- Email: [email protected]
Thiazide diuretics for hypertension: Prescribing practices and predictors of use in 194,761 elderly patients with hypertension
IC Tu et al.
The Amer@an Journal of Geriatric Pharmacotherapy
Brief Report Thiazide Diuretics for Hypertension: Prescribing Practices and Predictors of Use in 194,761 Elderly Patients with Hypertension Karen Tu, MDI,2; Norman R.C. Campbell, MD3; Zhongliang Chen, MDI; and FinlayA. McAlister, MD 4 I Institute for Clinical Evaluative Sciences,Toronto, Ontario, Canada; 2University Health Network. Toronto Western Hospital Family Medicine Centre, University of Toronto, Toronto, Ontario, Canada; 3University of Calgary, Canada Departments of Medicine and Pharmacology and Therapeutics, Calgary, Canada; and 4Division of General Internal Medicine, University of Alberta, Edmonton, Alberta, Canada
ABSTRACT Background: Although several small studies have reported underuse of thiazide diuretics for elderly hypertensive patients, those factors which influence initial choice of first-line antihypertensive treatment are unknown. Objectives: The objective of this study was to explore prescribing practices for antihypertensives in the elderly and determine which factors are associated with thiazide diuretic use as first-line treatment. Methods: This population-based cohort study used linked administrative databases for all elderly patients (>66 years of age) first treated for hypertension between July 1, 1994, and March 31, 2002, in Ontario, Canada. Results: Of the 194,761 patients in our cohort, 68,858 (35%) were prescribed a thiazide diuretic as their first antihypertensive agent. On multivariate analysis, factors associated with being prescribed a thiazide as first-line treatment included age (adjusted odds ratio [AOR], 1.72 [95% CI, 1.67-1.78] for octogenarians compared with patients aged 66-69 years) and having multiple comorbidities (AOR, 1.24 [95% CI, 1.16-1.29] for Charlson scores of 2 and AOR, 1.52 [95% CI, 1.37-1.61] for Charlson scores of>3). On the other hand, men (AOR, 0.64 [95% CI 0.63-0.65 ]) and hypertensives with diabetes (AOR, 0.22 [95% CI, 0.21-0.23 ]) were substantially less likely to be prescribed thiazide diuretics as first-line treatment. Socioeconomic status was not associated with use of thiazide diuretics. Conclusions: One third of initial antihypertensive prescriptions for elderly patients were for thiazides in our publicly funded health care system with universal drug coverage. Socioeconomic status did not influence use of thiazides, but age, sex, and comorbidities did. (AmJ Geriatr Pharmacother. 2006;4:161-167) Copyright © 2006 Excerpta Medica, Inc. Key words: antihypertensives, prescribing patterns, elderly, thiazides. Accepted for publication April 13, 2006.
doi:l 0, 1016/j,amjopharm,2006,06,004 1543 5946/06/$19,00
Printed in the USA. Reproduction in whole or part is not permitted.
Copyright © 2006 Excerpta Medic& Inc.
Volume 4
• Number 2
June 2006
I
161
The American Journal of Geriatric Pharmacotherapy
K. Tu et al.
INTRODUCTION
Worldwide, it is estimated that >50% of all older individuals (those aged >65 years) have hypertension. 1 The log-linear relationship between blood pressure and vascular outcomes is strong across all age strata. Antihypertensive treatment has been associated with a reduction in cardiovascular events, and optimal control of blood pressure would prevent almost half of all coronary events (and an even greater proportion of strokes), with the most impact in older individuals and those with isolated systolic hypertension. 2 5 In response to this evidence, guidelines 6 for the treatment of hypertension have been modified to emphasize the importance of diagnosis, treatment, and control of hypertension in elderly patients. Since 1993, the Canadian national hypertension guidelines 7 have recommended thiazide diuretics as first-line treatment in elderly patients with hypertension. However, the Cardiovascular Health Study, s the Framingham Heart Study,9 and the Health Survey for England 1° have reported that, over the past decade, approximately one third of elderly patients with hypertension were prescribed thiazides as first-line monotherapy. It is unknown to what extent the results from these relatively small cohort studies (1242, 4919, and 2827 elderly patients with hypertension, respectively) conducted in specific geographic areas are generalizable to other locations. The factors that influence the choice of initial antihypertensive drug class in elderly patients are also unknown. In contradistinction to some other national guidelines, 6 since 1999 the Canadian national hypertension guidelines 7 have recommended any of 4 drug classes for initial antihypertensive treatment in the elderly: thiazides, long-acting calcium-channel blockers (CCBs), angiotensin-converting enzyme (ACE) inhibitors, or angiotensin receptor blockers, n Canada has a publicly funded health care system with universal coverage for antihypertensive agents in patients aged >65 years. In Ontario, agents from each of the 4 recommended drug classes are available at the same cost to all registered elderly patients; those with incomes less than Can $16,000/y pay $2 per prescription regardless of the cost of the drug prescribed, and those with an income of $ 1 6 , 0 0 0 / y or more pay the first $100 and $6.11 per prescription thereafter each year regardless of the drug prescribed. Therefore, Ontario is one of the few locales in which clinical factors that influence the choice of drug for initial treatment of hypertension can be assessed without being affected by the bias created by cost to the patient, local standardized practice, or a
162
strong endorsement of one class over another by local or national hypertension guidelines. In an earlier cohort analysis, we reported a 30% increase in the population-adjusted rate of new antihypertensive prescriptions in seniors in Ontario between 1994 and 2002 (from 17.9:1000 to 23.3:1000). 12 In the present article, we explore these changes in antihypertensive prescribing practices by drug class over time in more detail and explore the factors that appear to influence whether thiazide diuretics are chosen as firstline treatment. MATERIALS AND
METHODS
As outlined in detail elsewhere, 12 wc identified a cohort of all Ontario residents _>66 years of age who received a new prescription for an antihypertensive agent between July 1, 1994, and March 31, 2002. This cohort was derived by linking the Ontario Drug Benefit prescription drugs database with the Ontario Health Insurance Plan physician claims database (for outpatient visits), the Canadian Institute for Health Information (CIHI) hospitalization database (for inpatient visits), and the Registered Persons Database. A unique encrypted Ontario health card number preserves the exact identification of individuals but allows for the examination of individuals across administrative databases. Patients who were prescribed an antihypertensive agent were deemed to have newly diagnosed hypertension if they did not have another indication for that agent (eg, a history of myocardial infarction or angina, heart failure, arrhythmias, renal disease, esophageal varices, transient ischemic attack or stroke, hyperthyroidism, or migraines using International Classification of Diseases, N i n t h Revision, Clinical Modification codes 13 from the CIHI hospitalization database in the previous 4 years, or in the outpatient visit database in the previous 3 years, or a prescription for a marker medication indicative of those conditions in the Ontario Drug Benefit database in the previous year). 12 For the purposes of this study, we excluded those patients who were prescribed a combination agent as their first antihypertensive prescription. Studies on the validity of these administrative databases have confirmed their high degree of accuracy and comprehensiveness (>95% for both). 14 We examined initial prescribed antihypertensive drug class in 5 a priori defined age categories: 66 to 69, 70 to 74, 75 to 79, 80 to 84, _>85 years of age. In particular, we examined rates of initiation of thiazide and thiazide-like diuretics (which included chlorthalidone, hydrochlorothiazide, or indapamide) versus all other antihypertensive drug classes.
