapel Hill, NC, US OBJECTIVE: The objective was to investigate the compliance and expenditure impact of pharmacist’s cognitive services (i.e., services that require clinical skills, such as patient education). METHODS: Data came from the Cognitive Activities and Reimbursement Effectiveness Project (CARE), and from Washington State Medicaid. There were three groups of pharmacies: a group that was paid to provide cognitive services (n 5 110), a group that was not paid but did document the provision of cognitive service (n 5 90), and a control group (n 5 100) matched to patients in the paid and nonpaid groups. Data were collected for 6 months of both baseline and follow-up. There were 992 patients from the paid group pharmacies, 554 from the nonpaid, and 6,021 from the control group. The outcomes of interest were 1) medication noncompliance, 2) utilization of services, and 3) expenditures. The analysis focused on seven target diseases (epilepsy, depression, heart disease, thrombolic disease, hypertension, diabetes, and schizophrenia). RESULTS: The paid and nonpaid groups were 31% less likely to be noncompliant (OR 5 0.69; 95% CI 0.51– 0.94) than the control group. The likelihood of compliance increased with increasing numbers of services per patient. Total cost was not significantly increased (p . 0.05) for the paid or nonpaid group. There was an increase in prescription cost and use for target diseases in the paid and nonpaid group (p , 0.05), but there was also a contemporaneous trend (p . 0.05) toward decreased cost of other components of care.

CONCLUSION: Paying for cognitive services may be cost neutral overall, while increasing the likelihood of favorable clinical outcomes through gains in medication compliance.

PMC2

RUNNING COST OF TREATING A PATIENT IN A GENERAL MEDICAL WARD IN SULTAN QABOOS UNIVERSITY HOSPITAL IN OMAN Al-Zakwani I, Rao S, Al-Riyamy K, Balkhair A Sultan Qaboos University Hospital, Muscat, Oman OBJECTIVE: To determine the baseline running cost of maintaining a patient in a general medical ward for 1 day. METHOD: Forty-three patients (19 males, 24 females) were retrospectively randomly selected from January through July 1997 from medical admission lists of the following specialties: cardiology, neurology, rheumatology, endocrinology, gastroenterology, infectious diseases, and hematology. Data on patients’ demographics, medications, clinical and laboratory investigations and procedures were collected from patients’ files. Fixed and variable costs were identified and costed. Fixed costs included nursing, biomedical engineering, laundry, catering, cleaning, maintenance, and administration overheads, but excluded the costs of the hospital building per se. Among the variable costs were professional fees, including those of physicians, clinical pharmacists, physiotherapists, and dieticians. The average daily running cost was calculated and comparative analysis was made for differences between sex and age groups. RESULTS: The average daily running cost (in US dollars) of maintaining a patient in a general medical ward was $202. There was no statistical difference (p 5 0.05) for either sex ($196: $203, male: female) or age group ($193: $204, (30: .30). CONCLUSION: This study determined the daily baseline running cost of maintaining a patient in a general medical ward. Both fixed and variable costs were comparable for sex and age groups, but the fixed costs were almost double the variable costs. This was a helpful pilot study and provided a base for further future larger studies in determining the average daily running costs of maintaining patients in general medical wards.

PMC3

SURVIVAL GAINS AND DEGREE OF THERAPEUTIC INNOVATION Messori A, Trippoli S Laboratorio SIFO di Farmacoeconomia, c/o Centro Informazione Farmaci, Azienda Ospedaliera Careggi, Florence, Italy OBJECTIVE: In the comparison between an innovative treatment versus a standard reference treatment, modern cost-effectiveness studies evaluate the survival data of the two treatments and determine the survival gain, using a long-term time horizon. These studies are often conducted in the area of biotechnology drugs (e.g., the new-

Abstracts

41

est anti-cancer drugs, interferons, and recombinant DNA agents, etc.); the methodology for performing these studies generally requires separate estimations of the mean lifetime survival (MLS) for the innovative treatment and for the reference treatment. METHODS: Mean survival gain (MSG), which is expressed in life-years (or months) gained per patient using the new treatment as opposed to the other, is the difference between the two values of MLS observed with the two treatments and is generally calculated according to the methods of Markov or Gompertz. We have conducted a literature search of studies reporting these estimates of MSG. RESULTS: The table below illustrates some values of MSG for a series of different therapeutic problems: CONCLUSION: The MSG can be proposed as a simple and objective index to quantify the degree of innovation of new treatments.

PMC4

LINKING HRQOLS TO LARGE NATIONAL DATA SETS: EUROQOL AND HUI INDICES Rizzo JA, Sindelar JL Department of Epidemiology and Public Health, Yale University School of Medicine, New Haven, CT, US Health-related quality of life (HRQoL) measures are increasingly important for evaluating health programs and health outcomes. Understanding the clinical and socioeconomic determinants of HRQoL is a related, but relatively little studied, topic. It is often difficult to study such relationships for large samples of individuals, because large health survey data sets rarely include HRQoL measures. OBJECTIVE: This paper attempts to address this problem by demonstrating the feasibility and utility of linking algorithms for well-known quality of life indicators to large, nationally representative databases. This process allows the advantages of HRQoL indicators to be combined with the benefits of large, nationally representative data sets. METHODS: We illustrate this approach using the National Medical Expenditure Survey (NMES) and the EuroQol and drawing on a previous match of the Health Utility Index (HUI) with NMES. Although the NMES data allow a good match to the EuroQol and the HUI, the matches are not perfect. Thus we conduct sensitivity analysis using several related specifications of each indi-

cator. We analyze the correlation of the several specifications within each set of specifications of HUI and EuroQol and across the two, HUI and EuroQol. RESULTS: By examining the within domain correlations and the between correlations, we find that the alternate specifications within domain are very similar and that the two HRQoL indicators are similar in many but not all ways. We analyze construct validity by examining the performance of the HUI and EuroQols in empirical situations in which we have knowledge about the relationships (e.g., health decreases with aging). CONCLUSIONS: The results suggest good construct validity.

