- Email: [email protected]
Drug prescriptions after acute myocardial infarction: Dosage, compliance, and persistence
Drug prescriptions after acute myocardial infarction: Dosage, compliance, and persistence Ewurabena Simpson, BSc,a Christine Beck, MSc,a Hugues Richard, MSc,a Mark J. Eisenberg, MD, MPH,b and Louise Pilote, MD, MPH, PhDa Montreal, Quebec, Canada
Background Although it has been well documented that aspirin, -blockers, angiotensin-converting enzyme (ACE) inhibitors, and lipid-lowering drugs are under-prescribed for patients with acute myocardial infarction (Am Heart J 2003; 145:438-44.) , few studies have examined dosage and long-term compliance and persistence patterns for the use of these drugs after AMI.
Methods Using Quebec administrative data on all elderly (aged ⱖ65 years) survivors of hospital admissions for AMI between 1996 and 1998 (n ⫽ 14,057), we studied the discharge prescriptions, dosages, patient compliance, and persistence during this period for aspirin, -blockers, ACE inhibitors, and lipid-lowering drugs.
Results Rates of discharge medications were suboptimal (aspirin 65%, -blockers 54%, ACE inhibitors 45%, lipidlowering drugs 21%). Most patients with prescriptions for aspirin and ACE inhibitors were prescribed dosages equivalent to those administered in clinical trials (99% and 88%, respectively). In contrast, only 20% of patients with -blocker prescriptions and 48% of patients with lipid-lowering drug prescriptions were prescribed clinical trial doses. For patients with discharge prescriptions, 1-year compliance rates were high (aspirin 74%, -blockers 74%, ACE inhibitors 70%, lipid-lowering drugs 84%), as were the 1-year persistence rates (aspirin 71%, -blockers 72%, ACE inhibitors 69%, lipid-lowering drugs 80%).
Conclusion Although cardiac drugs are under-prescribed to patients with AMI, once prescribed, patients are likely to adhere to these prescriptions, with high rates of compliance and persistence. (Am Heart J 2003;145:438-44.) The survival benefits of aspirin (acetylsalicylic acid), -adrenergic receptor blocker medications (-blockers), angiotensin-converting enzyme inhibitors (ACE inhibitors), and lipid-lowering drugs have been well documented in several clinical trials.1-4 However, it has been reported that these medications are under-prescribed for patients with acute myocardial infarction (AMI).5-12 Although several studies have described low prescription rates for these medications, there are few data evaluating long-term compliance and persistence patterns among patients with AMI who have received these prescriptions outside of the context of a clinical trial. The effectiveness of any pharmacologic treatment is directly related to the physician’s choice of treatment and patient compliance with the given medication reg-
imen. After a patient has sustained an AMI, failure to comply with even some aspects of a recommended regimen can lead to preventable complications, rehospitalizations, or mortality. Previous research has shown that most cardiac-related morbidity and mortality occurs among elderly patients and that this population is at a higher risk for low compliance in taking cardiac medications.10 It is therefore crucial to characterize and understand the patterns of use of these drugs within this patient population. For this reason, we examined the prescription patterns for aspirin, -blockers, ACE inhibitors, and lipid-lowering drugs in the year after hospital discharge among patients aged ⱖ65 years who were admitted to Quebec hospitals for their first AMI between 1996 and 1998.
Materials and methods a
From the Division of Clinical Epidemiology, Montreal General Hospital, Montreal, Quebec, and the bDivisions of Cardiology and Clinical Epidemiology, Jewish General Hospital, McGill University, Montreal, Quebec, Canada. Dr Pilote is an investigator for the Canadian Institutes for Health Research (CIHR). Dr Eisenberg is an investigator for the Fonds Recherche en Sante´ du Que´bec. Supported in part by a grant from CIHR (#53181). Submitted February 27, 2002; accepted June 4, 2002. Reprint requests: Louise Pilote, MD, MPH, PhD, Division of Clinical Epidemiology, Montreal General Hospital, 1650 Cedar Ave, Montreal, Quebec, H3G 1A4, Canada. E-mail: [email protected] Copyright 2003, Mosby, Inc. All rights reserved. 0002-8703/2003/$30.00 ⫹ 0 doi:10.1067/mhj.2003.143
Study population A cohort of elderly (aged ⱖ65 years) patients who sustained an AMI was constructed by use of the administrative hospital discharge summary database for the province of Quebec (Med-Echo). The Med-Echo database contains records on hospital admissions for all citizens and legal residents in Quebec. We identified all elderly patients who were admitted with a main discharge diagnosis of AMI (International Classification of Diseases, 9th revision code 410) between Jan 1, 1996, and Dec 31, 1998, and who were discharged alive (n ⫽ 16,988). To increase comparability of patients, we then
American Heart Journal Volume 145, Number 3
Simpson et al 439
Table I. Characteristics of elderly (ⱖ65 years) patients who filled and did not fill outpatient prescriptions for cardiac medications after acute myocardial infarction in Quebec, 1996-1998 Characteristic Number of subjects* (%) Male (%) Age (median years) Length of stay (median days) Comorbidities (% of patients) Congestive heart failure Cerebrovascular disease Pulmonary edema Diabetes mellitus Malignant cancer Acute renal failure Shock Cardiac dysrhythmias Physician specialties (% patients) General Cardiologist Internist Other Inhospital procedures (% patients) Angiography PTCA CABG
No prescription
Aspirin
ACE inhibitors
-Blockers
Lipid-lowering drugs
1463 (10.4) 55.4 75 10
9134 (65.0) 56.7 74 9
6287 (44.7) 54.6 75 10
7595 (54.0) 57.0 73 9
2970 (21.1) 57.8 71 9
29.3 10.7 2.9 24.3 3.6 8.1 2.1 27.9
23.6 6.4 1.9 22.2 2.6 4.2 1.0 17.8
39.4 7.9 3.0 27.5 2.7 5.4 1.5 23.7
19.4 6.8 1.5 21.0 2.4 3.7 0.8 16.6
21.2 7.5 1.3 24.0 2.0 4.1 0.9 18.3
46.1 43.2 8.6 2.1
44.1 44.9 10.0 1.0
43.2 45.4 10.0 1.4
42.7 47.1 9.3 0.9
39.0 48.3 11.7 1.0
21.3 7.1 7.3
27.3 12.1 4.9
23.1 9.8 3.3
29.7 13.3 4.1
37.7 16.2 6.7
No prescription, Patients without a prescription filled for any of the four cardiac drug classes within 30 days of discharge; ACE, angiotensin-converting enzyme; PTCA, percutaneous transluminal coronary angioplasty; CABG, coronary artery bypass graft surgery. *The total number of study patients was 14,057.