K. Tu et al.
Statistical Analysis
Z2 tests (including the Hosmer and Lemeshow Goodness-of-Fit Test) were used for assessing differences between age groups and other baseline features, including sex, diabetes status (by crosslinking to the Ontario Diabetes Database15), long-term care status (as identified by a long-term care indicator in the Ontario databases), Charlson comorbidity index score (using both the Ontario hospitalization and physician billing databases to identify comorbid conditions), the number of drugs prescribed in the prior year, the year the antihypertensive was first prescribed, and socioeconomic status (according to income quintile based on 2001 census data derived from postal code census geography provided by Statistics Canada, as per previously published and validated methods). 16,17 We performed multivariate logistic regression to determine which factors were independently associated with the initial choice of a thiazide diuretic versus another antihypertensive drug class. Interactions of age with socioeconomic status and age with sex were examined in the model. All analyses were performed with the SPSS statistical software package, version 11.5 (SPSS Inc., Chicago, Illinois). RESULTS
Our cohort consisted of 194,761 elderly patients first prescribed an antihypertensive agent for uncomplicated hypertension between July 1, 1994, and March 31, 2002 (30,433 patients had concomitant diabetes mellitus). As previously described, 12 although prescribing rates changed over time, the 2 most frequently prescribed first-choice antihypertensive drug classes in our cohort were ACE inhibitors and diuretics in all years of our study (Figure): between 1994 and 2002, prescribing rates increased by 81% for ACE inhibitors, 10% for diuretics, and 27% for ]3-blockers, while CCBs decreased by 22%. Overall, 68,858 of these patients (35%) were prescribed a thiazide or thiazide-like diuretic as their first antihypertensive agent. The proportion of newly treated elderly hypertensives started on thiazides as first-line agents increased with age (from 31% of all new antihypertensive prescriptions in those aged 66 to 69 years to 51% of all new antihypertensive prescriptions in those >85 years, P < 0.001), even after adjustment for comorbidities, sex, year of prescription, total number of medications, and socioeconomic status (Table). Septuagenarians (adjusted odds ratio [AOR] 1.20 [95% CI, 1.18-1.23]) and octogenarians (AOR, 1.72 [95% CI, 1.67-1.78]) were significantly more likely to be prescribed thiazides than
The American Journal of Geriatric Pharmacotherapy
patients aged 66 to 69 years, even after adjusting for sex and other factors in the table. Multivariate analyses also demonstrated that longterm care residents and patients with more comorbidity were more likely to be prescribed thiazide or thiazidelike diuretics as their first antihypertensive agent (both, P < 0.001). As there was an interaction between sex and age, we confirmed that the association between older age and increased likelihood of being prescribed a thiazide as first-line treatment was present in both women and men. On the other hand, men and individuals with diabetes were less likely to be prescribed thiazides. There was no gradient in prescribing practices across socioeconomic status. DISCUSSION
Our data confirm the findings of 3 smaller cohort studies in other locations that thiazides are the initial agent of choice in less than half of all elderly hypertensives. However, while the Cardiovascular Health Study s reported a 37% decrease in thiazide prescribing between 1990 and 1999 in 4 centers in the United States, we found that thiazide prescriptions as initial monotherapy for elderly hypertensives in Canada increased by 10% between 1994 and 2002. Over one third of our patients were treated first with a thiazide, and our data demonstrate a clear age gradient in both men and women (with patients in their 70s and 80s more likely to be prescribed a thiazide as first-line t r e a t m e n t - echoing results from 1990-1999 prescribing data in X -O-0--~-D-
"o
Diuretics ACE inhibitors Calcium-channelblockers [3-Blockers Angiotensin II antagonists
'~D
o o_
1994
1996
1998
2000
2002
Year Figure. Annual rate of initial antihypertensive prescriptions for elderly hypertensive patients newly treated in Ontario, Canada, from 1994 through 2002.ACE = angiotensin-converting enzyme.
163
The American Journal of Geriatric Pharmacotherapy
IC Tu et al.
Table. Rates of thiazide use as first-line antihypertensive treatment in various subgroups and odds ratio (OR) estimates by logistic regression model, in 194,761 elderly patients with hypertension. OR Using Diuretics Patients Using Diuretics (n 68,858),
Crude Estimates
Adjusted Estimates
N
no. (%)
(95% CI)
(95% CI)
70,917 58,407 36,778 18,389 10,270
21,701 (30.6) 19,975 (34,2) 13,976 (38,0) 7999 (43,5) 5207 (50.7)
1.00 (Reference) 1,18 (I,15 1,21) 1,39 (i,36 1,43) 1,75 (1,69 1,80) 2.33 (2.24 2.43)
1.00 (Reference) 1,14 (I,II 1,17) 1,30 (i,26 1,34) 1,59 (1,54 1,65) 1.99 (I.90 2.08)
118,901 75,860
47,679 (40, I) 21,179 (27,9)
1,00 (Reference) 0,58 (0,5~0,59)
1,00 (Reference) 0,64 (0,63 0,65)
Diabetes No
164,537
64,476 (39,2)
Yes
30,224
Variable
Age group, y 66 69 70 74 7s 79 80 84 85 Sex Female Male
4382 (14,5)
1,00 (Reference)
1,00 (Reference)
0,26 (0,25 0,27)
0,22 (0,21 0,23)
Long term care resident No Yes
192,281 2480
67,597 (35,2) 1261 (50,8)
1,00 (Reference) 1,90 (I,76 2,06)
1,00 (Reference) 1,41 (I,28 1,56)
Charlson score C~I
185,052
65,044 (35, I)
1,00 (Reference)
1,00 (Reference)
6569 3140
2536 (38,6) 1278(40,7)
1,16(I,II 1,22) 1,26(I,18 1 , 3 6 )
1,24(I,16 1,29) 1,52(I,3~1,61)
1,00 0,98 0,98 1,00 1,06
1,00 0,99 1,00 1,03 1,06
2 _>3 Socioeconomic status I st Quintile 2nd Quintile 3rd Quintile 4th Quintile 5th Quintile Missing economic data
39,753 41,724 37,102 32,976 34,794 8412
14,033 14,520 12,91 I II ,608 12,735
(35,3) (34,8) (34,8) (35,2) (36,6)
(Reference) (0,95 1,01) (0,95 1,01) (0,9~1,03) (I,03 1,09)
(Reference) (0,96 1,02) (0,9~1,03) (I ,0C~1,06) (I,02 1,09)
3051 (36,3)
Note: While year of prescription and total number of drugs were included in the multivariate models for adjustment (OR reported in last column), they were not independently associated with thiazide prescriptions and thus are not featured in this table. Hosmer and Lemeshow Goodness of FitTest:Z 2 34.83 with 8 degrees of freedom, P < 0.001.
