- Email: [email protected]
IN2 COST-EFFECTIVENESS OF DIFFERENT STRATEGIES FOR TUBERCULOSIS CONTROL PROGRAMS IN THAILAND
A504 delivered guidance on five of these, with three being recommended. In Australia, 10/21 of these drugs had licensing approval and seven underwent assessment by the Pharmaceutical Benefits Advisory Committee. All seven received approval for listing on the public reimbursement scheme. Similarly, CADTH assessed 16 drugs by HS in 2004. Of these, 10 received licensing approval but only one gained approval to be listed by the Canadian Expert Drug Advisory Committee. Of these 16 drugs, 14 had licensing approval in Australia and 12 underwent assessment by the PBAC, with only two being declined approval for listing on the public reimbursement scheme. CONCLUSIONS: The current process of pharmaceutical assessment appears to give Australian patients timely access to publicly funded pharmaceuticals and that the introduction of HS for new and emerging drugs would neither decrease the time to access for patients or better inform policymakers than the system already in place. HT2 COST-EFFECTIVENESS OF INTERVENTIONS FOR REDUCING ROAD TRAFFIC INJURIES RELATED TO DRIVING UNDER THE INFLUENCE OF ALCOHOL Ditsuwan V1, Veerman JL2, Bertram M2, Vos T2 1 Thaksin University, Phatthalung, Thailand; 2The University of Queensland, Brisbane, Queensland, Australia OBJECTIVES: To determine the cost-effectiveness of interventions to reduce road traffic injuries caused by driving under the influence of alcohol in Thailand. METHODS: This study used generalized cost-effectiveness analysis. We calculated costs from a health sector perspective. The time-horizon for intervention implementation was 1 year with health outcomes measured over a lifetime. The model covers road traffic crash victims who were injured, disabled, or died due to road traffic crashes. We obtained proportions of alcohol-related crashes, by age and sex of the victims, from the Thai Injury Surveillance system. Random breath testing (RBT), selective breath testing (SBT) and mass media campaigns compared to a “do nothing” scenario. We calculated intervention costs and cost offsets of prevented treatment costs in 2004 Thai Baht and measured benefits in terms of disability-adjusted life-years (DALYs) averted. RESULTS: The study found similar incremental cost-effectiveness ratios (ICERs) of SBT and RBT. SBT was cost saving while RBT had an ICER of 1000 Baht (95% uncertainty interval [UI]: cost saving to 7200) per DALY averted if costoffsets are included. The ICER of mass media campaigns was 8300 (95% UI: cost saving to 45,100). A combination of SBT or RBT and mass media campaigns is very cost-effective compared to a willingness to pay threshold of one times gross domestic product per capita (110,000 Baht) in 2004. CONCLUSIONS: The result shows all interventions are cost-effective, but wide variation is seen in the uncertainty ranges. Sobriety checkpoints and mass media campaigns reduce burden of alcohol-related RTI by 38,000 to 41,000 DALYs. HT3 COST-EFFECTIVENESS OF BLOOD PRESSURE LOWERING WITH A FIXED COMBINATION OF PERINDOPRIL AND INDAPAMIDE IN TYPE 2 DIABETES MELLITUS: A TRIAL-BASED ANALYSIS USING THE ADVANCE STUDY Clarke PM1, Glasziou P2, Alexander J3, Chalmers J4, Patel A5 1 University of Sydney, Sydney, NSW, Australia; 2University of Oxford, Oxford, Oxfordshire, UK; 3University of Queensland, Brisbane, Australia; 4George Institute, Sydney, Australia; 5 George Institute, Sydney, Australia OBJECTIVES: To determine the cost-effectiveness of routine administration, irrespective of blood pressure, of a fixed-dose combination of perindopril and indapamide to patients with Type 2 diabetes. METHODS: Prospective cost-effectiveness analysis within a 20 country randomized trial of 11,140 patients with Type 2 diabetes randomized to perindopril plus indapamide or placebo. We calculated cost per death averted at 4.3 years average follow-up and estimated cost per life-year gained by extrapolation. RESULTS: The ADVANCE trial showed a 14% relative risk reduction in all cause mortality (P < 0.03) and an 18% relative risk reduction (0.8% absolute reduction) in cardiovascular mortality (P < 0.03). Hospital admissions for coronary heart disease and coronary revascularization were reduced by 5%. Per patient perindopril-indapamide cost AUS$1368, but reduced total hospitalization costs by AUS$410 and other medication costs (mainly other blood pressure lowering drugs) by AUS$332. Quality of life, measured by the EQ-5D, was 0.80 (on a 0–1 scale); with no difference between groups. Absolute reduction in all-cause mortality was 1.1%, giving a cost per life saved of AUS$49,200, and life-time extrapolation estimated a cost per