included only patients with a first episode of AMI by ascertaining the absence of a hospital admission for AMI since January 1, 1988 (n ⫽ 1760). Next, using encrypted Medicare numbers, we linked the discharge summary database with the physician and drug claims database for the province of Quebec (RAMQ). This database contains information on all outpatient prescriptions for patients aged ⱖ65 years and on all inpatient and outpatient diagnostic and therapeutic procedures in Quebec. A number of exclusions were then made sequentially when 1) the main diagnosis of AMI was given after transfer from another hospitalization (n ⫽ 220), 2) there were incomplete data related to a change in Medicare number (n ⫽ 136), 3) the total length of stay was ⬍3 days (because this may indicate an unclear AMI diagnosis [n ⫽ 247]), or 4) patients were discharged to a long-term care institution, a rehabilitation center, or another province (because their prescriptions could not be ascertained [n ⫽ 568]). After these exclusions were made, our study population consisted of 14,057 patients, 82.7% of the original cohort.
Patient characteristics We grouped the study patients according to whether they had ⱖ1 class of cardiac medications prescribed at the time of hospital discharge. Patients who had obtained a prescription of at least one of the 4 major classes of cardiac medications (aspirin, -blockers, ACE inhibitors, lipid-lowering drugs) within 30 days of hospital discharge were classified as having an AMI drug prescription at discharge. It was thus possible for patients to be grouped into ⬎1 group when they had prescriptions from ⬎1
cardiac drug class. We then compared crude demographic and clinical characteristics, rates of cardiac procedure use, and specialties of the treating physicians for each group (Table I).
Time trends, compliance, and persistence Overall and annual rates of discharge prescriptions were assessed from 1996 to 1998. For each patient with a discharge prescription, we calculated their daily dosages on the basis of their prescriptions within one year after discharge. For each prescription, dosage results were grouped as either “clinical trial dosages” or “below clinical trial dosages.” A clinical trial dosage was defined as any dosage greater than or equal to the evaluated dosages administered in clinical trials, whereas a below clinical trial dosage was defined as any dosage that was less than the range of evaluated dosages in randomized clinical trials (Appendix, Table II).13-31 For example, we defined the clinical trial dosage for captopril as ⱖ12.5 mg/d, which was the lowest dosage administered in the Fourth International Study of Infarct Survival (ISIS-4) trial.14 When available, we selected only trials that evaluated secondary prevention after AMI. When clinical trials were not available, we defined “clinical trial dosages” according to the recommended starting dosages for patients with AMI in pharmaceutical guidelines. After we identified each prescription dosage as a clinical trial dosage or below clinical trial dosage, we placed patients in a dosage category for each of the 4 cardiac drug classes. To evaluate possible up- or down-titration of dosages, we also performed a number of additional analyses. First, we measured the dosages for each of the first 11 prescriptions
American Heart Journal March 2003
440 Simpson et al
Table II. Dosages of cardiac prescriptions, according to drug class Drug class Below clinical trial dosage (%) Discharge prescription Prescription at 1 year No dose switching at 1 year (%) Below clinical trial dosage* Above clinical trial dosage* Dose switching at 1 year (%) Below clinical trial dosage* Above clinical trial dosage* Dosage persistence at 1 year (%) Below clinical trial dosage* Above clinical trial dosage*
Aspirin
-Blockers
ACE inhibitors
Lipid-lowering drugs
0.5 0.7
80.4 79.0
11.7 11.1
52.4 44.7
0.3 99.0
73.1 14.7
7.6 83.3
40.0 42.4
0.2 0.5
7.4 4.9
4.1 5.0
12.4 5.3
78.4 99.9
95.5 85.6
78.3 97.6
87.7 95.1
Clinical trial dosages were defined as dosages for discharge prescriptions that were greater than or equal to the evaluated dosages administered in clinical trials. Below clinical trial dosages were defined as dosages for discharge prescriptions that were less than the range of evaluated dosages in clinical trials (Appendix). Dosage persistence was defined as the proportion of prescriptions filled that were of the same dosage category as discharge prescription dose, among all the prescriptions filled within the 1 year after discharge. *Dosage category at discharge.
after the discharge prescription and for the last prescription filled within one year after discharge. Second, we examined the proportions of patients who received discharge prescriptions that switched or did not switch dosage categories at any time during the year after discharge. Third, we calculated patients’ persistence on the dosage prescribed at discharge. Dosage persistence was defined as the proportion of prescriptions filled that were of the same dosage category as the discharge prescription dosage category, among all the prescriptions filled within the year after discharge. Compliance was calculated to determine the proportion of patients who continued to fill their prescriptions for each of the drug classes at one year after discharge. Compliance was calculated only for patients who were alive at one year and who received a discharge prescription. These patients were classified as “compliant” at one year when they had filled at least one prescription within 305 to 365 days after discharge. Similarly to earlier studies of persistence, we defined high persistence as the proportion of patients who received a discharge prescription and whose prescriptions covered at least 80% of the days within the year after discharge.10 For each patient, persistence was measured by means of the actual duration of medication use (in days) divided by the maximum possible duration of medication use (in days). For patients who died within one year after discharge, persistence was calculated as the actual duration of medication use divided by the total duration of medication use until the date of death. This analysis was performed for each patient according to drug class. All analyses were performed by use of the STATA (version 5, College Station, Tex) statistical software package.