the Framingham H e a r t Study9). Finally, while residents of long-term care facilities and patients with multiple comorbidities were 40% to 50% more likely to be prescribed a thiazide as the first agent for their hypertension, men and diabetic hypertensive individuals were one third to three fourths less likely to have a thiazide chosen as their first antihypertensive agent. Our finding of decreased use of thiazide diuretics in men compared with w o m e n is similar to other studies, which raises the question of why it is so. 9,1°,18 20 The data on sex-related differences in adverse events and
164
effectiveness of treatment are inconclusive. Many trials do not report efficacy or adverse events by sex and when reported, those data are often inconsistent. For example, the Medical Research Council Trial of Mild Hypertension 21 reported that men had more adverse events with diuretics than with ~-blockers (the opposite pattern was seen in women). The Treatment of Mild Hypertension Study 22 reported that quality of life improved after diuretic treatment only in women. The Second Australian National Blood Pressure Study investigators 23 reported that diuretics were less efficacious
IC Tu et al.
than ACE inhibitors in older men (but were more efficacious than ACE inhibitors in older women). However, these sex-related differences in the efficacy and safety of thiazides were not observed in the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial. 24 Moreover, although some argue that thiazides might have special advantages in elderly women due to their association with reduced osteoporotic fractures,25 other authors have pointed out that women might be more at risk for thiazide-induced hyponatremia. 26 Regardless, hypertension guidelines available at the time of this study did not identify treatment regimens specifically related to sex. The existing literature related to the impact of socioeconomic status on choice and use of cardiovascular and preventive therapies is conflicting.27 30 This might be due to the variability of access to, and coverage for, health care that can confound the impact of socioeconomic status on treatment decisions. Therefore, we are not surprised that in Ontario, where all elderly individuals are covered by the province's drug benefit program, we did not observe any clinically significant impact of socioeconomic status on the type of antihypertensive prescribed. It might be reasonable to suppose that in other jurisdictions where patients have to pay a greater portion of drug costs out of pocket, a gradient in antihypertensive choices between income quintiles might be present.
Study Limitations While our study examined antihypertensive prescribing over the past decade in a large, representative, and population-based sample of all adults >66 years of age in Canada's largest province, there are a number of limitations inherent to the use of administrative databases. First, while data on prescriptions and cardiovascular comorbidities are easily and accurately identifiable, blood pressure readings and reasons why physicians chose 1 agent over another are not recorded in administrative records. For example, we do not know which patients had noncardiovascular diagnoses that may have impacted the choice of initial antihypertensive agent and are possibly undercoded in administrative databases (ie, gout, benign prostatic hypertrophy, renal stones, or a history of syndrome of inappropriate antidiuretic hormone secretion). Second, our cohort was limited to those patients who actually filled their prescription for an antihypertensive. However, as noncompliance with physician recommendations is unlikely to systematically differ between drug classes before the patient has taken the prescription to the pharmacy, it is reasonable to use our data to impute prescribing practices. Third, since
The American Journal of Geriatric Pharmacotherapy
not all hypertensive Canadians are prescribed antihypertensive drugs, our data cannot be used for estimating the incidence or prevalence of hypertension in this population. 31 However, the positive predictive value of using antihypertensive prescriptions in patients with only blood pressure-lowering indications for these agents to define a hypertensive cohort is -85% (based on data from the National Public Health and Community Health Surveys in Canada from 2000-2003) 32 and is an accepted method in hypertension pharmacotherapy research) 2 Finally, it should be acknowledged that any cohort study using administrative databases is essentially a cross-sectional study, and although we followed each patient in our cohort for 2 years after the index presentation into the cohort, we deliberately did not link with outcomes such as death, myocardial infarction, or stroke (as we felt residual confounding by indication would influence the relationship between prescribed medications and clinical outcomes). CONCLUSIONS
We found that long-term care residents, older patients, and those with multiple comorbidities were more likely to be prescribed thiazides as first-line agents of choice for initial antihypertensive treatment, whereas men and diabetic patients were less likely to be prescribed these agents. We also found that, in this study, socioeconomic status did not affect the treatment selection. Thiazides were chosen as the first-line agent in only 35% of our study population. ACKNOWLEDGMENTS
This research was supported by a Grant-in-Aid (#NA 5459) from the Heart and Stroke Foundation of Ontario. Dr. Tu is supported by a Canadian Institutes of Health Research (CIHR) Short-Term Clinician Investigator Award. Dr. McAlister is supported by an Alberta Heritage Foundation for Medical Research Population Health Scholar Award, a C I H R New Investigator Award, and the University of Alberta/Merck Frosst/Aventis Chair in Patient Health Management. None of the funding agencies had input into the design, conduct, analysis, or reporting of this study. Dr. McAlister reports operating grant funding support from Pfizer Canada (for a trial cofunded by the Heart and Stroke Foundation of Canada and unrelated to this manuscript or topic), and Dr. Campbell reports having received speaking fees and serving as a consultant/ advisory board member for sanofi-aventis, Bayer, BristolMyers Squibb, Pfizer, Servier, Bioval, Fournier, Abbott, Solvay, and Merck Frosst.