life-year saved of AUS$8,470 (discounted at 3%). CONCLUSIONS: The combination of perindopril and indapamide in patients with type 2 diabetes reduces all-cause mortality and appears cost-effective. HT4 HEALTH TECHNOLOGY ASSESSMENT DATABASE IN THAILAND 1 2 2 2 2 2 Kapol N , Lochid-amnuay S , Chalongsuk R , Maitreemit P , Sribundit N , Amrumpai Y 1 Silpakorn University, Nakorn Pathom, Thailand; 2Silpakorn University, A. Muang, Nakhon Pathom, Thailand BACKGROUND: Health technology assessment is a pivotal tool to assist policymakers to decide whether a health technology is efficient. However, the accessibility of health technology assessment information in Thailand is problematic. Attempts to develop Thai health technology assessment database has been raised. The database aims to increase accessibility and utility to policymakers, researchers and practitioners. From January 2008, Thai health technology assessment database has been available online at http://www.db.hitap.net. OBJECTIVES: The objectives of this study are to
4th Asia-Pacific Abstracts determine current contents of the database and evaluate the usefulness of the database. METHODS: A review of the Thai health technology assessment database was conducted to determine the main characteristics of studies, the number of included studies, the number of database members, and number of visitors. Moreover, the use of the database was evaluated by an electronic mail survey to 260 website members. RESULTS: By the end of 2009, the database consisted of 732 articles including three types of studies, 155 full economic evaluation studies, 471 randomized control trials studies and 106 quality of life studies. The number of database members is 711. The number of visitors increases from 1123 to 1717 hits per month in 2008 and 2009, respectively. For the survey results, of 260 members, 41 respondents (15.77%) replied to the electronic mail survey. Most of them used the database for studying and/ or conducting a research. The usefulness of the database was found most in 65.86% of respondents. CONCLUSIONS: Thai health technology assessment database provided useful information to users. However, some types of studies should be included to support potential users’ needs. Moreover, the registration system for the members should be further improved to follow up and effectively disseminate information to users.
PODIUM SESSION I: INFECTIOUS DISEASE STUDIES IN1 ECONOMIC AND CLINICAL BURDEN OF PNEUMOCOCCAL DISEASES AND ACUTE OTITIS MEDIA IN TAIWAN: A NATIONWIDE POPULATION-BASED DATABASE ANALYSIS Chang CJ1, Wang PC2, Huang YC3, Wu BS4 1 Chang Gung University, Taoyuan, Taiwan; 2Cathay General Hospital, Taipei, Taiwan; 3Chang Gung Memorial Hospital, Taoyuan, Taiwan; 4National Yang-Ming University, Taipei, Taiwan OBJECTIVES: Streptococcus pneumonia can cause invasive diseases such as meningitis and bacteremia, and noninvasive diseases such as pneumonia and acute otitis media (AOM), leading to high morbidity and mortality in infants and the elderly. Limited data are available for the burden of these diseases in Taiwan. This study investigates economic and clinical burden of pneumococcal diseases and AOM in Taiwan. METHODS: A retrospective population-based National Health Insurance Reimbursement Database (NHIRD) study was performed to estimate the incidence and direct medical costs of pneumococcal meningitis/bacteremia/pneumonia and AOM from 2002 to 2007 from healthcare provider perspective, including only NHIcovered costs. RESULTS: The incidence of pneumococcal meningitis is 0.1 per 100,000 and highest at age <1 (2.4 per 100,000) and decreases with age. The incidence of pneumococcal bacteremia is 0.9 per 100,000 with peak at 7.1 per 100,000 at age 75 and above, followed by 3.7 at age 2 and 3. The incidence of pneumococcal pneumonia is 165.1 per 100,000 with peak at 1060.9 per 100,000 at age 75 and above, followed by 1230.7 per 100,000 at age 1, followed by 1076.9 per 100,000 at age 2. The AOM incidence is 2,729.1 per 100,000 with peak at 23,078.8 per 100,000 at age 1, and high in children group (9,641 per 100,000 at age 12 and below). Annually, 120 deaths are estimated to cause by pneumococcal diseases. The total annual medical cost is US$110,886,546. The annual cost for pneumococcal meningitis is US$193,021, mainly for inpatients (US$189,794). The annual cost for pneumococcal bacteremia is US$744,573, mainly for inpatients (US$732,282). For pneumococcal pneumonia, the total annual cost is US$74,029,821 mainly for inpatients (US$66,353,242). For AOM, the annual cost is US$35,919,131, US$18,059,540 for inpatients and US$17,859,590 for out-patients. CONCLUSIONS: Our study shows huge burden of pneumococcal diseases and AOM in