Results Patient characteristics Overall prescription rates for aspirin, -blockers, ACE inhibitors, and lipid-lowering medications were subopti-
mal at discharge (65%, 54%, 45%, and 21%, respectively), and 10% of patients received no discharge prescriptions. The proportions of patients receiving prescriptions for any combination of 1, 2, or 3 of the 4 drugs were 24%, 40%, and 22%, respectively. Only 4% of patients received discharge prescriptions for all 4 drugs. Clinical and demographic characteristics were similar across the 5 prescription groups (Table I). However, there were more patients with ACE inhibitor prescriptions and congestive heart failure than patients with congestive heart failure in the remaining prescription groups (39.4% vs 19.4%, 23.6%, and 21.2% for -blockers, aspirin, and lipid-lowering drugs, respectively). Patients without prescriptions for any of the 4 drug classes had higher rates of comorbidities than patients with prescriptions. For the 4 drug classes, the group with prescriptions for lipid-lowering medications had the highest proportion of patients who underwent angiography, percutaneous transluminal coronary angioplasty, or coronary artery bypass grafting (37.7%, 16.2%, and 6.7%, respectively). A greater proportion of patients who had prescriptions were treated by cardiologists or internists on hospitalization for AMI than patients who did not receive any prescriptions. Similarly, the group without prescriptions and the ACE inhibitor drug group had the lowest proportion of patients who underwent angiography or percutaneous transluminal coronary angioplasty.
Time trends, compliance, and persistence There were temporal changes in the rates of cardiac prescriptions between 1996 and 1998 (Figure 1). During the study period, there was a marked increase in the proportion of patients with lipid-lowering drug prescriptions at discharge (11.3% difference). In contrast, there was a
American Heart Journal Volume 145, Number 3
Simpson et al 441
Figure 1
Time trends in discharge cardiac prescriptions for elderly patients with acute myocardial infarction. For each year, patients who were prescribed at least one of the 4 major classes of cardiac medications within 30 days of hospital discharge were classified as having an acute myocardial infarction prescription at discharge. Number of patients: 1996, n ⫽ 4732; 1997, n ⫽ 4711; 1998, n ⫽ 4614. -blockers, -adrenergic receptor blockers; ACE, angiotensin-converting enzyme.
decrease in the proportion of patients who received aspirin prescriptions after discharge (7.8% difference). Finally, there were only modest increases in rates of -blocker and ACE inhibitor prescriptions (1.7% and 2.7% differences, respectively). For aspirin, -blockers, and ACE inhibitors, most patients received an initial prescription immediately after hospital discharge (median time to initial prescription, 0 days for aspirin, -blockers, and ACE inhibitors). However, it took longer for patients to receive prescriptions for lipid-lowering medications (median time to initial prescription, 20 days). When the rates for discharge prescriptions were compared with rates for cardiac prescriptions within one year after discharge, there was no substantial increase in the rates of aspirin, -blocker, and ACE inhibitor prescriptions (prescription rates at 1 year: aspirin 74.9%, -blockers 60.4%, ACE inhibitors 53.2%). The rate of lipid-lowering drug prescriptions, however, was higher at one year than at the time of discharge (prescription rate at 1 year 36.4%). Similarly, of the patients who did not receive discharge prescriptions but who were alive at one year, only a small number were “late starters” and filled a prescription within 11 to 12 months after discharge for -blockers, ACE inhibitors, and aspirin (4.7%, 6.4%, and 7.9%, respectively). However, greater proportions of patients had late prescriptions for lipidlowering drugs filled (14.1%).
Whether patients with cardiac prescriptions received initial dosages that were within the range of dosages administered in clinical trials depended on the drug class (Table II). Of patients to whom aspirin and ACE inhibitors were prescribed, ⬎85% received clinical trial dosages for these drug classes (99.4% and 88.3%, respectively). In contrast, initial prescription dosages for -blockers and lipid-lowering drugs were often lower than those used in randomized controlled trials. Of the patients who received discharge prescriptions for -blockers and lipid-lowering drugs, only 19.6% and 47.6%, respectively, received prescriptions that were within clinical trial dosages. When the prescriptions at one year were assessed, there were only modest increases in the proportions of patients receiving clinical trial dosage prescriptions for aspirin, ACE inhibitors, and -blockers compared with initial prescriptions. However, there were larger increases for patients with lipid-lowering drugs (Table II). The same trends were evident when looking at each of the first 11 prescriptions after the initial discharge prescription (data not shown). The high dosage persistence rates provide additional evidence that patients tended to stay within the same dosage category in the year after discharge (Table II). When receiving a prescription for AMI, patients showed high rates of compliance and persistence for these drugs (Table III). The one-year compliance rates
American Heart Journal March 2003
442 Simpson et al
Table III. Compliance and persistence with cardiac prescriptions at one year, according to drug class
Drug class Compliance (%) High persistence (%)
LipidACE lowering Aspirin -Blockers inhibitors drugs 74.4 71.4
73.7 71.6
69.6 69.4
84.0 79.7
Compliance was defined as the proportion of patients who filled a discharge prescription and who filled at least 1 prescription within 305 to 365 days after discharge. High persistence was defined as the proportion of patients who filled a discharge prescription and whose prescriptions covered at least 80% of the days within the year after discharge.
were high for each of the 4 drug categories. The highest rate of compliance was associated with the use of lipid-lowering medications (84.0%, 95% CI 82.6%85.4%), followed by aspirin (74.4%, 95% CI 73.5%75.4%), -blockers (73.7%, 95% CI 72.6%-74.7%), and ACE inhibitors (69.6%, 95% CI 68.3%-70.8%). Measurements of patient persistence showed that there were high proportions of patients who maintained their drug prescriptions for at least 80% of the one-year study period (high persistence) (Table III). A total of 79.7% of patients (95% CI 78.2%-81.2%) who had lipid-lowering drug prescriptions filled at discharge had high persistence rates, the highest rate for any of the 4 drug classes. For the remaining drug categories, 71.6% of patients (95% CI 70.6%-72.6%) with -blocker prescriptions, 71.4% of patients (95% CI 70.5%-72.3%) with aspirin prescriptions, and 69.4% of patients (95% CI 68.2%-70.5%) with prescriptions for ACE inhibitors had high persistence rates.