165
The American Journal of Geriatric Pharmacotherapy
K. Tu et al.
The authors thank Siraj Khan for his assistance in the preparation o f this manuscript. REFERENCES
1. Joft?es MR, Hamet P, MacLean DR, et al. Distribution of blood pressure and hypertension in Canada and the United States. A m J Hypertens. 2001 ;14:1099-1105. 2. Lewington S, Clarke R, Qizilbash N, et al, for the Prospective Studies Collaboration. Age-specific relevance of usual blood pressure to vascular mortality: A metaanalysis of individual data for one million adults in 61 prospective studies [published correction appears in Lancet. 2003;361:1060]. Lancet. 2002;360:1903-1913. 3. Staessen JA, Gasowski J, Wang JG, et al. Risks of untreated and treated isolated systolic hypertension in the elderly: Meta-analysis of outcome trials [published correction appears in Lancet. 2001;357:724]. Lancet. 2000;355:865-872. 4. Gueyffier F, Bulpitt C, Boissel JP, et al, for the INDANA Group. Antihypertensive drugs in very old people: A subgroup meta-analysis of randomised controlled trials. Lancet. 1999;353:793-796. 5. Wong ND, Thakral G, Franklin SS, et al. Preventing heart disease by controlling hypertension: Impact of hypertensive subtype, stage, age, and sex. A m Heart J. 2003;145:888-895. 6. Chobanian AV, Bakris GL, Black HR, et al. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. The JNC 7 Report. J A M A . 2003;289:25602572. 7. Canadian Hypertension Guidelines. 2005 Canadian recommendations for the management of hypertension. Available at: http://www.hypertension.ca/Documentation/ Recommandation05_va.pdf. Accessed May 22, 2006. 8. Psaty BM, Manolio TA, Smith NL, et al, for the Cardiovascular Health Study. Time trends in high blood pressure control and the use of antihypertensive medications in older adults: The Cardiovascular Health Study. Arch Intern Med. 2002;162:2325-2332. 9. Lloyd-Jones DM, Evans JC, Levy D. Hypertension in adults across the age spectrum: Current outcomes and control in the community. J A M A . 2005;294:4664:72. 10. Primatesta P, Poulter NR. Hypertension management and control among English adults aged 65 years and older in 2000 and 2001. JHypertens. 2004;22:1093-1098. 11. Khan NA, McAlister FA, Lewanczuk RZ, et al, for the Canadian Hypertension Education Program. The 2005 Canadian Hypertension Education Program recommendations for the management of hypertension: Part I I - therapy. Can J Cardiol. 2005;21:657-672.
166
12. Tu K, Campbell NR, Duong-Hua M, McAlister FA. Hypertension management in the elderly has improved: Ontario prescribing trends, 1994 to 2002. Hypertension. 2005;45:1113-1118. 13. International Classification of Diseases, Ninth Revision, ClinicalModification. 6th ed. Washington, DC: US Dept of Health and Human Services, Centers for Disease Control and Prevention, Centers for Medicare and Medicaid Services; 2005. 14. Williams JI, Young W. A summary of studies on the quality of health care administrative databases in Canada. In: Goel V, Williams JI, Anderson GM, et al, eds. Patterns of Health Care in Ontario. The ICES Practice Atlas. 2nd ed. Ottawa, Ontario, Canada: Canadian Medical Association; 1996:339-345. 15. Hux JE, Dis F, Flintoft V, Bica A. Diabetes in Ontario: Determination of prevalence and incidence using a validated administrative data algorithm. Diabetes Care. 2002; 25:512-516. 16. Wilkins R. PCCF + Version 4D User's Guide: Automated Geographic Coding Based on the Statistics Canada Postal Code Conversion Files, Including Postal Codes to December 2003. Ottawa, Ontario, Canada: Health Analysis and Measurement Group, Statistics Canada; July 2004. 17. Mustard CA, Derksen S, Berthelot JM, Wolfson M. Assessing ecologic proxies for household income: A comparison of household and neighbourhood level income measures in the study of population health status. Health Place. 1999;5:157-171. 18. Usher C, Bennett K, Feely J. Evidence for a gender and age inequality in the prescribing of preventative cardiovascular therapies to the elderly in primary care. Age Ageing. 2004;33:500-502. 19. Pears E, Hannaford PC, Taylor MW. Gender, age and deprivation differences in the primary care management of hypertension in Scotland: A cross-sectional database study. Faun Pract. 2003 ;20:22-31. 20. Onder G, Gambassi G, Landi F, et al, for the Investigators of the GIFA Study (SIGG-ONLUS). Trends in antihypertensive drugs in the elderly: The decline of thiazides. J H u m Hypertens. 2001;15:291-297. 21. Medical Research Council Working Party. MRC trial of treatment of mild hypertension: Principal results. BMJ (Clin Res Ed). 1985;291:97-104. 22. Neaton JD, Grimm RH Jr, Prineas RJ, et al, for the Treatment of Mild Hypertension Study Research Group. Treatment of Mild Hypertension Study. Final results. J A M A . 1993;270:713-724. 23. Wing LM, Reid CM, Ryan P, et al, for the Second Australian National Blood Pressure Study Group. A comparison of outcomes with angiotensin-converting-
K. Tu et al.
24.
25.
26. 27.
enzyme inhibitors and diuretics for hypertension in the elderly. N E n g l J M e d . 2003;348:583-592. ALLHAT Officers and Coordinators for the A L L H A T Collaborative Research Group. Major outcomes in highrisk hypertensive patients randomized to angiotensinconverting enzyme inhibitor or calcium channel blocker vs diuretic: The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) [published corrections appear in J A M A . 2003;289:178 and J A M A . 2004;291:2196]. J A M A . 2002;288:29812997. Schoofs MW, van der Klift M, Hofman A, et al. Thiazide diuretics and the risk for hip fracture. A n n Intern Med. 2003;139:476-482. Chow KM, Szeto CC, Wong TY, et al. Risk factors for thiazide-induced hyponatraemia. Q,JM. 2003;96:911-917. Ahmed A, Allman RM, DeLong IF, et al. Age-related underutilization of angiotensin-converting enzyme inhibitors in older hospitalized heart failure patients. South Med J. 2002;95:703-710.
The American Journal of Geriatric Pharmacotherapy
28. Solberg LI, Brekke ML, Kottke TE. Are physicians less likely to recommend preventive services to low-SES patients? Prev Med. 1997;26:350-357. 29. Scott A, SNell A, King M. Is general practitioner decision making associated with patient socio-economic status? Soc Sci Med. 1996;42:35-46. 30. McAlister FA, Murphy NF, Simpson CR, et al. Influence of socioeconomic deprivation on the primary care burden and treatment of patients with a diagnosis of heart failure in general practice in Scotland: Population based study. BMJ. 2004;328:1110. 31. Joff?es MR, Ghadirian P, Fodor JG, et al. Awareness, treatment, and control of hypertension in Canada. A m J Hypertens. 1997;10:1097-1102. 32. Campbell NR, Onysko J, Maxwell C, et al, for the Canadian Hypertension Education Program. Increases in the diagnosis and drug treatment of hypertensive Canadians [abstract]. Presented at: 58th Annual Meeting of the Canadian Cardiovascular Society, October 22-26, 2005; Montreal, Quebec, Canada.
