Prescribing patterns for elderly community-dwelling heavy medicinal drug users in Manitoba, Canada and Jämtland, Sweden

08954356(94)00189-8

J Clin

Vol. 48, No. 6, pp. 825-83 I, 1995 Copyright0 1995ElsevierScienceLtd Printedin Great Britain. All rights reserved

Epidemiol

0895-4356/95

$9.50 + 0.00

PRESCRIBING PATTERNS FOR ELDERLY COMMUNITY-DWELLING HEAVY MEDICINAL DRUG USERS IN MANITOBA, CANADA AND JiiMTLAND, SWEDEN DANIEL S. SITAR’*, MRAN BOj?THIUS2-‘, ULF BERGMAN3 and PAUL A. MITENKO’ ‘Clinical Pharmacology Section, University of Manitoba, Winnipeg, Manitoba, Canada *Department of Pulmonary Medicine, &tersund Hospital, ostersund and ‘Department of Clinical Pharmacology, Karolinska institutet, Huddinge University Hospital, Huddinge, Sweden (Received in revised form 31 May 1994)

Abstract-Patterns of drugs prescribed for elderly community-dwelling heavy medicinal drug users were determined from random samplesof data basesand compared between Manitoba, Canada and JImtland, Sweden for the year 1981. Qualitatively, there was 80% concordance between the 20 most frequently prescribed drugs in Manitoba and in Jimtland. Gender differences were rare, but there were notable quantitative differences for prescription of specific drugs between the two jurisdictions. Drugs to treat cardiovascular diseaseswere prominent in both groups with thiazides, triamterene and a-methyldopa more frequently prescribed for Manitoba patients, and furosemide, potassium supplements and digoxin more frequently prescribed for patients from JSimtland. The frequent prescribing of codeine in combination analgesics in Manitoba and phenothiazines in Jgmtland appears to represent geographically disparate approaches to the use of these drugs. Overall, the concordance of prescribed drugs for elderly heavy medicinal drug users from these two jurisdictions appears to outweigh the differences. These results indicate that data from studies of heavy medicinal drug users, at least in the industrialized world, may be more widely applicable than to the geographical location from where they were obtained. Drug use

Geriatrics

Heavy medicinal drug users

INTRODUCTION

In the United States, data from the Boston Collaborative Drug Surveillance Program indicated that there were likely to be millions of Americans who suffered from adverse drug reactions, and that this would predict hundreds of thousands of hospital admissions and tens of thousands of deaths. It was argued that this high prevalence reflected the extensive use *All correspondence should be addressedto: Dr. D. S. Sitar, Clinical Pharmacology Section, University of Manitoba, 770 Bannatyne Ave., Winnipeg, MB R3E OW3 Canada.

Prescribing patterns

rather than the intrinsic toxic potential of drugs [l]. Subsequently, studies on drug use and morbidity were conducted by several groups to assess more critically the potential contribution of prescribed agents on the health of patients [2-81. Several studies have been published which show that prescribing of drugs increases with the age of the patient [g-13]. Problems associated with heavy drug use by older patients have been appreciated for a long time [ 141. It seems reasonable to conclude that identification of the characteristics that mark this group would help in the development of strategies to prevent or

825

826

Daniel S. Sitar et al.

at least minimize drug-related complications. Although different definitions can be applied, the prescription of six or more pharmaceutical preparations per year seems to mark the transition from occasional and reasonable drug use by an elderly individual to heavy and potentially problematic drug use [ 151. Previous studies of drug use and morbidity in Manitoba indicated that about one quarter of acute hospital admissions to a medical service could be associated with prescribed drugs, and that the risk of a drug-related adverse patient event increased with the number of drugs prescribed [S]. Using the hospital admission data, and assessing patterns of prescribed drugs for community-dwelling heavy drug users in Manitoba over a 10 year period, we were able to estimate a risk for drug-associated hospital admissions in this cohort [16]. The purpose of the present study was to determine whether demographic and pharmacologic features described for the Manitoba elderly cohort might be reflected more broadly in older patients residing elsewhere in the industrialized world. MATERIALS