Taiwan. Prevention of these diseases will help reduce clinical and economic burden. IN2 COST-EFFECTIVENESS OF DIFFERENT STRATEGIES FOR TUBERCULOSIS CONTROL PROGRAMS IN THAILAND Hunchangsith P, Barendregt JJ, Vos T, Bertram M The University of Queensland, Brisbane, Queensland, Australia OBJECTIVES: To evaluate the cost-effectiveness of different strategies to control tuberculosis (TB) in Thailand. METHODS: Strategies included directly observed treatment (DOT) by a health worker, community member, or family member, and a mobile phone “contact-reminder” system, compared to self-administered treatment (SAT). Cost-effectiveness analysis was undertaken using a decision tree model, which had three stages of treatment; initial treatment, re-treatment, and multi-drug resistant TB (MDR-TB) treatment. Costs (2005 international dollars: I$) were calculated based on treatment periods and treatment outcomes. Health outcomes were estimated over the lifetime of smear-positive pulmonary TB patients in disability-adjusted life-years (DALYs). Both costs and health outcomes were discounted at 3%. RESULTS: Costeffectiveness results did not clearly indicate a preference for any of the interventions analyzed. Although the median cost-effectiveness ratio for each DOT intervention was favorable, the uncertainty ranges surrounding the health benefits were wide, including a sizeable probability that SAT could lead to more health gain than DOT strategies. The health gain in DALY for family-member DOT was 9400 DALY (95% uncertainty interval −7200 to 25,000), for community-member DOT was 13,000 (−21,000 to 37,000) and health-worker DOT was 7900 (−50,000 to 43,000). There were costsavings (from less re-treatment or MDR-TB treatment) associated with family-member DOT (–I$9 million (–I$12 million to –I$5 million)) as the trial treatment failure rate was significantly lower than for SAT. The mobile phone reminder system was not cost-effective, as the mortality rate in the small trial of this intervention was much
A505
4th Asia-Pacific Abstracts higher than for each of the other treatment strategies but this was likely to have been due to the selection of an elderly population in this trial rather than a result of the intervention. CONCLUSIONS: The large uncertainty around the potential benefits of DOT strategies prohibits a conclusive statement regarding their cost-effectiveness in comparison to SAT. IN3 ASSESSMENT OF KNOWLEDGE ABOUT TUBERCULOSIS AMONG LIBYAN POPULATION IN NORTH EAST LIBYA Solliman MM1, Hassali MA1, Al-Haddad M1, Hadidan MM2 1 Universiti Sains Malaysia, Minden, Penang, Malaysia; 2Alfatih Medical Sciences University, Tripoli, Libya OBJECTIVES: To evaluate the knowledge of North Eastern Libyan population with regards to the etiology and treatment of tuberculosis (TB). METHODS: A cross sectional study was undertaken in 2009. A pre-validated questionnaire was sent to 1400 residents residing in five cities of North East Libya. At the end of the study a total of 1000 questionnaire was collected. Questionnaire includes questions related to respondent’s demographics, general knowledge, transmission, diagnosis, risk factors, treatment and prevention of TB. All data was analyzed using SPSS version 15.00 Software package (SPSS Inc, Chicago, IL, USA) Chi-square test and one-way ANOVA were used as whenever appropriate, P-value less than 0.05 was considered significant. RESULTS: Almost all subjects (n = 965, 96.5%) in this study had heard about TB. The main sources of knowledge were television (n = 447, 44.7%), health workers (n = 242, 24%) and family members (n = 189, 19%). Majority of the respondents were Libyans (n = 883, 88.3%). Fifty percent of the respondents were males. For the purpose of this research the maximum score of the knowledge which can be obtained by any respondent is 23. Assessment of knowledge score reveals that the mean knowledge score was significantly higher among Libyans (11.7 ± 3.8) than non Libyans (9.7 ± 4.7, t = 26.13), (P < 0.001). In addition those respondents with tertiary educations scored significantly higher knowledge score (11.8 ± 3.5) compared to those of intermediate (11.6 ± 4.4) and illiterate (7.7 ± 5.5), [F = 19.34, P = 0.001]. No significant differences between the demographic variables and knowledge score. CONCLUSIONS: The present study findings suggested that the level of knowledge about TB among residents in Libya was low. Therefore, there is a need for massive health education campaign to be undertaken by policymakers in order to improve the population’s knowledge toward TB. IN4 HERPES ZOSTER-ASSOCIATED ILLNESSES, QUALITY OF LIFE AND HEALTH-CARE COSTS AMONG 180 THAI PATIENTS Aunhachoke K1, Bussaratid V2, Chirachanakul P3, Chua-Intra B3, Dhitavat