Address c o r r e s p o n d e n c e to: Finlay A. McMistcr, MD, 2E3.24 Walter Mackcnzic Ccntrc, University of Alberta Hospital, 8440 112 Street, Edmonton, Mbcrta T6G 2R7, Canada. E-mail: Finlay'McMistcr@ualbcrta'ca
167
The Amer@an Journal of Geriatric Pharmacotherapy
Brief Report Thiazide Diuretics for Hypertension: Prescribing Practices and Predictors of Use in 194,761 Elderly Patients with Hypertension Karen Tu, MDI,2; Norman R.C. Campbell, MD3; Zhongliang Chen, MDI; and FinlayA. McAlister, MD 4 I Institute for Clinical Evaluative Sciences,Toronto, Ontario, Canada; 2University Health Network. Toronto Western Hospital Family Medicine Centre, University of Toronto, Toronto, Ontario, Canada; 3University of Calgary, Canada Departments of Medicine and Pharmacology and Therapeutics, Calgary, Canada; and 4Division of General Internal Medicine, University of Alberta, Edmonton, Alberta, Canada
ABSTRACT Background: Although several small studies have reported underuse of thiazide diuretics for elderly hypertensive patients, those factors which influence initial choice of first-line antihypertensive treatment are unknown. Objectives: The objective of this study was to explore prescribing practices for antihypertensives in the elderly and determine which factors are associated with thiazide diuretic use as first-line treatment. Methods: This population-based cohort study used linked administrative databases for all elderly patients (>66 years of age) first treated for hypertension between July 1, 1994, and March 31, 2002, in Ontario, Canada. Results: Of the 194,761 patients in our cohort, 68,858 (35%) were prescribed a thiazide diuretic as their first antihypertensive agent. On multivariate analysis, factors associated with being prescribed a thiazide as first-line treatment included age (adjusted odds ratio [AOR], 1.72 [95% CI, 1.67-1.78] for octogenarians compared with patients aged 66-69 years) and having multiple comorbidities (AOR, 1.24 [95% CI, 1.16-1.29] for Charlson scores of 2 and AOR, 1.52 [95% CI, 1.37-1.61] for Charlson scores of>3). On the other hand, men (AOR, 0.64 [95% CI 0.63-0.65 ]) and hypertensives with diabetes (AOR, 0.22 [95% CI, 0.21-0.23 ]) were substantially less likely to be prescribed thiazide diuretics as first-line treatment. Socioeconomic status was not associated with use of thiazide diuretics. Conclusions: One third of initial antihypertensive prescriptions for elderly patients were for thiazides in our publicly funded health care system with universal drug coverage. Socioeconomic status did not influence use of thiazides, but age, sex, and comorbidities did. (AmJ Geriatr Pharmacother. 2006;4:161-167) Copyright © 2006 Excerpta Medica, Inc. Key words: antihypertensives, prescribing patterns, elderly, thiazides. Accepted for publication April 13, 2006.
doi:l 0, 1016/j,amjopharm,2006,06,004 1543 5946/06/$19,00
Printed in the USA. Reproduction in whole or part is not permitted.
Copyright © 2006 Excerpta Medic& Inc.
Volume 4
• Number 2
June 2006
I
161
The American Journal of Geriatric Pharmacotherapy
K. Tu et al.
INTRODUCTION
Worldwide, it is estimated that >50% of all older individuals (those aged >65 years) have hypertension. 1 The log-linear relationship between blood pressure and vascular outcomes is strong across all age strata. Antihypertensive treatment has been associated with a reduction in cardiovascular events, and optimal control of blood pressure would prevent almost half of all coronary events (and an even greater proportion of strokes), with the most impact in older individuals and those with isolated systolic hypertension. 2 5 In response to this evidence, guidelines 6 for the treatment of hypertension have been modified to emphasize the importance of diagnosis, treatment, and control of hypertension in elderly patients. Since 1993, the Canadian national hypertension guidelines 7 have recommended thiazide diuretics as first-line treatment in elderly patients with hypertension. However, the Cardiovascular Health Study, s the Framingham Heart Study,9 and the Health Survey for England 1° have reported that, over the past decade, approximately one third of elderly patients with hypertension were prescribed thiazides as first-line monotherapy. It is unknown to what extent the results from these relatively small cohort studies (1242, 4919, and 2827 elderly patients with hypertension, respectively) conducted in specific geographic areas are generalizable to other locations. The factors that influence the choice of initial antihypertensive drug class in elderly patients are also unknown. In contradistinction to some other national guidelines, 6 since 1999 the Canadian national hypertension guidelines 7 have recommended any of 4 drug classes for initial antihypertensive treatment in the elderly: thiazides, long-acting calcium-channel blockers (CCBs), angiotensin-converting enzyme (ACE) inhibitors, or angiotensin receptor blockers, n Canada has a publicly funded health care system with universal coverage for antihypertensive agents in patients aged >65 years. In Ontario, agents from each of the 4 recommended drug classes are available at the same cost to all registered elderly patients; those with incomes less than Can $16,000/y pay $2 per prescription regardless of the cost of the drug prescribed, and those with an income of $ 1 6 , 0 0 0 / y or more pay the first $100 and $6.11 per prescription thereafter each year regardless of the drug prescribed. Therefore, Ontario is one of the few locales in which clinical factors that influence the choice of drug for initial treatment of hypertension can be assessed without being affected by the bias created by cost to the patient, local standardized practice, or a
162
strong endorsement of one class over another by local or national hypertension guidelines. In an earlier cohort analysis, we reported a 30% increase in the population-adjusted rate of new antihypertensive prescriptions in seniors in Ontario between 1994 and 2002 (from 17.9:1000 to 23.3:1000). 12 In the present article, we explore these changes in antihypertensive prescribing practices by drug class over time in more detail and explore the factors that appear to influence whether thiazide diuretics are chosen as firstline treatment. MATERIALS AND
METHODS
As outlined in detail elsewhere, 12 wc identified a cohort of all Ontario residents _>66 years of age who received a new prescription for an antihypertensive agent between July 1, 1994, and March 31, 2002. This cohort was derived by linking the Ontario Drug Benefit prescription drugs database with the Ontario Health Insurance Plan physician claims database (for outpatient visits), the Canadian Institute for Health Information (CIHI) hospitalization database (for inpatient visits), and the Registered Persons Database. A unique encrypted Ontario health card number preserves the exact identification of individuals but allows for the examination of individuals across administrative databases. Patients who were prescribed an antihypertensive agent were deemed to have newly diagnosed hypertension if they did not have another indication for that agent (eg, a history of myocardial infarction or angina, heart failure, arrhythmias, renal disease, esophageal varices, transient ischemic attack or stroke, hyperthyroidism, or migraines using International Classification of Diseases, N i n t h Revision, Clinical Modification codes 13 from the CIHI hospitalization database in the previous 4 years, or in the outpatient visit database in the previous 3 years, or a prescription for a marker medication indicative of those conditions in the Ontario Drug Benefit database in the previous year). 12 For the purposes of this study, we excluded those patients who were prescribed a combination agent as their first antihypertensive prescription. Studies on the validity of these administrative databases have confirmed their high degree of accuracy and comprehensiveness (>95% for both). 14 We examined initial prescribed antihypertensive drug class in 5 a priori defined age categories: 66 to 69, 70 to 74, 75 to 79, 80 to 84, _>85 years of age. In particular, we examined rates of initiation of thiazide and thiazide-like diuretics (which included chlorthalidone, hydrochlorothiazide, or indapamide) versus all other antihypertensive drug classes.