PMC5

USING A SURROGATE SAMPLE TO DERIVE PATIENT PREFERENCES FOR A COST-UTILITY ANALYSES IN A RANDOMIZED CLINICAL TRIAL Lustig S1, Lipscomb J2, Salsburg D3, Trost C1, Turner R4 1 Pfizer, Groton, CT, US; 2Duke University, Durham, NC, US; 3 New London, CT, US; 4Phase V Technologies, Inc., Wellesley Hills, MA, US Cost-utility analyses conducted in phase III clinical trials have typically obtained preference values for health outcomes from one of two sources: direct interviews of trial subjects, or a generic health status index to assign preference weights to subjects. But the interview approach has been criticized because of concerns about administrative feasibility and generalizability. The generic instrument approach usually requires the investigator to make subjective judgments in mapping a subject’s clinical profile into a (pre-computed) preference weight. OBJECTIVE: This paper introduces a third approach and tests its viability in a trial comparing rheumatoid arthritis therapies. METHODS: Specifically, we: (1) identified a “surrogate sample” of arthritis patients (N 5 123) who were clinically comparable to trial subjects; (2) elicited their preference (utility) ratings for their own current (arthritis-affected) state of health; (3) estimated a multivariable regression model relating the surrogate-sample utility scores to these patients’ functional status, clinical symptoms, and demographics; (4) used these regressions to assign utility scores to trial subjects at each scheduled encounter; and (5) computed encounter-specific mean utility scores for each treatment group.

Disease condition

Innovative treatment

Standard treatment

Acute myocardial infarction High-risk resected cutaneous melanoma Node-positive breast cancer Advanced ovarian cancer Hepatitis C Advanced pancreatic cancer

TPA Interferon Adjuvant CMF Paclitaxel 1 cisplatin Interferon (6 mo) Gemcitabine

Streptokinase No adjuvant therapy No adjuvant chemotherapy Standard chemotherapy No specific treatment Fluorouracil

*with 3% annual discounting; **with 5% annual discounting.

Mean survival gain (mo per pt) 1.08** 15.9** 12.7** 5.2** 3.0** 3.0*

42

Abstracts

RESULTS AND CONCLUSION: The conclusions are: (1) it is feasible to use a surrogate sample to estimate regression models that satisfactorily relate patient utility to readily collectible predictor variables; (2) these models can be applied to assign plausible utility scores to subjects in a phase III clinical trial; (3) the resulting ranking of therapies is in accord with clinical findings from the present trial; and (4) this strategy may be especially attractive for costutility analyses that pool data from multiple trials.

PMC6

ADVANCED SENSITIVITY ANALYSIS BY THE INTEGRATION OF POINT-SENSITIVITY AND RANGE-SENSITIVITY USING PROBABILITY DISTRIBUTIONS Nuijten MJC The Lewin Group, Hoofddorp, The Netherlands The level of uncertainty regarding the outcome of a pharmacoeconomic study cannot be covered completely by the statistical methods routinely employed to handle uncertainty in clinical research. Sensitivity analysis is the most common methodology to deal with the extra uncertainty associated with pharmacoeconomics, which has also been incorporated in recent guidelines on health care evaluation. However, the execution of a sensitivity analysis and the interpretation of the results is not yet standardized, which may lead to subjectivity and consequently weakening of the value of economic evaluations. OBJECTIVE: The concept of point-sensitivity was used to develop a method which allows a more objective judgment of the robustness of a model by avoiding the subjectiveness in most current economic studies. RESULTS: However, a limitation of this method is that the sensitivity measurement is based on a uniform distribution of the variables, which may actually have different distributions. A second limitation is that the overall sensitivity measure is based on a subjectively chosen range, excluding the impact of values outside the range on overall sensitivity. CONCLUSION: The present study presents a refinement of the method by the incorporation of probability distributions, which allow a more accurate assessment of the level of uncertainty in the model. In addition, a bootstrap method is used to create probability distribution for fixed input variables based on a limited number of data points.

PMC7

DIMINISHING THE COST OF HIGHLY PRESCRIBED DRUGS IN THE COMUNIDAD VALENCIANA (SPAIN) Larruga J, Diaz R, Garcia J, Moreno R Conselleria de Sanidad, Generalitat Valenciana, Barcelona, Spain OBJECTIVES: Our objective is to diminish the cost of a group of highly prescribed drugs in the Comunidad Valenciana (Spain) through the promotion of prescribing