J2, Jaisathaporn K4, Kaewkungwal J2, Kampirapap K4, Khuhaprema T4, Pairayayutakul K3, Pitisuttithum P2, Sindhvananda J4, Thaipisuttikul Y5 1 Phramongkutklao Hospital, Bangkok, Bangkok, Thailand; 2Mahidol University, Bangkok, Thailand; 3 Ministry of Public Health, Nonthaburi, Nonthaburi, Thailand; 4Ministry of Public Health, Bangkok, Thailand; 5Rajvithi Hospital, Bangkok, Thailand OBJECTIVES: To determine the incidence of zoster-associated symptoms, as well as the impact of zoster on pain, quality of life, work loss, health-care utilization and direct health-care costs among 180 Thai patients seeking medical care for herpes zoster. METHODS: Study design: prospective observational study conducted in seven hospitals in Bangkok and vicinity, Thailand Study population: patients recently diagnosed herpes zoster, and meeting at least one other criterion, as follows: aged ≥50 years; aged ≥20 years with HIV-infection or receiving chemotherapy. All patients were scheduled for five visits (i.e., at Day1, Day7, Month1, Month3, Month 6) for questionnaire based interview to determine zoster-associated pain (Zoster Brief Pain Inventory questionnaire), quality of life (EUROQOL 5-dimension questionnaire), work and productivity (multi-response questionnaire), and health-care utilization and cost. Statistical analysis: Descriptive statistics and Spearman rank correlation coefficients were used. RESULTS: Of 180 patients, whose mean (SD) age was 58.9 (13.82) years, 138 (76.7%), 34 (18.9%), and 8 (4.4%) were age ≥50 years, HIV-infected and immunosuppressed, respectively. Thirteen (7.2%) and 35 (20.6%) patients had zoster ophthalmicus and post-herpetic neuralgia (pain that persisted beyond 3 months of illness), respectively. Peak QoL lost was observed during the first week of study and declined thereafter. Correlations between worst-pain scores and QoL were moderately high, r = 0.54. Of those 77 working patients, 57 reported that they were absent from work, either entired day or part of day, during zoster illness. Thirty patients reported sick leave and the median number of days off was 5.5 (1–52) days. Most patients required only two OPD visits for zoster illnesses. Other health-care services requirement included emergency-room visits (n = 2), ambulance services (n = 3), or hospital admissions (n = 7). The mean (SD) direct health-care cost was 3083.39 (5047.03) Thai Baht. CONCLUSIONS: Herpes zoster causes a significant burden among those who develop it.
PODIUM SESSION I: RESEARCH ON METHODS ME1 METHODOLOGICAL ISSUES IN THE DEVELOPMENT OF AN AUSTRALIAN ALGORITHM FOR THE EQ-5D Viney R1, Norman R1, King M2, Street DJ1, Knox S1, Cronin P1 1 University of Technology, Sydney, Broadway, NSW, Australia; 2University of Sydney, Sydney, NSW, Australia OBJECTIVES: QALY weights for the EQ-5D are typically obtained through time trade-off (TTO) surveys using a sub-set (n = 17 or n = 43) of the 243 health states. Discrete choice experiments (DCEs) are a potentially more flexible approach. We explore the development of EQ-5D algorithms based on both approaches and the impact of selection of health states in each. METHODS: We explore experimental designs for the TTO and DCE approaches via, respectively, a simulation study and a pilot study. An on-line panel sample (n = 1000) completed the DCE. Each choice set presented two health profiles (EQ-5D state and survival duration), and death. A population sample (n = 417) completed a computer based TTO task and the DCE. The TTO incorporated 198 plausible health states. Respondents were randomly assigned to value 11 health states and the worst health state. Separate models were estimated for the DCE and the TTO. RESULTS: Results of the DCE pilot suggest that different experimental designs lead to differences in both the mean and variance of parameter estimates. There is a loss of statistical efficiency arising from using a reduced set of health states. The simulation study demonstrated that many interaction effects are not estimable from a TTO using only 43 health states. Models were estimated based on the functional form for the published algorithm for the EQ-5D, and allowing for interaction terms. While relatively few interaction terms are significant, comparison tests demonstrate that inclusion of interaction terms improves the fit of the model. CONCLUSIONS: Both approaches are broadly consistent with previously published EQ-5D algorithms, with some important differences. DCEs allows exploration of interactions between health states and duration that cannot be estimated with the TTO approach. In both approaches a design that allows for more extensive coverage of the EQ-5D space is appropriate. ME2 VALIDATING THE ACCURACY OF A NOVEL METHOD FOR IDENTIFYING