K. Tu et al.
Statistical Analysis
Z2 tests (including the Hosmer and Lemeshow Goodness-of-Fit Test) were used for assessing differences between age groups and other baseline features, including sex, diabetes status (by crosslinking to the Ontario Diabetes Database15), long-term care status (as identified by a long-term care indicator in the Ontario databases), Charlson comorbidity index score (using both the Ontario hospitalization and physician billing databases to identify comorbid conditions), the number of drugs prescribed in the prior year, the year the antihypertensive was first prescribed, and socioeconomic status (according to income quintile based on 2001 census data derived from postal code census geography provided by Statistics Canada, as per previously published and validated methods). 16,17 We performed multivariate logistic regression to determine which factors were independently associated with the initial choice of a thiazide diuretic versus another antihypertensive drug class. Interactions of age with socioeconomic status and age with sex were examined in the model. All analyses were performed with the SPSS statistical software package, version 11.5 (SPSS Inc., Chicago, Illinois). RESULTS
Our cohort consisted of 194,761 elderly patients first prescribed an antihypertensive agent for uncomplicated hypertension between July 1, 1994, and March 31, 2002 (30,433 patients had concomitant diabetes mellitus). As previously described, 12 although prescribing rates changed over time, the 2 most frequently prescribed first-choice antihypertensive drug classes in our cohort were ACE inhibitors and diuretics in all years of our study (Figure): between 1994 and 2002, prescribing rates increased by 81% for ACE inhibitors, 10% for diuretics, and 27% for ]3-blockers, while CCBs decreased by 22%. Overall, 68,858 of these patients (35%) were prescribed a thiazide or thiazide-like diuretic as their first antihypertensive agent. The proportion of newly treated elderly hypertensives started on thiazides as first-line agents increased with age (from 31% of all new antihypertensive prescriptions in those aged 66 to 69 years to 51% of all new antihypertensive prescriptions in those >85 years, P < 0.001), even after adjustment for comorbidities, sex, year of prescription, total number of medications, and socioeconomic status (Table). Septuagenarians (adjusted odds ratio [AOR] 1.20 [95% CI, 1.18-1.23]) and octogenarians (AOR, 1.72 [95% CI, 1.67-1.78]) were significantly more likely to be prescribed thiazides than
The American Journal of Geriatric Pharmacotherapy
patients aged 66 to 69 years, even after adjusting for sex and other factors in the table. Multivariate analyses also demonstrated that longterm care residents and patients with more comorbidity were more likely to be prescribed thiazide or thiazidelike diuretics as their first antihypertensive agent (both, P < 0.001). As there was an interaction between sex and age, we confirmed that the association between older age and increased likelihood of being prescribed a thiazide as first-line treatment was present in both women and men. On the other hand, men and individuals with diabetes were less likely to be prescribed thiazides. There was no gradient in prescribing practices across socioeconomic status. DISCUSSION
Our data confirm the findings of 3 smaller cohort studies in other locations that thiazides are the initial agent of choice in less than half of all elderly hypertensives. However, while the Cardiovascular Health Study s reported a 37% decrease in thiazide prescribing between 1990 and 1999 in 4 centers in the United States, we found that thiazide prescriptions as initial monotherapy for elderly hypertensives in Canada increased by 10% between 1994 and 2002. Over one third of our patients were treated first with a thiazide, and our data demonstrate a clear age gradient in both men and women (with patients in their 70s and 80s more likely to be prescribed a thiazide as first-line t r e a t m e n t - echoing results from 1990-1999 prescribing data in X -O-0--~-D-
"o
Diuretics ACE inhibitors Calcium-channelblockers [3-Blockers Angiotensin II antagonists
'~D
o o_
1994
1996
1998
2000
2002
Year Figure. Annual rate of initial antihypertensive prescriptions for elderly hypertensive patients newly treated in Ontario, Canada, from 1994 through 2002.ACE = angiotensin-converting enzyme.
163
The American Journal of Geriatric Pharmacotherapy
IC Tu et al.
Table. Rates of thiazide use as first-line antihypertensive treatment in various subgroups and odds ratio (OR) estimates by logistic regression model, in 194,761 elderly patients with hypertension. OR Using Diuretics Patients Using Diuretics (n 68,858),
Crude Estimates
Adjusted Estimates
N
no. (%)
(95% CI)
(95% CI)
70,917 58,407 36,778 18,389 10,270
21,701 (30.6) 19,975 (34,2) 13,976 (38,0) 7999 (43,5) 5207 (50.7)
1.00 (Reference) 1,18 (I,15 1,21) 1,39 (i,36 1,43) 1,75 (1,69 1,80) 2.33 (2.24 2.43)
1.00 (Reference) 1,14 (I,II 1,17) 1,30 (i,26 1,34) 1,59 (1,54 1,65) 1.99 (I.90 2.08)
118,901 75,860
47,679 (40, I) 21,179 (27,9)
1,00 (Reference) 0,58 (0,5~0,59)
1,00 (Reference) 0,64 (0,63 0,65)
Diabetes No
164,537
64,476 (39,2)
Yes
30,224
Variable
Age group, y 66 69 70 74 7s 79 80 84 85 Sex Female Male
4382 (14,5)
1,00 (Reference)
1,00 (Reference)
0,26 (0,25 0,27)
0,22 (0,21 0,23)
Long term care resident No Yes
192,281 2480
67,597 (35,2) 1261 (50,8)
1,00 (Reference) 1,90 (I,76 2,06)
1,00 (Reference) 1,41 (I,28 1,56)
Charlson score C~I
185,052
65,044 (35, I)
1,00 (Reference)
1,00 (Reference)
6569 3140
2536 (38,6) 1278(40,7)
1,16(I,II 1,22) 1,26(I,18 1 , 3 6 )
1,24(I,16 1,29) 1,52(I,3~1,61)
1,00 0,98 0,98 1,00 1,06
1,00 0,99 1,00 1,03 1,06
2 _>3 Socioeconomic status I st Quintile 2nd Quintile 3rd Quintile 4th Quintile 5th Quintile Missing economic data
39,753 41,724 37,102 32,976 34,794 8412
14,033 14,520 12,91 I II ,608 12,735
(35,3) (34,8) (34,8) (35,2) (36,6)
(Reference) (0,95 1,01) (0,95 1,01) (0,9~1,03) (I,03 1,09)
(Reference) (0,96 1,02) (0,9~1,03) (I ,0C~1,06) (I,02 1,09)
3051 (36,3)
Note: While year of prescription and total number of drugs were included in the multivariate models for adjustment (OR reported in last column), they were not independently associated with thiazide prescriptions and thus are not featured in this table. Hosmer and Lemeshow Goodness of FitTest:Z 2 34.83 with 8 degrees of freedom, P < 0.001.