AND METHODS

Study population Individuals 65 years of age or older, who were prescribed more than six pharmaceutical preparations in 198 1, had their prescription records reviewed for the number and types of drugs prescribed. A pharmaceutical preparation can consist of a single drug entity or of several active drugs in a combination product. In addition to determining single or combination drug use, drug prescriptions were characterized as systemic or topical, and as short term or longer term (more than 120 days/year). These features were related to the age and gender of the recipients [ 161. Manitoba. Prescription records were obtained from claims submitted under Manitoba’s universal Pharmacare plan, by which partial reimbursement was provided for annual prescription costs greater than $50.00 [16]. Of Manitoba’s 1,079,OOO population in 1981, 11.5% were older than 65 years and 44% of them spent more than $50.00 on drugs [17]. A random sample provided prescription information on 104 males and 202 females who used six or more drugs in 198 1. Jiimtland. Prescription records were obtained through the continuing outpatient prescription survey maintained by the Swedish Apoteksbo-

laget for the County of Jlmtland in which 18.8% of the population was older than 65 years in 1981. Approximately one-seventh of the County’s 135,000 population is continually monitored for prescription use, and these records have been maintained since 1970 [ 181.A random sample from these records yielded information on 342 males and 539 females 65 years or older who were prescribed six or more drugs in 1981. Methods Prescribed drugs were classified under 160 categories [16], each of which contained one or more drugs with similar pharmacological properties. Frequency of drug use was categorized as a function of age and gender of patients from both geographical locations. Data analyses. Overall frequencies of prescription for specific drugs stratified by gender and age group were compared by the MantelHaenszel Chi-square test for each geographical location. For this analysis, age was grouped by 5-year intervals until 85 years, and all remaining older subjects were allocated to a single residual group. Data were reanalyzed stratified by geographical area and age group to assess potential gender associations with prescription of specific drugs. For both analyses, the Breslow-Day test was completed first to insure that the odds ratio calculated was consistent with a homogeneous distribution. These analyses were completed with the Proc Freq procedure of version 6.08 of the SAS statistical analyses program. The presence of a significant difference was accepted when p < 0.025 to maintain the risk of a type I error at 5%. RESULTS

Sample characteristics In Jlmtland, women comprised 53% of the population older than 65 years, but accounted for 61% of elderly heavy medicinal drug users. Similarly, in Manitoba, women constituted 56% of the older population but accounted for 66% of older heavy medicinal drug users. Thus in women were over repboth jurisdictions, resented among heavy medicinal drug users. The Swedish sample was marginally older than the Canadian sample (75.9 vs 74.3 years respectively), and women were older than men only in the Manitoba sample (74.8 vs 73.3 years respectively). No differences in age distribution

Prescription Patterns for Elderly Heavy Medicinal Drug Users

between the sample populations or between the genders in either Jlimtland or Manitoba were detected. In Jlmtland, 60% of the population older than 65 years received prescribed drugs, and of

827

these, 34% (20% of the total) were prescribed six or more drugs. The proportion of heavy medicinal drug users rose with age from 24% of all drug users at age 65 years to 50% at age 90 years (p < 0.01; r2 = 0.72). In Manitoba, 44% Manitoba

100 80 60 40

0-I 65-69

I 70-74

i 75-79 Age

I 80-84

I 85-89

I 90+

‘:]: ‘i]__. ‘i]: 0

1 70-74

65-69

I 75-79

1 80-84

I 85-89

Males

I 90+

(yrl Manitoba

Females

100 60 60 40 g

20

zz 0 El00 f ;

80 60

E 40 ‘;

20

2

0

!i$i 0

65-69

70-74

75-79

80-84

85-89

90+

65-69

I 70-74

Age (yr) Fig. I(a) and (b)--legend overleaf

I 75-79

I 80-84

I 85-89

90+

Daniel S. Sitar et al.

828

Jlmtland

Males

80 60 40 2

20

iif Q ZlOO f ;

80 60

0’ 65-69

I 70-74

I 75-79

I 80-84

I 85-89

1 90+

0-J 65-69

I 70-74

I 75-79

I 80-84

I 85-89

I 90+

Age (yr) Jiimtland

Females

$00 o 2

80 60

Q) 4o .s 20 $0 El00 $

80 60 40 20 0-I 65-69

I 70-74

I 75-79 Age

I 80-84

I 85-89

I 90+

0-J. 65-69

I 70-74

I 75-79

I 80-84

I 85-89

(yr)

Fig. l(c) and (d). Fig. I. Cumulative frequency distribution by age group (open symbols) for selected prescribed drugs relative to the sample distribution (solid line) for males (panel a) and females (panel b) from Manitoba, and for males (panel c) and females (panel d) from Jgmtland.