HEALTH CARE-ASSOCIATED INFECTIONS Lee J, Imanaka Y, Sekimoto M, Nishikawa H, Ikai H, Motohashi T Kyoto University, Kyoto, Japan OBJECTIVES: The identification of health care-associated infections (HAIs) is a necessary first step for elucidating infection patterns and evaluating control measures. While the potential advantages to using administrative data for this purpose are recognized, previous attempts have proved unsuccessful. We have developed a novel method of HAI identification based on a series of criteria involving antibiotic utilization patterns available in administrative data. The objective of this study was to validate the accuracy of our HAI identification method using comparative analyses with chart reviewbased identification. METHODS: We retrospectively and contemporaneously identified HAIs in 584 gastrectomy patients admitted to four Japanese hospitals (A~D) using both our method and chart reviews. Chart review analysis was based on CDC criteria for nosocomial infections. The accuracy of our method was tested using Cohen’s Kappa coefficient to quantify the non-random agreement between the two methods, as well as sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV). RESULTS: Overall HAI incidence was 21.6% based on chart review identification. Cohen’s Kappa coefficient was 0.78, and sensitivity and specificity were 0.93 and 0.91, respectively. PPV was 0.75 and NPV was 0.98. CONCLUSIONS: These results indicate that our method of identification has high accuracy, and may be used to estimate HAIs in large populations of patients at the hospital level and above. While the NPV was observed to be higher than the PPV, the relatively low prevalence of HAI incidence may have influenced this result. It is important to note that this method was not designed to pinpoint individual infections, but rather to identify identifications in a large number of patients using administrative data with a far higher accuracy than previously available. Therefore, our method has applications for HAI identification in large groups of patients, elucidating HAI patterns and trends, downstream economic evaluations and multi-institutional comparisons. ME3 COST METHOD: A SIMPLE FRAMEWORK AND WORKING TOOL FOR BUDGET IMPACT ANALYSIS Lo PC1, Shau WY1, Tarn YH2, Chern HD1 1 Center for Drug Evaluation, Taipei, Taiwan; 2Taiwan Pharmacist Association, Taipei, Taiwan BACKGROUND: Budget impact analysis was required for new drugs listing application to the National Health Insurance (NHI) in Taiwan over the last decade. However, a working model has not been proposed until now. OBJECTIVES: To provide a simple framework and working tool to estimate the likely financial impact to NHI due to the submitted new drug. METHODS: To estimate the potential 5-year financial impact annually if a new drug was reimbursed, proposed framework starts with the size of patient population and potential claim amount of existing drugs that the new drug is going to target, and taking into account the clinical position and usage of the new relative to the existing drugs. A simple working tool by using Excel spreadsheet was developed accompanying the framework. Only nine cells need to be input. They are grouped into three parts: 1) predicted annual number of new patients receiving new drug; 2) treatment duration; and 3) drug cost. The first part should reference to local
4th Asia-Pacific Abstracts determine current contents of the database and evaluate the usefulness of the database. METHODS: A review of the Thai health technology assessment database was conducted to determine the main characteristics of studies, the number of included studies, the number of database members, and number of visitors. Moreover, the use of the database was evaluated by an electronic mail survey to 260 website members. RESULTS: By the end of 2009, the database consisted of 732 articles including three types of studies, 155 full economic evaluation studies, 471 randomized control trials studies and 106 quality of life studies. The number of database members is 711. The number of visitors increases from 1123 to 1717 hits per month in 2008 and 2009, respectively. For the survey results, of 260 members, 41 respondents (15.77%) replied to the electronic mail survey. Most of them used the database for studying and/ or conducting a research. The usefulness of the database was found most in 65.86% of respondents. CONCLUSIONS: Thai health technology assessment database provided useful information to users. However, some types of studies should be included to support potential users’ needs. Moreover, the registration system for the members should be further improved to follow up and effectively disseminate information to users.