them by drug (generic) name (not registered [brand] names), and the corresponding dispensation at pharmacies of a selected group of registered drugs with the same composition and size and lower price (i.e., reference drugs). METHODS: The drugs representing about 50% of the drug prescription costs for the Health Service in June 1995 were selected through analysis of pharmacy bills. The most prescribed sizes for each drug were studied in order to select those with minimum prices as reference drugs. An agreement was concluded with pharmacies to dispense these trade names when the prescription was made by drug name. Physicians were informed in order to promote prescription by drug name. RESULTS: Thirty-four drugs were responsible for 45% of the drug prescription bill of the Health Service in the Comunidad Valenciana in June 1995. The first 10 of them were Omeprazole, Nimodipine, Ciprofloxacin, Enalapril, Calcitonin, Diltiazem, Fluoxetine and Nifedipine. Globally, prescriptions by drug name have increased from 0.2% to 0.34% in 5 months. Prescription of registered names of minimum price has increased markedly for nearly each drug. For example, with Amoxicyllin, prescription of reference drugs in 1995 was 14.62% of the total of Amoxicyllin prescriptions compared with 35.92% in 1996; for Diclofenac, prescriptions of reference drugs represented 15.02% of the total of Diclofenac prescriptions in 1995 compared with 31.85% in 1996. CONCLUSIONS: The minimum increase in dispensation of prescription by drug names is influenced by the fact that in Spain no generic drugs are available, and so there is a strong tradition of prescribing by registered drug names. The agreement with pharmacies represents the dispensation of a group of selected trade names of minimum prices (reference drugs); physicians have preferred prescribing directly the names included in this group. The effect of diminishing costs has been obtained with this intervention, but not by the increase in prescriptions by drug name, as the impact of this measure in prescribing habits has been slight.

PMC8

META-ANALYSIS OF ECONOMIC EVALUATIONS: A METHODOLOGICAL APPROACH AND CASE STUDY Pang F, Drummond M Centre for Health Economics, University of York, York, UK Meta-analysis has become an increasingly accepted research tool for producing overviews of the results of clinical studies. Recently, there has been a rapid increase in the number of economic evaluations performed, and there is both the need and the potential to produce overviews for policy makers. OBJECTIVE: The purpose of this research was to investigate whether standard meta-analytical approaches (as applied to clinical data) could be applied to economic evaluations. As a case study, an economic meta-analysis (EMA) was conducted of cost-effectiveness analyses involving

Abstracts antithrombotic agents (standard heparin, LMW heparin, warfarin) as prophylaxis against deep-vein thrombosis (DVT) following total hip arthroplasty. METHODS: The EMA was based on an adapted Cochrane Review framework. The objectives were (1) to determine whether prophylaxis versus no prophylaxis is cost-effective; (2) to determine the most cost-effective agent; and (3) to explore differences in cost-effectiveness between settings. Before the literature search (databases/hand searches), selection criteria were specified. Each eligible study was assessed for economic and clinical quality using modified checklists, and various clinical variables (e.g., DVT, PE rates; mortality) and economic variables (e.g., drug acquisition costs, treatment costs) were extracted and pooled. All costs were standardized to US dollars (1997). RESULTS: Thirty-five economic studies were retrieved, of which 14 met the criteria for inclusion (US 5 5, Swe-

43 den 5 4, Canada 5 2, Denmark 5 2, UK 5 1). Economic quality was generally high (71–86%), but the clinical quality was less certain (3 “A”, 4 “B”, 7 “ungradeable”). Although considerable variability existed across studies, the pooled analysis showed an overall reduction in both costs and DVT, PE and mortality rates associated with prophylaxis versus no prophylaxis. In preventing DVT and PE, LMW heparin was more effective than standard heparin, but similar to warfarin. Although the acquisition cost of LMW heparin was greater than both alternatives, this result was offset by the reduced treatment cost for complications. Costs tended to be higher in the US. CONCLUSION: From this case study, we have employed a systematic approach for reviewing economic evaluations, and have highlighted ways the standard meta-analytical methodology would need to be adapted to achieve reliable and useful overviews.

Volume 2 • Number 1 • 1999 VALUE IN HEALTH

Contributed Workshop Presentations

W5

TRACK 1: DECISION SUPPORT SYSTEMS AND MODELS

W4

MONTE CARLO SIMULATION IN HEALTHCARE MODELS Richter A, Mauskopf JA Research Triangle Institute, Research Triangle Park, NC, US In this workshop we will focus on Monte Carlo disease simulations and how they can be used to perform economic evaluations of healthcare interventions. Monte Carlo disease simulation is a modeling technique that operates on a patient level basis, explicitly estimating the effect of variability among patients in both underlying disease progression patterns and in individual responsiveness to treatments. Typical outputs from these simulations are patient functional status, life-years, quality-adjusted life-years, and associated costs, all of which can be appropriately discounted. The output information is presented in the form of distributions, which can be used to estimate mean or median values and confidence intervals for the outcomes of interest. These results can then be used to compute cost-effectiveness ratios and other drug value measures. Monte Carlo disease simulation also allows decision-makers to address the question of risk that is associated with smaller populations who may not tend to the “average” results generated by Markov models or simulations of large populations. In this workshop we describe how to create a Monte Carlo simulation model and how different types of uncertainty can be incorporated into the model. We will develop a basic Monte Carlo simulation of HIV/AIDS in @Risk showing the step-by-step creation of the model. We will run the simulation and show the typical outcomes that can easily be tracked by this type of simulation. This workshop should be beneficial to outcomes researchers and healthcare decision-makers who need to incorporate uncertainty about the natural history of a disease and the impact of alternative disease management strategies for individual patients into their drug value analyses.