Discussion In this study, we describe the trends for post-AMI drug use in Quebec in the 3 years from 1996 to 1998. Our results suggest that recommended cardiac drugs are prescribed at suboptimal rates and dosages for patients with AMI. However, once prescribed, patients showed high rates of compliance and persistence for these medications in the year after discharge. To date, there are limited data about patient compliance and persistence for cardiac drug use after AMI. Unlike in earlier studies of persistence of lipid-lowering medication use after AMI,10 in our study patients showed high persistence for lipid-lowering drugs and for aspirin, -blockers, and ACE inhibitors. Their high levels of persistence suggest that, to maximize the benefits of cardiac medications, physicians should ensure that these drugs are prescribed to all eligible patients at discharge. However, although the rates of patient compliance were high, there remained approximately 25% of patients to whom aspirin, -blockers, or ACE inhibitors were prescribed who did not continue to renew these prescrip-
tions at one year after discharge. These findings suggest that new strategies must be found to achieve higher compliance rates. Even with evidence of benefit, the annual prescription rates for aspirin, -blockers, and ACE inhibitors did not change substantially with time. However, there was a marked increase in the rate of lipid-lowering drug prescriptions during the study period. This increase might be explained by the advent of new evidence demonstrating the efficacy of lipid-lowering treatment within 3 to 6 months of hospital discharge after AMI and by the availability of novel lipid-lowering medications.32 However, room for improvement remains. On the basis of the general lack of contraindications to lipid-lowering medications, especially statins, most patients who survive an AMI should have lipid-lowering medications prescribed at discharge.33 There was also a modest decline in aspirin prescription rates. This change may be related to the drug co-payment policies initiated by the Quebec government in 1996, which could have increased the likelihood that patients would obtain over-the-counter aspirin preparations rather than prescribed aspirin. Of the 4 drug classes, -blockers were the least often prescribed at the same dosages as those administered in clinical trials. These results are similar to those found in an earlier observational study, in which almost 50% of patients receiving prescriptions for -blockers were treated with dosages that were less than half the dosages administered in randomized clinical trials.34 In addition, the administered dosages for aspirin, ACE inhibitors, -blockers, and—to a lesser extent—lipid-lowering medications did not change in the year after discharge, suggesting that physicians did not increase dosages on the basis of tolerance. However, whether lower dosages of aspirin, -blockers, ACE inhibitors, or lipid-lowering drugs would be equally effective at reducing post-AMI mortality rates has yet to be determined. It is also possible that elderly patients require lower than trial dosages of medications, in comparison with younger patients. The assessment of patient characteristics indicated that the decision to prescribe certain cardiac drug classes instead of others did not depend solely on disease severity. Other unexplained factors likely play a role in the suboptimal use of cardiac prescriptions after AMI. Nonetheless, comparisons of the clinical characteristics for patients with no cardiac prescriptions versus patients with at least one prescription indicated that the physician choice to either prescribe or forgo cardiac medications altogether was influenced by the presence of associated comorbidities. It is possible that the higher rates of comorbidities among patients without prescriptions may have led their physicians to forgo the use of preventive therapies in view of a perceived poor prognosis.
American Heart Journal Volume 145, Number 3
Our study was limited because, due to the nature of the drugs claims database, we did not have sufficient inhospital data pertaining to clinical characteristics of the study patients. As a result, we were unable to accurately exclude patients with contraindications from our analyses. Consequently, we may have underestimated the prescription rates for the 4 drug classes studied. However, because there is no single set of criteria that define contraindications to each of the studied drug classes, contraindication assessment tends to be very subjective. As a result, certain contraindications may be completely preclusive to drug therapy (“absolute contraindications”), whereas other contraindications may be deemed as permissive to therapy with certain conditions (“relative contraindications”). The looseness of this definition made it very difficult to estimate accurately patient contraindications to cardiac drug treatment, especially from an administrative database. Additionally, we could not identify the reasons for suboptimal prescription practices from our administrative databases. Finally, we cannot determine from our databases whether all the patients who continued to renew their cardiac prescriptions were actually taking their medication as it was prescribed. However, the assessment of prescription refill patterns does provide an accurate and practical means of approximating medication use in large study populations.10 Despite the limitations of using administrative databases, previous studies have clearly demonstrated that these data provide an accurate standardized tool for assessing patterns of medication use in large patient populations.35-38 Specifically for the Quebec hospital summary and prescription claims databases, recent reports have demonstrated that their predictive value in identifying patients with AMI is high and that this information gives an accurate summary of prescriptions dispensed to patients aged ⱖ65 years.39,40 The results of this study suggest that, if physicians prescribe medications according to evidence-based guidelines, patients are likely to adhere to these prescriptions in the long term. Thus, although we are not certain of all factors that are responsible for long-term prescription patterns after AMI, these results suggest that the suboptimal adherence to guidelines evident in Quebec from 1996 to 1998 arose primarily because of low rates of discharge prescriptions on the part of physicians.
References 1. Jamali AH, Wilson Tang WH, Khot UN, et al. The role of angiotensin receptor blockers in the management if chronic heart failure. Arch Int Med 2001;161:667-72. 2. Collaborative overview of randomized trials of antiplatelet therapy—I: prevention of death, myocardial infarction, and stroke by prolonged antiplatelet therapy in various categories of patients. Antiplatelet Trials Collaboration. BMJ 1994;308:81-106. 3. Gottlieb SS, McCarter RJ, Vogel RA. Effect of beta-blockade on mortality among high-risk and low-risk patients after myocardial infarction. New Engl J Med 1998;339:489-97.