PODIUM SESSION I: INFECTIOUS DISEASE STUDIES IN1 ECONOMIC AND CLINICAL BURDEN OF PNEUMOCOCCAL DISEASES AND ACUTE OTITIS MEDIA IN TAIWAN: A NATIONWIDE POPULATION-BASED DATABASE ANALYSIS Chang CJ1, Wang PC2, Huang YC3, Wu BS4 1 Chang Gung University, Taoyuan, Taiwan; 2Cathay General Hospital, Taipei, Taiwan; 3Chang Gung Memorial Hospital, Taoyuan, Taiwan; 4National Yang-Ming University, Taipei, Taiwan OBJECTIVES: Streptococcus pneumonia can cause invasive diseases such as meningitis and bacteremia, and noninvasive diseases such as pneumonia and acute otitis media (AOM), leading to high morbidity and mortality in infants and the elderly. Limited data are available for the burden of these diseases in Taiwan. This study investigates economic and clinical burden of pneumococcal diseases and AOM in Taiwan. METHODS: A retrospective population-based National Health Insurance Reimbursement Database (NHIRD) study was performed to estimate the incidence and direct medical costs of pneumococcal meningitis/bacteremia/pneumonia and AOM from 2002 to 2007 from healthcare provider perspective, including only NHIcovered costs. RESULTS: The incidence of pneumococcal meningitis is 0.1 per 100,000 and highest at age <1 (2.4 per 100,000) and decreases with age. The incidence of pneumococcal bacteremia is 0.9 per 100,000 with peak at 7.1 per 100,000 at age 75 and above, followed by 3.7 at age 2 and 3. The incidence of pneumococcal pneumonia is 165.1 per 100,000 with peak at 1060.9 per 100,000 at age 75 and above, followed by 1230.7 per 100,000 at age 1, followed by 1076.9 per 100,000 at age 2. The AOM incidence is 2,729.1 per 100,000 with peak at 23,078.8 per 100,000 at age 1, and high in children group (9,641 per 100,000 at age 12 and below). Annually, 120 deaths are estimated to cause by pneumococcal diseases. The total annual medical cost is US$110,886,546. The annual cost for pneumococcal meningitis is US$193,021, mainly for inpatients (US$189,794). The annual cost for pneumococcal bacteremia is US$744,573, mainly for inpatients (US$732,282). For pneumococcal pneumonia, the total annual cost is US$74,029,821 mainly for inpatients (US$66,353,242). For AOM, the annual cost is US$35,919,131, US$18,059,540 for inpatients and US$17,859,590 for out-patients. CONCLUSIONS: Our study shows huge burden of pneumococcal diseases and AOM in Taiwan. Prevention of these diseases will help reduce clinical and economic burden. IN2 COST-EFFECTIVENESS OF DIFFERENT STRATEGIES FOR TUBERCULOSIS CONTROL PROGRAMS IN THAILAND Hunchangsith P, Barendregt JJ, Vos T, Bertram M The University of Queensland, Brisbane, Queensland, Australia OBJECTIVES: To evaluate the cost-effectiveness of different strategies to control tuberculosis (TB) in Thailand. METHODS: Strategies included directly observed treatment (DOT) by a health worker, community member, or family member, and a mobile phone “contact-reminder” system, compared to self-administered treatment (SAT). Cost-effectiveness analysis was undertaken using a decision tree model, which had three stages of treatment; initial treatment, re-treatment, and multi-drug resistant TB (MDR-TB) treatment. Costs (2005 international dollars: I$) were calculated based on treatment periods and treatment outcomes. Health outcomes were estimated over the lifetime of smear-positive pulmonary TB patients in disability-adjusted life-years (DALYs). Both costs and health outcomes were discounted at 3%. RESULTS: Costeffectiveness results did not clearly indicate a preference for any of the interventions analyzed. Although the median cost-effectiveness ratio for each DOT intervention was favorable, the uncertainty ranges surrounding the health benefits were wide, including a sizeable probability that SAT could lead to more health gain than DOT strategies. The health gain in DALY for family-member DOT was 9400 DALY (95% uncertainty interval −7200 to 25,000), for community-member DOT was 13,000 (−21,000 to 37,000) and health-worker DOT was 7900 (−50,000 to 43,000). There were costsavings (from less re-treatment or MDR-TB treatment) associated with family-member DOT (–I$9 million (–I$12 million to –I$5 million)) as the trial treatment failure rate was significantly lower than for SAT. The mobile phone reminder system was not cost-effective, as the mortality rate in the small trial of this intervention was much
A505