© ISPOR 1098-3015/99/$10.50/44 44–50

A NATURAL HISTORY OF ALZHEIMER’S DISEASE: PATH-DEPENDENT EVALUATION OF CLINICAL AND ECONOMIC OUTCOMES USING DECISION DATABASE METHODOLOGY Arcelus A1, Bhattacharjya A2, Longo AE1, Kosik KS3, Finkelstein SN4 1 Analysis Group/Economics, Cambridge, MA, US; 2Janssen Research Foundation, Titusville, NJ, US; 3Harvard Medical School, Boston, MA, US; 4Massachusetts Institute of Technology, Cambridge, MA, US The model was developed to describe and simulate the natural history of Alzheimer’s Disease over a 2-year period, using an innovative decision database methodology. The decision database models the path-dependent nature of Alzheimer’s Disease, whereby a patient’s disease outcome is dependent on the specific secondary conditions experienced. The decision model tracks several aspects of Alzheimer’s Disease: participation in an intervention program, disease stage, side effects, secondary conditions, clinical complications, institutionalization, and disease progression. This model reports the number and percent of patients experiencing disease outcomes and the total direct and indirect costs associated with treating those patients. The decision database methodology overcomes the “memory-less” limitation of Markov-based models, the extensive time requirement of Monte Carlo models, and the inability of decision tree software to deal with large, complex models. The workshop is designed to provide an overview of the decision database methodology and will focus on the capability of the methodology to analyze the path-dependent nature of Alzheimer’s Disease. The workshop will also describe the model development process, including identification of relevant parameters, model structure, data sources, and application software. In addition, the workshop will demonstrate the decision database software’s ability to create customized user interfaces and reports. The decision database methodology enables researchers to examine clinical and economic aspects of Alzheimer’s Disease, using a rigorous, probabilistic tool that is able to analyze the interaction and path-dependent nature of disease variables. This methodology can be applied to the study of other diseases.

44

Abstracts

45 W6

THE APPROPRIATENESS OF COMMON MEDICAL DATA SOURCES: RANDOMIZED TRIALS, METAANALYSES, EPIDEMIOLOGICAL AND OBSERVATIONAL STUDIES FOR PHARMACOECONOMIC ANALYSES Hasford J, Sangha O Department of Medical Informatics, Biometry and Epidemiology, University of Munich, School of Medicine, Munich, Germany The validity of pharmacoeconomic analyses relies on the quality and appropriateness of the medical data sources used. Randomized trials, meta-analyses, epidemiological and observational studies will be presented and the pros and cons discussed, with special emphasis on topics such as internal and external validity, comparator treatments, surrogate measures of outcome, assessment of adverse reactions, the place of diagnostics, and censored observations. Our own analyses, along with examples from the literature, will be presented to describe the pitfalls encountered when using common medical data sources. We aim firstly to sensitize the economist about shortcomings when medical data are used for purposes they were not originally collected for, and secondly to enable physicians to critically assess pharmacoeconomic studies that are based on empirical data.

W11

DECISION ANALYSIS IN CANCER THERAPY, A NEW WAY FOR EVALUATION OUTCOMES WITH THE ANALYTIC HIERARCHY PROCESS (AHP) Reibnitz C, Silva, Saavedra M School of Public Health, University of Bielefeld, Bielefeld, Germany The analytic hierarchy process (AHP) is a decision analysis technique for evaluating complex multi-attributed alternatives with conflicting objectives among one or more scores. The AHP uses judgments from the decision maker and hierarchical decomposition to derive a set of ratio-scaled utility measures for decision alternatives. With the AHP, there is no need to estimate a utility function explicitly, because the AHP deals with stated preferences at each step. The model may be used to give a stage 1 breast cancer patient the chance to discover the full extent and weight of the decision for selecting breast cancer treatment, though similar models could be built for any type of treatment (e.g., cost-benefit analysis). The model allows for the opinions of the physician and the legitimate concerns of the patient. Another problem inherent in the medical field is the lag between new treatments and their long-term statistical confirmation. For this reason, it is imperative that physician and patient alter relevant weights, given advancements in medical knowledge, particular situational needs and patient concerns. The issues of concern in the model were identified as the total of psychological effects from

the treatment; expected length of time that the patient can reasonably expect to live a normal life after a surgery decision is acted upon; the quality of preservation of the breast tissue and surrounding structures; side effects which may develop as a result of the surgery; and patient’s background, medical status, and cost. Though this model evaluates general treatment options of surgical, non-surgical, combination, or no treatment, specific alternatives of therapies could also be analyzed.

W12

DEVELOPING INTERACTIVE COMPUTER SOFTWARE FOR EVALUATING THE ECONOMIC AND CLINICAL OUTCOMES OF VARIOUS TREATMENT STRATEGIES McGhan WF University of the Sciences at Philadelphia. Philadelphia, PA, US The objective of this workshop is to demonstrate how interactive software programs are useful tools in healthcare decision-making as well as educational tools on specific diseases and the decision-making process. In this workshop, a specific disease, gastroesophageal reflux disease (GERD), will be used as the example. Decision support systems are tools that allow systematic approaches to decision-making under conditions of uncertainty. There are four steps in decision system development: 1) to identify and bound the decision problem, 2) to structure the decision problem over time, 3) to characterize the information needed to structure the decision, and 4) to perform the analysis to reveal the preferred course of action. In this workshop, a decision tree analysis will be demonstrated. The branches of the tree describing the clinical pathway treatment options will be discussed. The decision nodes (treatment options), probability nodes (healing rates), and resulting outcomes will be demonstrated. Sensitivity analysis (changing one variable, such as one of the costs, or changing two or three variables at one time) will be demonstrated. Interactive software programs have the additional benefit as educational tools to self-instruct on the clinical issues of a specific disease. Attendees will see how the use of photos, graphics, and dictionaries within the decision support software program can enhance the value of the program as a user-friendly educational tool.