Simpson et al 443
4. Crouch MA. Effective use of statins to prevent coronary heart disease. Am Fam Physician 2001;63:309-20. 5. Mendelson G, Aronow WS. Underutilization of angiotensin-converting enzyme inhibitors in older patients with Q-wave anterior myocardial infarction in an academic hospital-based geriatrics practice. Am Geriatr Soc 1998;46:751-2. 6. Heller DA, Ahern FM, Kozak M. Changes in rates of -blocker use between 1994 and 1997 among elderly survivors of acute myocardial infarction. Am Heart J 2000;140:663-71. 7. Krumholz HM, Radford MJ, Ellerbeck EF, et al. Aspirin for secondary prevention after acute myocardial infarction in the elderly: prescribed use and outcomes. Ann Intern Med 1996;124:292-8. 8. Bown Luzier A, Navsarikar A, Wilson MF, et al. Patterns of prescribing ACE inhibitors after myocardial infarction. Pharmacotherapy 1999;19:655-60. 9. Rochon PA, Anderson GM, Tu JV, et al. Use of -blocker therapy in older patients after acute myocardial infarction in Ontario. CMAJ 1999;161:1403-8. 10. Avorn J, Monette J, Lacour A, et al. Persistence of use of lipidlowering medications. JAMA 1998;279:1458-62. 11. Beck CA, Lauzon C, Eisenberg MJ, et al. Discharge prescriptions following admission for acute myocardial infarction at tertiary care and community hospitals in Quebec. Can J Cardiol 2001;17:33-40. 12. Krumholz HM, Radford MJ, Wang Y, et al. National use and effectiveness of -blockers for the treatment if elderly patients after acute myocardial infarction. National Cooperative Cardiovascular Project. JAMA 1998;280:623-9. 13. Canadian Pharmacists Association Group. Bruce LD, editor. Compendium of pharmaceuticals and specialties. 33rd ed. Ottawa: Canadian Pharmacists Association; 1998. 14. ISIS-4 (Fourth International Study of Infarct Survival) Collaborative Group. ISIS-4: a randomised factorial trial assessing early oral captopril, oral mononitrate, and intravenous magnesium sulphate in 58,050 patients with suspected acute myocardial infarction. Lancet 1995;345:669-85. 15. Swedberg K, Held P, Kjekshus J, et al. Effects of the early administration of enalapril on mortality in patients with acute myocardial infarction: results of the Cooperative New Scandinavian Enalapril Survival Study II (CONSENSUS II). New Engl J Med 1992;327:678-84. 16. The SOLVD Investigators. Effect of enalapril on survival in patients with reduced left ventricular ejection fractions and congestive heart failure. New Engl J Med 1991;325:293-302. 17. Borghi C, Marino P, Zardini P, et al. Short- and long-term effects of early fosinopril administration in patients with acute anterior myocardial infarction undergoing intravenous thrombolysis: results from the Fosinopril in Acute Myocardial Infarction Study. FAMIS Working Party. Am Heart J 1998;136:213-25. 18. Gruppo Italiano per lo Studio della Sopravvivenza nell’infarto Miocardico. GISSI-3: effects of lisinopril and transdermal glyceryl trinitrate singly and together on 6-week mortality and ventricular function after acute myocardial infarction. Lancet 1994;343:1115-22. 19. Anonymous. PREAMI: Perindopril and Remodelling in Elderly with Acute Myocardial Infarction: study rationale and design. Cardiovasc Drugs Ther 2000;14:671-9. 20. The Acute Infarction Ramipril Efficacy (AIRE) Study Investigators. Effect of ramipril on mortality and morbidity of survivors of acute myocardial infarction with clinical evidence of heart failure. Lancet 1993;342:821-8. 21. Neil HA, Fowler G, Patel H, et al. An assessment of the efficacy of atorvastatin in achieving LDL cholesterol target levels in patients
American Heart Journal March 2003
444 Simpson et al
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36. 37.
with coronary heart disease: a general practice study. Int J Clin Pract 1999;53:509-13. The BIP Study Group. Secondary prevention by raising HDL cholesterol and reducing triglycerides in patients with coronary artery disease: the Bezafibrate Infarction Prevention (BIP) study. Circulation 2000;102:21-7. Riegger G, Abletshauser C, Ludwig M, et al. The effect of fluvastatin on cardiac events in patients with symptomatic coronary artery disease during one year of treatment. Artherosclerosis 1999;144:263-7. Downs JR, Clearfield M, Weis S, et al. Primary prevention of acute coronary events with lovastatin in men and women with average cholesterol levels: results of AFCAPS/TexCAPS. JAMA 1998;279: 1615-22. The Long-Term Intervention with Pravastatin in Ischemic Disease (LIPID) Study Group. Prevention of cardiovascular events and death with pravastatin in patients with coronary heart disease and a broad range of initial cholesterol levels. New Engl J Med 1998; 339:1349-57. Scandinavian Simvastatin Survival Study Group. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet 1994;344:1383-9. American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on Management of Acute Myocardial Infarction). 1999 update: ACC/AHA guidelines for the management of patients with acute myocardial infarction: executive summary and recommendations. Circulation 1999;100:1016-30. Wilcox RG, Roland JM, Banks DC, et al. Randomised trial comparing propranolol with atenolol in immediate treatment of suspected myocardial infarction. BMJ 1980;280:885-8. Hjalmarson A, Herlitz J, Holmberg S, et al. The Goteborg metoprolol trial: effects on mortality and morbidity in acute myocardial infarction. Circulation 1983;67(1 Suppl):I26-32. -Blocker Heart Attack Trial Research Group. A randomized trial of propranolol in patients with acute myocardial infarction. JAMA 1982;247:1707-14. Norwegian Multicenter Study Group. Timolol-induced reduction in morbidity and reinfarction in patients surviving acute myocardial infarction. New Engl J Med 1981;304:801-7. Jackevicius CA, Anderson GM, Leiter L, et al. Use of the statins in patients after acute myocardial infarction: does evidence change practice? Arch Intern Med 2001;161:183-8. Expert Panel on Detection Evaluation, and Treatment of High Blood Cholesterol in Adults. Summary of the second report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel II). JAMA 1993;269:3015-23. Barron HV, Viskin S, Lundstrom RJ, et al. -blocker dosages and mortality after acute myocardial infarction. Arch Int Med 1998; 158:449-53. Bright RA, Avorn J, Everitt DE. Medicaid data as a resource for epidemiologic studies: strengths and limitations. J Clin Epidemiol 1989;42:937-45. Ray WA, Griffin MR. Use of Medicaid data for pharmacoepidemiology. Am J Epidemiol 1989;129:837-49. Romano PS, Roos LL, Luft HS, et al. A comparison of administrative versus clinical data: coronary artery bypass surgery as an ex-
ample. Ischaemic Heart Disease Patient Outcomes Research Team. J Clin Epidemiol 1994;47:249-60. 38. D’Hoore W, Bouckaert A, Tilquin C. Practical considerations on the use of the Charlson comorbidity index. J Clin Epidemiol 1996; 49:1429-33. 39. Tamblyn R, Lavoie G, Petrella L, et al. The use of prescription claims databases in pharmacoepidemiological research: the accuracy and comprehensiveness of the prescriptions claims database in Que´ bec. J Clin Epidemiol 1995;48:999-1009. 40. Levy AR, Tamblyn R, Fitchett D, et al. Coding accuracy of hospital discharge data for elderly survivors of myocardial infarction. Can J Cardiol 1999;15:1277-82.