4th Asia-Pacific Abstracts higher than for each of the other treatment strategies but this was likely to have been due to the selection of an elderly population in this trial rather than a result of the intervention. CONCLUSIONS: The large uncertainty around the potential benefits of DOT strategies prohibits a conclusive statement regarding their cost-effectiveness in comparison to SAT. IN3 ASSESSMENT OF KNOWLEDGE ABOUT TUBERCULOSIS AMONG LIBYAN POPULATION IN NORTH EAST LIBYA Solliman MM1, Hassali MA1, Al-Haddad M1, Hadidan MM2 1 Universiti Sains Malaysia, Minden, Penang, Malaysia; 2Alfatih Medical Sciences University, Tripoli, Libya OBJECTIVES: To evaluate the knowledge of North Eastern Libyan population with regards to the etiology and treatment of tuberculosis (TB). METHODS: A cross sectional study was undertaken in 2009. A pre-validated questionnaire was sent to 1400 residents residing in five cities of North East Libya. At the end of the study a total of 1000 questionnaire was collected. Questionnaire includes questions related to respondent’s demographics, general knowledge, transmission, diagnosis, risk factors, treatment and prevention of TB. All data was analyzed using SPSS version 15.00 Software package (SPSS Inc, Chicago, IL, USA) Chi-square test and one-way ANOVA were used as whenever appropriate, P-value less than 0.05 was considered significant. RESULTS: Almost all subjects (n = 965, 96.5%) in this study had heard about TB. The main sources of knowledge were television (n = 447, 44.7%), health workers (n = 242, 24%) and family members (n = 189, 19%). Majority of the respondents were Libyans (n = 883, 88.3%). Fifty percent of the respondents were males. For the purpose of this research the maximum score of the knowledge which can be obtained by any respondent is 23. Assessment of knowledge score reveals that the mean knowledge score was significantly higher among Libyans (11.7 ± 3.8) than non Libyans (9.7 ± 4.7, t = 26.13), (P < 0.001). In addition those respondents with tertiary educations scored significantly higher knowledge score (11.8 ± 3.5) compared to those of intermediate (11.6 ± 4.4) and illiterate (7.7 ± 5.5), [F = 19.34, P = 0.001]. No significant differences between the demographic variables and knowledge score. CONCLUSIONS: The present study findings suggested that the level of knowledge about TB among residents in Libya was low. Therefore, there is a need for massive health education campaign to be undertaken by policymakers in order to improve the population’s knowledge toward TB. IN4 HERPES ZOSTER-ASSOCIATED ILLNESSES, QUALITY OF LIFE AND HEALTH-CARE COSTS AMONG 180 THAI PATIENTS Aunhachoke K1, Bussaratid V2, Chirachanakul P3, Chua-Intra B3, Dhitavat J2, Jaisathaporn K4, Kaewkungwal J2, Kampirapap K4, Khuhaprema T4, Pairayayutakul K3, Pitisuttithum P2, Sindhvananda J4, Thaipisuttikul Y5 1 Phramongkutklao Hospital, Bangkok, Bangkok, Thailand; 2Mahidol University, Bangkok, Thailand; 3 Ministry of Public Health, Nonthaburi, Nonthaburi, Thailand; 4Ministry of Public Health, Bangkok, Thailand; 5Rajvithi Hospital, Bangkok, Thailand OBJECTIVES: To determine the incidence of zoster-associated symptoms, as well as the impact of zoster on pain, quality of life, work loss, health-care utilization and direct health-care costs among 180 Thai patients seeking medical care for herpes zoster. METHODS: Study design: prospective observational study conducted in seven hospitals in Bangkok and vicinity, Thailand Study population: patients recently diagnosed herpes zoster, and meeting at least one other criterion, as follows: aged ≥50 years; aged ≥20 years with HIV-infection or receiving chemotherapy. All patients were scheduled for five visits (i.e., at Day1, Day7, Month1, Month3, Month 6) for questionnaire based interview to determine zoster-associated pain (Zoster Brief Pain Inventory questionnaire), quality of life (EUROQOL 5-dimension questionnaire), work and productivity (multi-response questionnaire), and health-care utilization and cost. Statistical analysis: Descriptive statistics and Spearman rank correlation coefficients were used. RESULTS: Of 180 patients, whose mean (SD) age was 58.9 (13.82) years, 138 (76.7%), 34 (18.9%), and 8 (4.4%) were age ≥50 years, HIV-infected and immunosuppressed, respectively. Thirteen (7.2%) and 35 (20.6%) patients had zoster ophthalmicus and post-herpetic neuralgia (pain that persisted beyond 3 months of illness), respectively. Peak QoL lost was observed during the first week of study and declined thereafter. Correlations between worst-pain scores and QoL were moderately high, r = 0.54. Of those 77 working patients, 57 reported that they were absent from work, either entired day or part of day, during zoster illness. Thirty patients reported sick leave and the median number of days off was 5.5 (1–52) days. Most patients required only two OPD visits for zoster illnesses. Other health-care services requirement included emergency-room visits (n = 2), ambulance services (n = 3), or hospital admissions (n = 7). The mean (SD) direct health-care cost was 3083.39 (5047.03) Thai Baht. CONCLUSIONS: Herpes zoster causes a significant burden among those who develop it.