TRACK 2: GUIDELINES AND STANDARDS

W8

THE DEVELOPMENT AND IMPLEMENTATION OF FORMULARY SUBMISSION GUIDELINES FOR MANAGED CARE IN THE UNITED STATES Langley P Program in Pharmaceutical Economics and Health Systems Research, University of Colorado Health Sciences Center, Denver, CO, US

46

Abstracts

Managed care groups, physician management organizations, and independent practice associations in the United States are increasingly seeing formulary submission guidelines as a key management tool for cost and quality control within health systems. The purpose of this workshop is to describe the process of guideline development, the analytical and evaluative standards required from those making submissions and the implementation of guidelines within managed care health systems. The focus is on the guidelines recently put in place by Blue Cross Blue Shield of Colorado, Nevada, and New Mexico (with over 0.5 million members), which were developed at the University of Colorado, and similar guidelines in the process of development for the Academy of Managed Care Pharmacy in Washington, DC. Of particular importance in this presentation are the analytical and evaluative standards required in the guidelines. While the guidelines are similar in a number of respects to the revised November 1995 Australian guidelines, the central requirement is for projections to be made of the net cost and outcomes impact on a treating population from the introduction of a new drug or therapy intervention in disease or treatment areas. This is described as a systems impact assessment, which is quite different from the analytical focus of traditional pharmaceutical economics. Rather than base formulary decisions on implications drawn from synthetic decision models, the guidelines ask for cost and outcome impact assessments which are, in principle, verifiable. This means they can serve as a basis for monitoring health impact assessments of set targets and in the context of disease or care management interventions.

W10

PRO AND CON: USE OF DATA FROM FOREIGN PHARMACOECONOMIC INVESTIGATIONS IN RUSSIA Stratchounski L, Rozenson O Smolensk State Medical Academy, Smolensk, Russia Pharmacoeconomic data are extremely useful for the development of optimal treatment strategies in terms of clinical efficacy, safety, and cost of drug therapy. Although several pharmacoeconomic trials have been conducted in Russia, these data are quite scarce in our country. In the majority of cases, foreign data are the only available sources of information. The main advantages of these data are their methodology, design, and modern approaches to data analyses and interpretation. However, Russian doctors and health authorities should be very careful in using data obtained abroad. As far as we are concerned, there are at least four principal limitations for these data in Russia. Firstly, variations in epidemiology of given diseases exist among different countries. Secondly, we see a difference in drug prices; for instance, cefotaxime (1 g) costs 2.44 USD in Russia versus 11.7 USD in the US. Thirdly, the costs of medical service and hospitalization are much less costly in Russia than in Western countries—the cost per ICU day in Russia is about 100 USD (Smolensk Regional Hospital),

whereas in the US it is about 1000 USD (Millard Fillmore Suburban Hospital, Williamsville, NY). This may explain to some extent the lack of sufficient attention paid by Russian physicians to the duration of hospitalization. The total cost of treatment nowadays in Russia predominantly depends on drug therapy cost. Finally, there are differences in financing sources of health services between countries. The Russian healthcare system is predominantly funded by the Government and/or insurance companies. Moreover, there are often different sources of financing within one region.

W17

STANDARDIZATION IN THE PUBLIC HEALTH SYSTEM OF RUSSIA: THE FIRST EXPERIENCE Vorobyev PA1, Aksyuk AV1, Yakimov OS1, Novolodsky VM1, Kuzin VF1, Kobina SA2 1 Moscow Medical Academy, Laboratory of Outcomes and Standardization Research, Moscow, Russia; 2Rhône-Poulenc Moscow Representative Office, Pharmacoeconomics Department, Moscow, Russia Since 1998, the standardization system in Russia has been functioning in the public health system, which corresponds with the state system of standardization and with the international requirements in this field. This system is to solve the following major problems: 1) to assure that citizens shall exercise their rights to generally available free of charge medical care of high quality; 2) to provide the compulsory health insurance system in Russia with a standardizing base; and 3) to provide the certification and quality assessment system to be set up in the health public system with standards and specifications. At the first stage of development, 16 groups of standardization objects were identified and the basic principles formulated. The most important principles are a consensuality principle (i.e., a mutual agreement between all the parties engaged in the development and use of standards), an actuality principle (i.e., compliance of the elaborated documents with legislative and other international and state requirements), and a complexity principle (i.e., solution of problems of standardization in all spheres of citizen healthcare by uniform methods). At the second stage, a working program on the establishment and development of a public health standardization system in Russia was worked out and approved, highlighting the priorities of work, defining a list of priority regulatory documents, and executing this work. Particular emphasis is placed on a group of system-forming documents: a group of documents that regulate the terms, conditions, and quality of the medical services rendered; and a group of documents pertaining to drug provision, medical equipment, informational technologies, etc. Assimilation of the results of new studies in pharmacological economy and evidence-based medicine will be a very important element of development of the documents regulating the rendering of medical services. The specialists of the pharmaceutical company Rhône-Poulenc, whose collaboration helps to introduce mathematical programs and informational technologies into the development of a program

Abstracts

47

of standardization, were among the first to initiate new pharmacoeconomic and pharmacoepidemiological developments. At the third stage, the establishment of a standardization service is envisaged. In addition to regulatory documentation development proper, the service will be engaged in putting these standards and specifications into practice and in analyzing the results on the introduction of specific regulatory documents. The I. M. Sechenov Moscow Medical Academy (its Laboratory of Outcomes and Standardization Research Problems) has been appointed the central research institution to deal with standardization problems. It is entrusted with the duties of a coordinator of research developments, an expert in the existing documents and a developer of a standardizing base (now lacking). In 1999, a conference is to be held and a scientific journal covering standardization problems will begin publication.