Appendix Classification of clinical trial dosage therapies for all acute myocardial infarction medications from the Quebec physicians and drugs claims database (RAMQ). Classification of clinical trial dosage therapies for all AMI medications from the Quebec physicians and drugs claims database (RAMQ) Drug class ACE inhibitors Benazepril13 Captopril14 Cilazapril13 Enalapril15,16 Fosinopril17 Lisinopril18 Perindopril19 Quinapril13 Ramipril20 Lipid-lowering drugs Atorvastatin21 Bezafibrate22 Cerivastatin13 Cholestyramine13 Clofibrate13 Colestipol (suspension)13 Dextrothyroxine sodium13 Fenofibrate13 Fluvastatin23 Gemfibrozil13 Lovastatin24 Nicotinic acid13 Pravastatin25 Simvastatin26 Aspirin27 -Blockers Acebutolol13 Atenolol28 Metoprolol29 Nadolol13 Oxprenolol13 Pindolol13 Propranolol30 Timolol31
Clinical trial dosage (mg/day)
ⱖ10 ⱖ12.5 ⱖ1.25 ⱖ5 ⱖ5 ⱖ5 ⱖ2 ⱖ10 ⱖ5 ⱖ10 ⱖ400 ⱖ0.4 ⱖ4000 ⱖ2000 ⱖ2000 ⱖ1 ⱖ300 ⱖ40 ⱖ1200 ⱖ20 ⱖ1500 ⱖ40 ⱖ20 ⱖ160 ⱖ200 ⱖ100 ⱖ200 ⱖ80 ⱖ60 ⱖ10 ⱖ20 ⱖ10
Background Although it has been well documented that aspirin, -blockers, angiotensin-converting enzyme (ACE) inhibitors, and lipid-lowering drugs are under-prescribed for patients with acute myocardial infarction (Am Heart J 2003; 145:438-44.) , few studies have examined dosage and long-term compliance and persistence patterns for the use of these drugs after AMI.
Methods Using Quebec administrative data on all elderly (aged ⱖ65 years) survivors of hospital admissions for AMI between 1996 and 1998 (n ⫽ 14,057), we studied the discharge prescriptions, dosages, patient compliance, and persistence during this period for aspirin, -blockers, ACE inhibitors, and lipid-lowering drugs.
Results Rates of discharge medications were suboptimal (aspirin 65%, -blockers 54%, ACE inhibitors 45%, lipidlowering drugs 21%). Most patients with prescriptions for aspirin and ACE inhibitors were prescribed dosages equivalent to those administered in clinical trials (99% and 88%, respectively). In contrast, only 20% of patients with -blocker prescriptions and 48% of patients with lipid-lowering drug prescriptions were prescribed clinical trial doses. For patients with discharge prescriptions, 1-year compliance rates were high (aspirin 74%, -blockers 74%, ACE inhibitors 70%, lipid-lowering drugs 84%), as were the 1-year persistence rates (aspirin 71%, -blockers 72%, ACE inhibitors 69%, lipid-lowering drugs 80%).
Conclusion Although cardiac drugs are under-prescribed to patients with AMI, once prescribed, patients are likely to adhere to these prescriptions, with high rates of compliance and persistence. (Am Heart J 2003;145:438-44.) The survival benefits of aspirin (acetylsalicylic acid), -adrenergic receptor blocker medications (-blockers), angiotensin-converting enzyme inhibitors (ACE inhibitors), and lipid-lowering drugs have been well documented in several clinical trials.1-4 However, it has been reported that these medications are under-prescribed for patients with acute myocardial infarction (AMI).5-12 Although several studies have described low prescription rates for these medications, there are few data evaluating long-term compliance and persistence patterns among patients with AMI who have received these prescriptions outside of the context of a clinical trial. The effectiveness of any pharmacologic treatment is directly related to the physician’s choice of treatment and patient compliance with the given medication reg-
imen. After a patient has sustained an AMI, failure to comply with even some aspects of a recommended regimen can lead to preventable complications, rehospitalizations, or mortality. Previous research has shown that most cardiac-related morbidity and mortality occurs among elderly patients and that this population is at a higher risk for low compliance in taking cardiac medications.10 It is therefore crucial to characterize and understand the patterns of use of these drugs within this patient population. For this reason, we examined the prescription patterns for aspirin, -blockers, ACE inhibitors, and lipid-lowering drugs in the year after hospital discharge among patients aged ⱖ65 years who were admitted to Quebec hospitals for their first AMI between 1996 and 1998.