the Framingham H e a r t Study9). Finally, while residents of long-term care facilities and patients with multiple comorbidities were 40% to 50% more likely to be prescribed a thiazide as the first agent for their hypertension, men and diabetic hypertensive individuals were one third to three fourths less likely to have a thiazide chosen as their first antihypertensive agent. Our finding of decreased use of thiazide diuretics in men compared with w o m e n is similar to other studies, which raises the question of why it is so. 9,1°,18 20 The data on sex-related differences in adverse events and
164
effectiveness of treatment are inconclusive. Many trials do not report efficacy or adverse events by sex and when reported, those data are often inconsistent. For example, the Medical Research Council Trial of Mild Hypertension 21 reported that men had more adverse events with diuretics than with ~-blockers (the opposite pattern was seen in women). The Treatment of Mild Hypertension Study 22 reported that quality of life improved after diuretic treatment only in women. The Second Australian National Blood Pressure Study investigators 23 reported that diuretics were less efficacious
IC Tu et al.
than ACE inhibitors in older men (but were more efficacious than ACE inhibitors in older women). However, these sex-related differences in the efficacy and safety of thiazides were not observed in the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial. 24 Moreover, although some argue that thiazides might have special advantages in elderly women due to their association with reduced osteoporotic fractures,25 other authors have pointed out that women might be more at risk for thiazide-induced hyponatremia. 26 Regardless, hypertension guidelines available at the time of this study did not identify treatment regimens specifically related to sex. The existing literature related to the impact of socioeconomic status on choice and use of cardiovascular and preventive therapies is conflicting.27 30 This might be due to the variability of access to, and coverage for, health care that can confound the impact of socioeconomic status on treatment decisions. Therefore, we are not surprised that in Ontario, where all elderly individuals are covered by the province's drug benefit program, we did not observe any clinically significant impact of socioeconomic status on the type of antihypertensive prescribed. It might be reasonable to suppose that in other jurisdictions where patients have to pay a greater portion of drug costs out of pocket, a gradient in antihypertensive choices between income quintiles might be present.
Study Limitations While our study examined antihypertensive prescribing over the past decade in a large, representative, and population-based sample of all adults >66 years of age in Canada's largest province, there are a number of limitations inherent to the use of administrative databases. First, while data on prescriptions and cardiovascular comorbidities are easily and accurately identifiable, blood pressure readings and reasons why physicians chose 1 agent over another are not recorded in administrative records. For example, we do not know which patients had noncardiovascular diagnoses that may have impacted the choice of initial antihypertensive agent and are possibly undercoded in administrative databases (ie, gout, benign prostatic hypertrophy, renal stones, or a history of syndrome of inappropriate antidiuretic hormone secretion). Second, our cohort was limited to those patients who actually filled their prescription for an antihypertensive. However, as noncompliance with physician recommendations is unlikely to systematically differ between drug classes before the patient has taken the prescription to the pharmacy, it is reasonable to use our data to impute prescribing practices. Third, since
The American Journal of Geriatric Pharmacotherapy
not all hypertensive Canadians are prescribed antihypertensive drugs, our data cannot be used for estimating the incidence or prevalence of hypertension in this population. 31 However, the positive predictive value of using antihypertensive prescriptions in patients with only blood pressure-lowering indications for these agents to define a hypertensive cohort is -85% (based on data from the National Public Health and Community Health Surveys in Canada from 2000-2003) 32 and is an accepted method in hypertension pharmacotherapy research) 2 Finally, it should be acknowledged that any cohort study using administrative databases is essentially a cross-sectional study, and although we followed each patient in our cohort for 2 years after the index presentation into the cohort, we deliberately did not link with outcomes such as death, myocardial infarction, or stroke (as we felt residual confounding by indication would influence the relationship between prescribed medications and clinical outcomes). CONCLUSIONS
We found that long-term care residents, older patients, and those with multiple comorbidities were more likely to be prescribed thiazides as first-line agents of choice for initial antihypertensive treatment, whereas men and diabetic patients were less likely to be prescribed these agents. We also found that, in this study, socioeconomic status did not affect the treatment selection. Thiazides were chosen as the first-line agent in only 35% of our study population. ACKNOWLEDGMENTS
This research was supported by a Grant-in-Aid (#NA 5459) from the Heart and Stroke Foundation of Ontario. Dr. Tu is supported by a Canadian Institutes of Health Research (CIHR) Short-Term Clinician Investigator Award. Dr. McAlister is supported by an Alberta Heritage Foundation for Medical Research Population Health Scholar Award, a C I H R New Investigator Award, and the University of Alberta/Merck Frosst/Aventis Chair in Patient Health Management. None of the funding agencies had input into the design, conduct, analysis, or reporting of this study. Dr. McAlister reports operating grant funding support from Pfizer Canada (for a trial cofunded by the Heart and Stroke Foundation of Canada and unrelated to this manuscript or topic), and Dr. Campbell reports having received speaking fees and serving as a consultant/ advisory board member for sanofi-aventis, Bayer, BristolMyers Squibb, Pfizer, Servier, Bioval, Fournier, Abbott, Solvay, and Merck Frosst.