W20

STANDARD PROCEDURES FOR ECONOMIC EVALUATION: USING A COMPUTER PROGRAM FOR REPORTING ECONOMIC EVALUATIONS RESULTS (REER) Rovira J1,2, Brosa M2, Abbas I1,2 1 University of Barcelona, Barcelona, Spain; 2Soikos, Barcelona, Spain This workshop is intended for individuals with a basic understanding of the concepts and terminology of economic evaluation of healthcare who want to discuss applied standardization methodologies. Several software tools are available to assist in the development of pharmacoeconomic models. However, the lack of a standardized set of rules to present economic evaluation results may limit the comparability of studies carried out by different analysts. The objective of this session is to define a set of standardized procedures to improve the comparability of results of economic evaluation studies by means of user-friendly computer software for Reporting Economic Evaluation Results (the REER software). Two different prototypes will be presented: an application aimed at presenting economic evaluation results; and a health costs database (HCD). The main program (REER) is divided into three basic modules: 1) definition of options (treatment/intervention comparators, subjects); 2) data inputs (resources, unit costs, health outcomes, discounting and sensitivity analysis); and 3) outputs (differences in costs and effects, dominance analysis, incremental analysis, and sensitivity analysis). A prototype version of the REER software will be presented by means of two published pharmacoeconomic analyses. The Health Costs Database Software (HCDS) enables the collection, categorization, and storage of unit cost data to be used in different types of studies (e.g., cost analysis, economic evaluation). The database, which simplifies the retrieval of average and actualized cost figures, makes rapidly available updated unit cost data which can be used directly (e.g., within the REER software), as a first calculation, or can be contrasted with later cost estimations. Problems with obtaining standardized results of eco-

nomic evaluation studies and the usefulness of programs such as the REER software will be examined, and future development of complete and fully operational software tools will be discussed.

TRACK 3: QUALITY OF LIFE ASSESSMENT

W1

THE DARTS DIABETES DATABASE: HOW CAN QUALITY OF LIFE MEASUREMENTS FOR OUTCOMES RESEARCH BE INCORPORATED? Evans JMM, Ruta DA, Davey P, Morris AD The DARTS/MEMO Collaboration, Dundee University, Scotland The DARTS database is a validated, population-based register of 9,000 patients with diabetes in the population of Tayside, Scotland, compiled by the record-linkage of independent datasets. These include diabetes clinic records, hospital activity data, biochemistry results, and a database of dispensed prescriptions. The database can be used for epidemiological, outcomes, and health services research. Diabetes is a disease with high morbidity and mortality, for which treatments are onerous and time-consuming. Thus the nonclinical or psychological dimension should not be ignored, and so the incorporation of quality of life measures into DARTS has been proposed. A wide range of disease-specific and generic measures that tap into many different facets of quality of life have already been used in diabetes. The challenge is to identify a valid and reliable package of measures that could be practical for the diabetic patients of Tayside. In this workshop, the approach adopted to identify such a package for routine use in DARTS will be described and discussed. It will include a critical review of quality of life measures that have already been used in diabetic patients, the results of a survey of diabetic patients and diabetic health professionals to identify those aspects of quality of life deemed to be particularly relevant to diabetes, and a discussion of the validation of the measures chosen in a pilot group of patients. The workshop would thus be of interest to researchers working with quality of life outcome measures, and/or to those involved in the development and application of diseasespecific databases.

W13

THE USE OF PROXIES TO COLLECT HEALTHCARE RESOURCES AND QUALITY OF LIFE DATA Crawford B1, Evans C2, Abetz L3 1 MAPI Values, Boston, MA, US; 2Astra Pharmaceuticals, Westborough, MA, US; 3MAPI Values, Cheshire, UK The measurement of a product’s intrinsic value has become an essential part of a new technology’s reimbursement and marketing strategy. The key tactics in the evaluation of a

48

Abstracts

product’s value are the economic and humanistic consequences of the new therapy relative to the gold standard. However, in many circumstances, the patient is unable to provide an accurate assessment of healthcare resources consumed or their quality of life and functional status. Researchers have opted for the use of proxies in these instances, under the presumption that a larger sample size increases the power to detect meaningful differences. This presumption, however, neglects the inaccuracies inherent in proxy reporting and thus may inhibit the reliability of the estimates obtained. This workshop will review and evaluate the use of proxy respondents in the collection of resource utilization and quality of life data. A critical examination will be given to their potential influence on study results. Suggestions for overcoming these issues will also be presented and discussed. This session is directed at individuals in pharmaceutical firms, contract research organizations (CROs), and consultant companies responsible for the design and conduct of pharmacoeconomic evaluations.

W18

HEALTH STATUS AND QUALITY OF LIFE MEASURES IN INTERNATIONAL CLINICAL RESEARCH Marquis P1, Abetz L1, Conway K2, Mear I2 1 MAPI Values, Lyon, France; 2MAPI Research Institute, Lyon, France Assessing the impact of disease and the effect of treatment on patient quality of life (QoL) has become of major importance both to the pharmaceutical industry and the medical profession. The last 20 years have seen the development of a large number of QoL questionnaires, increasingly being used in clinical trials, with a growing emphasis on multinational applications. As a result, there is a continually expanding need for cross-nationally and cross-culturally valid, reliable and responsive QoL instruments. This course will provide key information and operational solutions to implement health status and QoL measures in clinical trials. The content of the course will cover definitions of health status and QoL; types of instruments used, including their strengths and weaknesses; identification of concepts to be measured; and selection of relevant instruments. The second part of the course will focus on strategy for instrument development, linguistic validation and psychometric validation, as well as interpretation guidelines. Participants will acquire a common baseline understanding of health and QoL evaluations, allowing them to become actively involved in development of optimal programs for clinical development and marketing purposes.