Materials and methods a
From the Division of Clinical Epidemiology, Montreal General Hospital, Montreal, Quebec, and the bDivisions of Cardiology and Clinical Epidemiology, Jewish General Hospital, McGill University, Montreal, Quebec, Canada. Dr Pilote is an investigator for the Canadian Institutes for Health Research (CIHR). Dr Eisenberg is an investigator for the Fonds Recherche en Sante´ du Que´bec. Supported in part by a grant from CIHR (#53181). Submitted February 27, 2002; accepted June 4, 2002. Reprint requests: Louise Pilote, MD, MPH, PhD, Division of Clinical Epidemiology, Montreal General Hospital, 1650 Cedar Ave, Montreal, Quebec, H3G 1A4, Canada. E-mail: [email protected] Copyright 2003, Mosby, Inc. All rights reserved. 0002-8703/2003/$30.00 ⫹ 0 doi:10.1067/mhj.2003.143
Study population A cohort of elderly (aged ⱖ65 years) patients who sustained an AMI was constructed by use of the administrative hospital discharge summary database for the province of Quebec (Med-Echo). The Med-Echo database contains records on hospital admissions for all citizens and legal residents in Quebec. We identified all elderly patients who were admitted with a main discharge diagnosis of AMI (International Classification of Diseases, 9th revision code 410) between Jan 1, 1996, and Dec 31, 1998, and who were discharged alive (n ⫽ 16,988). To increase comparability of patients, we then
American Heart Journal Volume 145, Number 3
Simpson et al 439
Table I. Characteristics of elderly (ⱖ65 years) patients who filled and did not fill outpatient prescriptions for cardiac medications after acute myocardial infarction in Quebec, 1996-1998 Characteristic Number of subjects* (%) Male (%) Age (median years) Length of stay (median days) Comorbidities (% of patients) Congestive heart failure Cerebrovascular disease Pulmonary edema Diabetes mellitus Malignant cancer Acute renal failure Shock Cardiac dysrhythmias Physician specialties (% patients) General Cardiologist Internist Other Inhospital procedures (% patients) Angiography PTCA CABG
No prescription
Aspirin
ACE inhibitors
-Blockers
Lipid-lowering drugs
1463 (10.4) 55.4 75 10
9134 (65.0) 56.7 74 9
6287 (44.7) 54.6 75 10
7595 (54.0) 57.0 73 9
2970 (21.1) 57.8 71 9
29.3 10.7 2.9 24.3 3.6 8.1 2.1 27.9
23.6 6.4 1.9 22.2 2.6 4.2 1.0 17.8
39.4 7.9 3.0 27.5 2.7 5.4 1.5 23.7
19.4 6.8 1.5 21.0 2.4 3.7 0.8 16.6
21.2 7.5 1.3 24.0 2.0 4.1 0.9 18.3
46.1 43.2 8.6 2.1
44.1 44.9 10.0 1.0
43.2 45.4 10.0 1.4
42.7 47.1 9.3 0.9
39.0 48.3 11.7 1.0
21.3 7.1 7.3
27.3 12.1 4.9
23.1 9.8 3.3
29.7 13.3 4.1
37.7 16.2 6.7
No prescription, Patients without a prescription filled for any of the four cardiac drug classes within 30 days of discharge; ACE, angiotensin-converting enzyme; PTCA, percutaneous transluminal coronary angioplasty; CABG, coronary artery bypass graft surgery. *The total number of study patients was 14,057.
included only patients with a first episode of AMI by ascertaining the absence of a hospital admission for AMI since January 1, 1988 (n ⫽ 1760). Next, using encrypted Medicare numbers, we linked the discharge summary database with the physician and drug claims database for the province of Quebec (RAMQ). This database contains information on all outpatient prescriptions for patients aged ⱖ65 years and on all inpatient and outpatient diagnostic and therapeutic procedures in Quebec. A number of exclusions were then made sequentially when 1) the main diagnosis of AMI was given after transfer from another hospitalization (n ⫽ 220), 2) there were incomplete data related to a change in Medicare number (n ⫽ 136), 3) the total length of stay was ⬍3 days (because this may indicate an unclear AMI diagnosis [n ⫽ 247]), or 4) patients were discharged to a long-term care institution, a rehabilitation center, or another province (because their prescriptions could not be ascertained [n ⫽ 568]). After these exclusions were made, our study population consisted of 14,057 patients, 82.7% of the original cohort.
Patient characteristics We grouped the study patients according to whether they had ⱖ1 class of cardiac medications prescribed at the time of hospital discharge. Patients who had obtained a prescription of at least one of the 4 major classes of cardiac medications (aspirin, -blockers, ACE inhibitors, lipid-lowering drugs) within 30 days of hospital discharge were classified as having an AMI drug prescription at discharge. It was thus possible for patients to be grouped into ⬎1 group when they had prescriptions from ⬎1
cardiac drug class. We then compared crude demographic and clinical characteristics, rates of cardiac procedure use, and specialties of the treating physicians for each group (Table I).
Time trends, compliance, and persistence Overall and annual rates of discharge prescriptions were assessed from 1996 to 1998. For each patient with a discharge prescription, we calculated their daily dosages on the basis of their prescriptions within one year after discharge. For each prescription, dosage results were grouped as either “clinical trial dosages” or “below clinical trial dosages.” A clinical trial dosage was defined as any dosage greater than or equal to the evaluated dosages administered in clinical trials, whereas a below clinical trial dosage was defined as any dosage that was less than the range of evaluated dosages in randomized clinical trials (Appendix, Table II).13-31 For example, we defined the clinical trial dosage for captopril as ⱖ12.5 mg/d, which was the lowest dosage administered in the Fourth International Study of Infarct Survival (ISIS-4) trial.14 When available, we selected only trials that evaluated secondary prevention after AMI. When clinical trials were not available, we defined “clinical trial dosages” according to the recommended starting dosages for patients with AMI in pharmaceutical guidelines. After we identified each prescription dosage as a clinical trial dosage or below clinical trial dosage, we placed patients in a dosage category for each of the 4 cardiac drug classes. To evaluate possible up- or down-titration of dosages, we also performed a number of additional analyses. First, we measured the dosages for each of the first 11 prescriptions
American Heart Journal March 2003
440 Simpson et al
Table II. Dosages of cardiac prescriptions, according to drug class Drug class Below clinical trial dosage (%) Discharge prescription Prescription at 1 year No dose switching at 1 year (%) Below clinical trial dosage* Above clinical trial dosage* Dose switching at 1 year (%) Below clinical trial dosage* Above clinical trial dosage* Dosage persistence at 1 year (%) Below clinical trial dosage* Above clinical trial dosage*
Aspirin
-Blockers
ACE inhibitors
Lipid-lowering drugs
0.5 0.7
80.4 79.0
11.7 11.1
52.4 44.7
0.3 99.0
73.1 14.7
7.6 83.3
40.0 42.4
0.2 0.5
7.4 4.9
4.1 5.0
12.4 5.3
78.4 99.9
95.5 85.6
78.3 97.6
87.7 95.1
Clinical trial dosages were defined as dosages for discharge prescriptions that were greater than or equal to the evaluated dosages administered in clinical trials. Below clinical trial dosages were defined as dosages for discharge prescriptions that were less than the range of evaluated dosages in clinical trials (Appendix). Dosage persistence was defined as the proportion of prescriptions filled that were of the same dosage category as discharge prescription dose, among all the prescriptions filled within the 1 year after discharge. *Dosage category at discharge.