I 90+

Prescription Patterns for Elderly Heavy Medicinal Drug Users

829

Table 1. Most commonly prescribed drugs in Manitoba and in JImtland for heavy drug users 65 years or older in 1981. The comparison is against the top 20 prescribed drugs in Manitoba as the reference Jlmtland Manitoba Odds ratio* (95% confidence interval) % use (rank) Drug group P Thiazides 49(l) 32 (6) 20) 20) 2.16(1.49-3.12) Ampicillin 18(13) 20) 20)
graphical locations were identified for approximately half of the drugs listed in the table. Thus patients in Manitoba were prescribed more thiazides, codeine, NSAIDs, topical steroids, ampicillin, cl-methyldopa and triamterene, while patients in Jlmtland were prescribed potassium supplements, acetaminophen, digoxin and furosemide more often. In Jimtland, the most commonly prescribed drugs not appearing in the Manitoba list in Table 1 included cotrimoxazole, with a prescripPrescription frequencies tion incidence of 13% and a ranking of 15 Cumulative prescription frequencies for (incidence in Manitoba of 12% with a ranking selected prescribed drugs relative to the of 21; not significantly different), and phenothisample distribution are presented in Fig. 1. The azines, with a prescription incidence of 17% and most commonly prescribed drug categories are a ranking of 14 (incidence in Manitoba of 7% shown in Table 1, along with the relative rank- and ranking undetermined; p < 0.001). ing in Manitoba and in JBmtland. The data Although there was no difference in the prefrom JBmtland have been standardized to the scription of all nitrates between Manitoba and age and gender characteristics of the Manitoba JBmtland, more patients were prescribed longsample, Qualitatively, there was 70% concor- acting nitrates in Manitoba (11 vs 2%, dance among the top ten prescribed drugs in p < 0.001). Manitoba between the two countries, and this When data were standardized to age group increased to 80% when the top 20 drugs pre- and geographical area, very few gender differscribed in Manitoba were compared with the ences for frequently prescribed drugs were data from JHmtland. Well represented in both detected. Benzodiazepines were prescribed countries were drugs acting on the central ner- more often for females with an odds ratio of vous system, cardiovascular drugs, respiratory 1.39 and a 95% confidence interval of 1.09-l .78 drugs and some miscellaneous agents. However, (p < 0.007). Nitrates were prescribed more highly significant quantitative differences in often for males with an odds ratio of 1.64 and prescription frequencies between the two geo- a 95% confidence interval of 1.20-2.22

of the older population applied for Pharmacare benefits, and of these 49% (21% of the total) were categorized as heavy drug users. The average number of drugs prescribed for older heavy medicinal drug users in JBmtland was 9.8 while in Manitoba it was 8.3 in 1981. There were no differences between the genders in either country with respect to the total number of drugs prescribed.

830

Daniel S. Sitar et al.

(p < O.OOl).This gender difference was reflected solely in the prescription of nitroglycerin.

DISCUSSION

Sample characteristics The two patient samples were remarkably similar in composition. Age and gender differences in the Manitoba population may be attributable to the participation of Canada in World War II, but these are minor. The proportion of heavy medicinal drug users in the older population in 1981 was about one-fifth in both Jimtland and Manitoba. The Jamtlanders received more prescription drugs than did the Manitobans, but this is probably explained by the restrictions on the Pharmacare plan in Manitoba which will not reimburse for some over-the-counter agents such as analgesics, antacids and vitamin preparations. These were therefore not fully recorded in the Manitoba sample, but were included in more extensive Sweden’s prescription plan. , Prescription frequencies Table 1 similarly indicates considerable congruence between the two populations and the type of drugs prescribed, and even the relative ranking of some drug categories is consistent. The most frequently prescribed drugs for older heavy medicinal drug users from Jlmtland in this study are concordant with previously reported findings of most commonly prescribed drugs in this jurisdiction for all patients [18]. Some differences do not have an easy explanation. For instance, the substantially higher use of digoxin in Jamtland is not readily understood. In other instances, the differences are not as real as they appear. Triamterene was used in Canada but was not available in Sweden, and potassium supplements were used in Sweden mostly in combination products with thiazides. It should be noted that the total use of diuretics (thiazides plus furosemide) was similar in both regions. As indicated above, the use of salicylates and acetaminophen is under-reported in Manitoba because of substantial over-the-counter availability of analgesic preparations. The popularity of analgesic preparations containing codeine probably explains the extensive use of this drug in Manitoba, since virtually none was prescribed as a single agent. The relatively frequent