TRACK 4: STUDY METHODS

W2

ESTIMATING DRUG EFFECTS: FROM CLINICAL TRIAL RESULTS TO ACTUAL PRACTICE Caro JJ1, Migliaccio-Walle K1, Thizon de Gaulle I2, Coniglio A2 1 Caro Research, Concord, MA, US; 2Bristol-Myers Squibb, Princeton, NJ, US An important element in judging the worthiness of a new drug’s effects is the translation of randomized trial results to actual clinical practice. A key input is the risk experienced by routinely managed patients. The risk rate is important, because the relative risk reduction typically estimated in trials only gains meaning when it is applied to such a reference risk. While the relative risk reduction is widely believed to be generalizable, the reference risk is not. In this workshop, an ongoing study, CAPRA (CAPRIE Actual Practice Rates Analysis) will be used as a case study— along with other published reports—to examine why considering information beyond that obtained in a clinical trial can be critical in assessing the value of a new therapy. The major reasons practice may diverge from trials, along with evidence of this fact, will be introduced. The methods for extending trial results to clinical practice will be demonstrated and discussed, along with alternative approaches to estimating actual practice results from clinical trial experience. The focus will be on the use of epidemiologic studies and databases such as that from Saskatchewan Health. Details of this analysis of records for 12,931 patients used to estimate the risk of first subsequent ischemic events (MI, ischemic stroke, vascular death) in actual practice will be shown. The magnitude of the distortion, with special reference to numbers needed to treat, and the implications for cost-effectiveness analysis will be presented. All available information and methods ought to be employed when considering a new therapy, as it is insufficient to evaluate a new therapy based on trial results alone.

W3

ESTIMATING POPULATION BUDGET AND HEALTH IMPACTS OF NEW TREATMENTS Mauskopf JA Research Triangle Institute, Research Triangle Park, NC, US Most pharmacoeconomic studies present estimates of the impact of new treatments on expected lifetime costs and health outcomes for typical individuals with a given disease. However, national or local healthcare decisionmakers also need to know what impact the new treatment will have on their annual budgets and on annual health outcomes for their patient populations. In this workshop, a method for estimating the population impacts of new treatments is presented. How information about the impact of a new treatment on individual patients can be combined with information about the number and type of patients in the patient population to esti-

Abstracts

49

mate the impact of the new treatment on annual budgets and health outcomes is shown. For chronic diseases, the impact of a new treatment on annual budget and health outcomes may change over the first few years, until a new steady state is reached. How these estimates vary with different assumptions about the extent of use of the new treatment is shown. A method for generating these estimates using a model developed to estimate the budget and health outcome impacts of new HIV treatments for state- or federally-funded programs is illustrated. Population estimates allow healthcare decision-makers to evaluate the impact on their budgets and patient health of 1) providing the new treatment to their patients, and 2) ensuring that they have sufficient funds available. This workshop will be of value both to industry pharmacoeconomists and national and/or local healthcare decisionmakers. How to generate analyses of economic and health outcomes in a population format that is likely to be of value for new treatment decisions will be shown.

W7

HANDLING SURVIVAL DATA IN COST-EFFECTIVENESS ANALYSIS Trippoli S, Messori A Laboratorio SIFO di Farmacoeconomia, c/o Centro Informazione Farmaci, Azienda Ospedaliera Careggi, Florence, Italy Handling survival data in cost-effectiveness (CE) studies generally implies an assessment of the survival data presented in clinical trials (measured survival), together with a long-term prediction of life expectancy for the same patients (predicted survival). Because different methods are needed to manage the data of measured survival and of predicted survival, this workshop will present an overview of these methods and discuss their relative advantages and disadvantages. 1. Measured survival: When a single clinical trial is the source of the survival information, traditional methods for constructing survival curves can be utilized in CE studies. When two or more clinical trials are available, combining the data requires a survival metaanalysis. Although the methodology of survival meta-analysis is still under development, some techniques in this area have adequately been tested and can therefore be proposed for general use. 2. Predicted survival: Traditional lifeexpectancy calculations remain the mainstay for predicting survival in healthy subjects or in “cured” patients. Specific methods, however, are needed for correcting normal life expectancy predictions on the basis of the presence of a disease condition. The Gompertz extrapolation technique can be used for conditions where the chance of cure is minimal. Furthermore, other methods have recently been proposed for combining the clinical evidence of a specific survival pattern with the assignment of a normal life expectancy to “long term survivors.” This workshop is designed to provide an overview of the foregoing methods; to report on their use in clinical and economic evaluation; and to present detailed examples of their application. It is

expected that workshop attendees will be primarily researchers and analysts concerned with clinical and economic evaluation, particularly where multi-center collaboration is involved.

W9

ECONOMIC EVALUATION OF REHABILITATION PROGRAMS: THE PERSPECTIVE OF SOCIAL OLD AGE INSURANCE FUNDS IN GERMANY Seitz R1, Wasem J2, Krauth C3 1 Department for Health Economics, University of Ulm, Germany; 2Institute for Medical Information Management, Biometrics and Epidemiology, University of Munich, Germany; 3 Department for Social Medicine and Epidemiology, University of Hannover, Hannover, Germany Society’s perspective of costs and benefits of healthcare programs has been well established in theory and methodology of outcomes research. Far less rigorously examined are issues from the perspective of the payers. For rehabilitation programs in Germany, the perspective of Social Old Age Insurance Funds (SOAIF) is of particular importance. For the workforce, SOAIF finances most rehabilitative interventions; they finance invalidity pensions in cases where their insured can no longer participate in the workforce; and rehabilitation programs are designed to avoid invalidity. In the workshop, theoretical and methodological issues of a payer’s perspective in general, and of SOAIF in particular, are discussed. Four perspectives of SOAIF can be distinguished: 1) minimizing SOAIF’s expenditure—this requires financing rehabilitative intervention in case it can be assumed that the expenditure for invalidity pensions saved through an intervention exceeds the expenditure for the intervention itself; 2) minimizing costs and cases of invalidity according to the (German) social code on rehabilitation— here the costs of interventions are not considered; 3) maximizing a social welfare function—as part of social insurance and the welfare state arrangements, SOAIF could be required to maximize a social welfare function, which could be identical to society’s perspective; and 4) maximizing the utility function of SOAIF’s bureaucrats—according to this perspective, as many rehabilitation interventions will be produced as is necessary to maximize power, prestige and income of the functionaries of the pension funds. Consequences for study designs are discussed.