PODIUM SESSION I: RESEARCH ON METHODS ME1 METHODOLOGICAL ISSUES IN THE DEVELOPMENT OF AN AUSTRALIAN ALGORITHM FOR THE EQ-5D Viney R1, Norman R1, King M2, Street DJ1, Knox S1, Cronin P1 1 University of Technology, Sydney, Broadway, NSW, Australia; 2University of Sydney, Sydney, NSW, Australia OBJECTIVES: QALY weights for the EQ-5D are typically obtained through time trade-off (TTO) surveys using a sub-set (n = 17 or n = 43) of the 243 health states. Discrete choice experiments (DCEs) are a potentially more flexible approach. We explore the development of EQ-5D algorithms based on both approaches and the impact of selection of health states in each. METHODS: We explore experimental designs for the TTO and DCE approaches via, respectively, a simulation study and a pilot study. An on-line panel sample (n = 1000) completed the DCE. Each choice set presented two health profiles (EQ-5D state and survival duration), and death. A population sample (n = 417) completed a computer based TTO task and the DCE. The TTO incorporated 198 plausible health states. Respondents were randomly assigned to value 11 health states and the worst health state. Separate models were estimated for the DCE and the TTO. RESULTS: Results of the DCE pilot suggest that different experimental designs lead to differences in both the mean and variance of parameter estimates. There is a loss of statistical efficiency arising from using a reduced set of health states. The simulation study demonstrated that many interaction effects are not estimable from a TTO using only 43 health states. Models were estimated based on the functional form for the published algorithm for the EQ-5D, and allowing for interaction terms. While relatively few interaction terms are significant, comparison tests demonstrate that inclusion of interaction terms improves the fit of the model. CONCLUSIONS: Both approaches are broadly consistent with previously published EQ-5D algorithms, with some important differences. DCEs allows exploration of interactions between health states and duration that cannot be estimated with the TTO approach. In both approaches a design that allows for more extensive coverage of the EQ-5D space is appropriate. ME2 VALIDATING THE ACCURACY OF A NOVEL METHOD FOR IDENTIFYING HEALTH CARE-ASSOCIATED INFECTIONS Lee J, Imanaka Y, Sekimoto M, Nishikawa H, Ikai H, Motohashi T Kyoto University, Kyoto, Japan OBJECTIVES: The identification of health care-associated infections (HAIs) is a necessary first step for elucidating infection patterns and evaluating control measures. While the potential advantages to using administrative data for this purpose are recognized, previous attempts have proved unsuccessful. We have developed a novel method of HAI identification based on a series of criteria involving antibiotic utilization patterns available in administrative data. The objective of this study was to validate the accuracy of our HAI identification method using comparative analyses with chart reviewbased identification. METHODS: We retrospectively and contemporaneously identified HAIs in 584 gastrectomy patients admitted to four Japanese hospitals (A~D) using both our method and chart reviews. Chart review analysis was based on CDC criteria for nosocomial infections. The accuracy of our method was tested using Cohen’s Kappa coefficient to quantify the non-random agreement between the two methods, as well as sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV). RESULTS: Overall HAI incidence was 21.6% based on chart review identification. Cohen’s Kappa coefficient was 0.78, and sensitivity and specificity were 0.93 and 0.91, respectively. PPV was 0.75 and NPV was 0.98. CONCLUSIONS: These results indicate that our method of identification has high accuracy, and may be used to estimate HAIs in large populations of patients at the hospital level and above. While the NPV was observed to be higher than the PPV, the relatively low prevalence of HAI incidence may have influenced this result. It is important to note that this method was not designed to pinpoint individual infections, but rather to identify identifications in a large number of patients using administrative data with a far higher accuracy than previously available. Therefore, our method has applications for HAI identification in large groups of patients, elucidating HAI patterns and trends, downstream economic evaluations and multi-institutional comparisons. ME3 COST METHOD: A SIMPLE FRAMEWORK AND WORKING TOOL FOR BUDGET IMPACT ANALYSIS Lo PC1, Shau WY1, Tarn YH2, Chern HD1 1 Center for Drug Evaluation, Taipei, Taiwan; 2Taiwan Pharmacist Association, Taipei, Taiwan BACKGROUND: Budget impact analysis was required for new drugs listing application to the National Health Insurance (NHI) in Taiwan over the last decade. However, a working model has not been proposed until now. OBJECTIVES: To provide a simple framework and working tool to estimate the likely financial impact to NHI due to the submitted new drug. METHODS: To estimate the potential 5-year financial impact annually if a new drug was reimbursed, proposed framework starts with the size of patient population and potential claim amount of existing drugs that the new drug is going to target, and taking into account the clinical position and usage of the new relative to the existing drugs. A simple working tool by using Excel spreadsheet was developed accompanying the framework. Only nine cells need to be input. They are grouped into three parts: 1) predicted annual number of new patients receiving new drug; 2) treatment duration; and 3) drug cost. The first part should reference to local