prescription of barbiturates in both jurisdictions is disturbing. Since 198 1, barbiturate-containing products have been withdrawn from the Swedish formulary. In Manitoba, the frequency of prescriptions for barbiturates has been declining steadily since 1975 [16], This pattern of prescribing in Manitoba probably reflects continuing use by patients habituated to barbiturates rather than prescriptions for new patients. At the time of this survey, phenothiazines were being extensively used in Sweden for minor psychiatric disturbances, e.g. anxiolytics, but this was in agreement with earlier national recommendations. However, such usage was recently abandoned in response to the deliberations of an expert committee [19]. Since then, the use of phenothiazines in the County of Jlmtland and in the rest of Sweden has declined considerably [20]. The infrequent use of antidepressants in both jurisdictions must be of some concern in the light of the reported high prevalence of depression in the elderly and the inferred consequences of its under treatment [21,22]. In conclusion, samples from relatively inclusive data banks in Manitoba and Jiimtland suggest that the types and frequencies of drug prescriptions for heavy medicinal drug users may be considerably concordant, at least for industrialized countries. However, application of these data to specific drug categories can be expected to change with time due to the availability of alternative drug therapies. This study supports the interpretation that comparison of prescribed drugs in a single jurisdiction may have broader application and may also identify differences in drug prescribing which deserve more intensive study. Acknowledgemenf-The authors thank the National Corporation of Swedish Pharmacies, and Mary Cheang for assistance with the statistical analyses of data in this manuscript.

REFERENCES 1. Jick H. Drugs-remarkably nontoxic. N Engl J Med 1974; 291: 824-828. 2. Miller RR. Hospital admissions due to adverse drug reactions. Arch Intern Med 1974; 134: 219-223. problems 3. Bergman U. Wiholm B-E. Drug-related cauiing admission to a medical,clinic. Euk J Clin Pharmacol 1981; 20: 193-200. 4. Black AJ, Somers K. Drug-related illness resulting in hosoital admission. J R Co11Physicians Land 1984: 18: 40--41. 5. Ives TJ, Bentz EJ, Gwyther RE. Drug-related admissions to a family medicine inpatient service. Arch Intern Med 1987; 147: 1117-I 120.

Prescription 6. 7.

8.

9. 10. II. 12. 13. 14.

Patterns for Elderly Heavy Medicinal

Westerholm B, AgenHs I, Dahlstrom M, Nordenstam I. Relation between drug utilization and morbidity pattern. Acta Med Scand 1984; (Suppl) 683: 95-99. Wiholm B-E, Westerholm B. Drug utilization and morbidity statistics for the evaluation of drug safety in Sweden. Acts Med &and 1984; (Suppl) 683: 107-l 17. Grymonpre RE, Mitenko PA, Sitar DS, Aoki FY, Montgomery PR. Drug-associated hospital admissions in older medical patients. J Am Geriatr Sot 1988; 36: 109221098. Boethius G. Recording of drug prescriptions in the county of Jlmtland, Sweden. Acta Med Stand 1977; 202: 241-251. Skoll SL, August RJ, Johnson GE. Drug prescribing for the elderly in Saskatchewan during 1976. Can Med Assoc J 1979; 121: 1074-1081. Hale WE, May FE, Marks RG, Stewart RB. Drug use in an ambulatory elderly population; a five-year update. DICP 1987; 21: 530-535. Landahl S. Drug treatment in 70-82-year-old persons. Acta Med Stand 1987; 221: 179-184. Nolan L, O’Malley K. Prescribing for the elderly: Part II. Prescribing patterns: differences due to age. J Am Geriatr Sot 1988; 36: 245-254. Hurwitz N. Predisposing factors in adverse reactions to drugs. BMJ 1969; I: 536-539.

IS. 16.

17.

18. 19. 20. 21. 22.

Drug Users

831

Williamson J, Chopin JM. Adverse reactions to prescribed drugs in the elderly: a multicentre investigation. Age Ageing 1980; 9: 73-80. Grymonpre RE, Sitar DS, Montgomery PR, Mitenko PA, Aoki FY. Prescribing patterns for older heavy drug users livine. in the community. DICP 1991; 25: 186-190. Manitoba Health Services Commission Annual Report 19Bl-g2. Manitoba Health Services Commission, 599 Empress Street, Winnipeg, Manitoba, Canada R3C 2T6. Boethius G, Wiman F. Recording of drug prescriptions in the county of Jiimtland, Sweden. Eur J Clin Pharmacol 1977; 12: 31-35. National Board of Health and Welfare Drug Information Committee, Sweden. Workshop: Treatment with Neuroleptics. 1990; 1: 143-l 54. Svensk LIk~medelsStatistik 92. Apoteksbolaget AB (National Cornoration of Pharmacies). 1992: 1499159. Blazer D, Williams CD. Epidemiology of dysphoria and depression in an elderly population. Am J Psychiat 1980; 137: 439-443. Isacsson G, Boethius G, Bergman U. Low level of antidepressant prescription for people who later commit suicide: 15 years of experience from a populationbased drug database in Sweden. Acta Psychiatr Stand 1992; 85: 444448.