W14

DESIGNING AN INTERNATIONAL PIGGYBACK STUDY: PROVIDING THE FOUNDATION Crawford B1, Evans C2 1 MAPI Values, Boston, MA, US; 2Astra Pharmaceuticals, Westborough, MA, US Pharmacoeconomic and quality of life analyses are becoming more familiar and hence more important to marketing and reimbursement. The use of piggyback studies as an an-

50

Abstracts

alytic technique has increased rapidly. A piggyback study is performed alongside an existing clinical trial, leaving the clinical components intact and undisturbed. Because clinical safety and efficacy are the drivers behind the trial, efforts are made to ensure that clinical endpoints are not compromised by the economic substudy. This growing emphasis on cost-effectiveness and improvements in humanistic attributes (QOL, ADLs, functioning) has caused many researchers to add poor economic and quality of life substudies in an attempt to tell compelling stories about their products. PE/QOL studies should be developed with a strong foundation. This workshop will provide a detailed review of the fundamental steps necessary to properly design a piggyback study in the international setting. These steps are: the development of a priori objectives and target audience; hypotheses; endpoints to be measured; definition of the target population; selection of appropriate comparators; the time-frame to be analyzed; determination of how the data shall be collected; creation of an analysis plan; costing the data; data analysis; and, lastly, the reporting of the results in a concise and meaningful way. Defining the objectives and hypotheses will be the focus of the workshop, with a review of the other steps. This session is directed at individuals in pharmaceutical firms, contract research organizations (CROs), and consultant companies who are responsible for the design and conduct of pharmacoeconomic evaluations.

W15

SELF-REPORTED RESOURCE UTILIZATION DATA IN PHARMACOECONOMIC TRIALS: THEIR USE AND IMPACT ON STUDY VALIDITY Crawford B1, Evans C2 1 MAPI Values, Boston, MA, US; 2Astra Pharmaceuticals, Westborough, MA, US This workshop offers a critical examination of the use of patient self-reports in pharmacoeconomic evaluations. It is directed at individuals in pharmaceutical firms, CROs, and consultant companies who are responsible for the design and conduct of pharmacoeconomic evaluations. This workshop advances current research by concentrating on how the data collection method—in prospective clinical economic evaluations—influences study results. Scant attention has been paid to the fact that the high level of internal validity found in prospective studies may be compromised by the application of inappropriate methodologies to data collection. Particular attention in this workshop is placed on how validity may be affected by the elapsed time between admission and reporting, the salience of the treatment event and the perceived social desirability of the condition. As part of this workshop, we will demonstrate

specific areas where estimates based on patient self-reports lead to either reliable or suspect values. In particular, we examine areas of hospitalization, outpatient consultations, medication use and indirect costs. The impact of patient self-reports on cost-effectiveness ratios is also discussed. Attendees of this workshop will gain an understanding of current methodological shortcomings in this area. Researchers and readers of pharmacoeconomic studies will gain the skills necessary to better design and evaluate the validity and potential bias in cost effectiveness analyses.

W19

THE BOOTSTRAP STATISTICAL METHOD FOR THE PHARMACOECONOMIC COST ANALYSIS OF SKEWED DATA LeLorier J1, Desgagné A1,2, Castilloux A-M1, Angers J-F1,2 1 Centre de recherche, Centre hospitalier de l’Université de Montréal, Campus Hôtel-Dieu, Canada; 2Département de mathématiques et de statistique, Université de Montréal, Canada In pharmacoeconomics, the comparison of costs generated by the use of two different drugs for the same treatment is of great interest. The problem is especially challenging when the drugs are likely to produce costly adverse effects in a small number of patients. The distribution of cost data is likely to be skewed, and therefore traditional statistical methods may be inappropriate to analyze the difference in mean costs. The objective of this workshop is to discuss the limitations of classical statistical methods used for cost analysis, and to present an alternative method appropriate to deal with skewed data. In this workshop, a pharmacoeconomic cost analysis example with skewed data will be presented. The bootstrap method will be presented as an alternative approach. Cost analysis will be demonstrated on an example with classical and alternative methods. The following statistical methods and their limitations will be discussed: 1) the Student’s t test, which assumes normality of data and is sensitive to data skewed for small to moderate sample sizes; 2) the t test on log-transformed data, which assumes log-normal distributions and equality of variances of log-transformed data for both samples; and 3) the nonparametric rank tests, which assumes that the distributions of data have the same shape and variance. The bootstrap method will be shown to produce the most reliable results for the example given. The use of the nonparametric bootstrap test for most pharmacoeconomic cost analyses with skewed data and with small to moderate sample sizes will be discussed. No restricting assumptions are needed for its application and its ability to deal with skewing makes it an appropriate alternative.