165
The American Journal of Geriatric Pharmacotherapy
K. Tu et al.
The authors thank Siraj Khan for his assistance in the preparation o f this manuscript. REFERENCES
1. Joft?es MR, Hamet P, MacLean DR, et al. Distribution of blood pressure and hypertension in Canada and the United States. A m J Hypertens. 2001 ;14:1099-1105. 2. Lewington S, Clarke R, Qizilbash N, et al, for the Prospective Studies Collaboration. Age-specific relevance of usual blood pressure to vascular mortality: A metaanalysis of individual data for one million adults in 61 prospective studies [published correction appears in Lancet. 2003;361:1060]. Lancet. 2002;360:1903-1913. 3. Staessen JA, Gasowski J, Wang JG, et al. Risks of untreated and treated isolated systolic hypertension in the elderly: Meta-analysis of outcome trials [published correction appears in Lancet. 2001;357:724]. Lancet. 2000;355:865-872. 4. Gueyffier F, Bulpitt C, Boissel JP, et al, for the INDANA Group. Antihypertensive drugs in very old people: A subgroup meta-analysis of randomised controlled trials. Lancet. 1999;353:793-796. 5. Wong ND, Thakral G, Franklin SS, et al. Preventing heart disease by controlling hypertension: Impact of hypertensive subtype, stage, age, and sex. A m Heart J. 2003;145:888-895. 6. Chobanian AV, Bakris GL, Black HR, et al. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. The JNC 7 Report. J A M A . 2003;289:25602572. 7. Canadian Hypertension Guidelines. 2005 Canadian recommendations for the management of hypertension. Available at: http://www.hypertension.ca/Documentation/ Recommandation05_va.pdf. Accessed May 22, 2006. 8. Psaty BM, Manolio TA, Smith NL, et al, for the Cardiovascular Health Study. Time trends in high blood pressure control and the use of antihypertensive medications in older adults: The Cardiovascular Health Study. Arch Intern Med. 2002;162:2325-2332. 9. Lloyd-Jones DM, Evans JC, Levy D. Hypertension in adults across the age spectrum: Current outcomes and control in the community. J A M A . 2005;294:4664:72. 10. Primatesta P, Poulter NR. Hypertension management and control among English adults aged 65 years and older in 2000 and 2001. JHypertens. 2004;22:1093-1098. 11. Khan NA, McAlister FA, Lewanczuk RZ, et al, for the Canadian Hypertension Education Program. The 2005 Canadian Hypertension Education Program recommendations for the management of hypertension: Part I I - therapy. Can J Cardiol. 2005;21:657-672.
166
12. Tu K, Campbell NR, Duong-Hua M, McAlister FA. Hypertension management in the elderly has improved: Ontario prescribing trends, 1994 to 2002. Hypertension. 2005;45:1113-1118. 13. International Classification of Diseases, Ninth Revision, ClinicalModification. 6th ed. Washington, DC: US Dept of Health and Human Services, Centers for Disease Control and Prevention, Centers for Medicare and Medicaid Services; 2005. 14. Williams JI, Young W. A summary of studies on the quality of health care administrative databases in Canada. In: Goel V, Williams JI, Anderson GM, et al, eds. Patterns of Health Care in Ontario. The ICES Practice Atlas. 2nd ed. Ottawa, Ontario, Canada: Canadian Medical Association; 1996:339-345. 15. Hux JE, Dis F, Flintoft V, Bica A. Diabetes in Ontario: Determination of prevalence and incidence using a validated administrative data algorithm. Diabetes Care. 2002; 25:512-516. 16. Wilkins R. PCCF + Version 4D User's Guide: Automated Geographic Coding Based on the Statistics Canada Postal Code Conversion Files, Including Postal Codes to December 2003. Ottawa, Ontario, Canada: Health Analysis and Measurement Group, Statistics Canada; July 2004. 17. Mustard CA, Derksen S, Berthelot JM, Wolfson M. Assessing ecologic proxies for household income: A comparison of household and neighbourhood level income measures in the study of population health status. Health Place. 1999;5:157-171. 18. Usher C, Bennett K, Feely J. Evidence for a gender and age inequality in the prescribing of preventative cardiovascular therapies to the elderly in primary care. Age Ageing. 2004;33:500-502. 19. Pears E, Hannaford PC, Taylor MW. Gender, age and deprivation differences in the primary care management of hypertension in Scotland: A cross-sectional database study. Faun Pract. 2003 ;20:22-31. 20. Onder G, Gambassi G, Landi F, et al, for the Investigators of the GIFA Study (SIGG-ONLUS). Trends in antihypertensive drugs in the elderly: The decline of thiazides. J H u m Hypertens. 2001;15:291-297. 21. Medical Research Council Working Party. MRC trial of treatment of mild hypertension: Principal results. BMJ (Clin Res Ed). 1985;291:97-104. 22. Neaton JD, Grimm RH Jr, Prineas RJ, et al, for the Treatment of Mild Hypertension Study Research Group. Treatment of Mild Hypertension Study. Final results. J A M A . 1993;270:713-724. 23. Wing LM, Reid CM, Ryan P, et al, for the Second Australian National Blood Pressure Study Group. A comparison of outcomes with angiotensin-converting-
K. Tu et al.
24.
25.
26. 27.
enzyme inhibitors and diuretics for hypertension in the elderly. N E n g l J M e d . 2003;348:583-592. ALLHAT Officers and Coordinators for the A L L H A T Collaborative Research Group. Major outcomes in highrisk hypertensive patients randomized to angiotensinconverting enzyme inhibitor or calcium channel blocker vs diuretic: The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) [published corrections appear in J A M A . 2003;289:178 and J A M A . 2004;291:2196]. J A M A . 2002;288:29812997. Schoofs MW, van der Klift M, Hofman A, et al. Thiazide diuretics and the risk for hip fracture. A n n Intern Med. 2003;139:476-482. Chow KM, Szeto CC, Wong TY, et al. Risk factors for thiazide-induced hyponatraemia. Q,JM. 2003;96:911-917. Ahmed A, Allman RM, DeLong IF, et al. Age-related underutilization of angiotensin-converting enzyme inhibitors in older hospitalized heart failure patients. South Med J. 2002;95:703-710.
The American Journal of Geriatric Pharmacotherapy
28. Solberg LI, Brekke ML, Kottke TE. Are physicians less likely to recommend preventive services to low-SES patients? Prev Med. 1997;26:350-357. 29. Scott A, SNell A, King M. Is general practitioner decision making associated with patient socio-economic status? Soc Sci Med. 1996;42:35-46. 30. McAlister FA, Murphy NF, Simpson CR, et al. Influence of socioeconomic deprivation on the primary care burden and treatment of patients with a diagnosis of heart failure in general practice in Scotland: Population based study. BMJ. 2004;328:1110. 31. Joff?es MR, Ghadirian P, Fodor JG, et al. Awareness, treatment, and control of hypertension in Canada. A m J Hypertens. 1997;10:1097-1102. 32. Campbell NR, Onysko J, Maxwell C, et al, for the Canadian Hypertension Education Program. Increases in the diagnosis and drug treatment of hypertensive Canadians [abstract]. Presented at: 58th Annual Meeting of the Canadian Cardiovascular Society, October 22-26, 2005; Montreal, Quebec, Canada.
Address c o r r e s p o n d e n c e to: Finlay A. McMistcr, MD, 2E3.24 Walter Mackcnzic Ccntrc, University of Alberta Hospital, 8440 112 Street, Edmonton, Mbcrta T6G 2R7, Canada. E-mail: Finlay'McMistcr@ualbcrta'ca
167