after the discharge prescription and for the last prescription filled within one year after discharge. Second, we examined the proportions of patients who received discharge prescriptions that switched or did not switch dosage categories at any time during the year after discharge. Third, we calculated patients’ persistence on the dosage prescribed at discharge. Dosage persistence was defined as the proportion of prescriptions filled that were of the same dosage category as the discharge prescription dosage category, among all the prescriptions filled within the year after discharge. Compliance was calculated to determine the proportion of patients who continued to fill their prescriptions for each of the drug classes at one year after discharge. Compliance was calculated only for patients who were alive at one year and who received a discharge prescription. These patients were classified as “compliant” at one year when they had filled at least one prescription within 305 to 365 days after discharge. Similarly to earlier studies of persistence, we defined high persistence as the proportion of patients who received a discharge prescription and whose prescriptions covered at least 80% of the days within the year after discharge.10 For each patient, persistence was measured by means of the actual duration of medication use (in days) divided by the maximum possible duration of medication use (in days). For patients who died within one year after discharge, persistence was calculated as the actual duration of medication use divided by the total duration of medication use until the date of death. This analysis was performed for each patient according to drug class. All analyses were performed by use of the STATA (version 5, College Station, Tex) statistical software package.
Results Patient characteristics Overall prescription rates for aspirin, -blockers, ACE inhibitors, and lipid-lowering medications were subopti-
mal at discharge (65%, 54%, 45%, and 21%, respectively), and 10% of patients received no discharge prescriptions. The proportions of patients receiving prescriptions for any combination of 1, 2, or 3 of the 4 drugs were 24%, 40%, and 22%, respectively. Only 4% of patients received discharge prescriptions for all 4 drugs. Clinical and demographic characteristics were similar across the 5 prescription groups (Table I). However, there were more patients with ACE inhibitor prescriptions and congestive heart failure than patients with congestive heart failure in the remaining prescription groups (39.4% vs 19.4%, 23.6%, and 21.2% for -blockers, aspirin, and lipid-lowering drugs, respectively). Patients without prescriptions for any of the 4 drug classes had higher rates of comorbidities than patients with prescriptions. For the 4 drug classes, the group with prescriptions for lipid-lowering medications had the highest proportion of patients who underwent angiography, percutaneous transluminal coronary angioplasty, or coronary artery bypass grafting (37.7%, 16.2%, and 6.7%, respectively). A greater proportion of patients who had prescriptions were treated by cardiologists or internists on hospitalization for AMI than patients who did not receive any prescriptions. Similarly, the group without prescriptions and the ACE inhibitor drug group had the lowest proportion of patients who underwent angiography or percutaneous transluminal coronary angioplasty.
Time trends, compliance, and persistence There were temporal changes in the rates of cardiac prescriptions between 1996 and 1998 (Figure 1). During the study period, there was a marked increase in the proportion of patients with lipid-lowering drug prescriptions at discharge (11.3% difference). In contrast, there was a
American Heart Journal Volume 145, Number 3
Simpson et al 441
Figure 1
Time trends in discharge cardiac prescriptions for elderly patients with acute myocardial infarction. For each year, patients who were prescribed at least one of the 4 major classes of cardiac medications within 30 days of hospital discharge were classified as having an acute myocardial infarction prescription at discharge. Number of patients: 1996, n ⫽ 4732; 1997, n ⫽ 4711; 1998, n ⫽ 4614. -blockers, -adrenergic receptor blockers; ACE, angiotensin-converting enzyme.
decrease in the proportion of patients who received aspirin prescriptions after discharge (7.8% difference). Finally, there were only modest increases in rates of -blocker and ACE inhibitor prescriptions (1.7% and 2.7% differences, respectively). For aspirin, -blockers, and ACE inhibitors, most patients received an initial prescription immediately after hospital discharge (median time to initial prescription, 0 days for aspirin, -blockers, and ACE inhibitors). However, it took longer for patients to receive prescriptions for lipid-lowering medications (median time to initial prescription, 20 days). When the rates for discharge prescriptions were compared with rates for cardiac prescriptions within one year after discharge, there was no substantial increase in the rates of aspirin, -blocker, and ACE inhibitor prescriptions (prescription rates at 1 year: aspirin 74.9%, -blockers 60.4%, ACE inhibitors 53.2%). The rate of lipid-lowering drug prescriptions, however, was higher at one year than at the time of discharge (prescription rate at 1 year 36.4%). Similarly, of the patients who did not receive discharge prescriptions but who were alive at one year, only a small number were “late starters” and filled a prescription within 11 to 12 months after discharge for -blockers, ACE inhibitors, and aspirin (4.7%, 6.4%, and 7.9%, respectively). However, greater proportions of patients had late prescriptions for lipidlowering drugs filled (14.1%).
Whether patients with cardiac prescriptions received initial dosages that were within the range of dosages administered in clinical trials depended on the drug class (Table II). Of patients to whom aspirin and ACE inhibitors were prescribed, ⬎85% received clinical trial dosages for these drug classes (99.4% and 88.3%, respectively). In contrast, initial prescription dosages for -blockers and lipid-lowering drugs were often lower than those used in randomized controlled trials. Of the patients who received discharge prescriptions for -blockers and lipid-lowering drugs, only 19.6% and 47.6%, respectively, received prescriptions that were within clinical trial dosages. When the prescriptions at one year were assessed, there were only modest increases in the proportions of patients receiving clinical trial dosage prescriptions for aspirin, ACE inhibitors, and -blockers compared with initial prescriptions. However, there were larger increases for patients with lipid-lowering drugs (Table II). The same trends were evident when looking at each of the first 11 prescriptions after the initial discharge prescription (data not shown). The high dosage persistence rates provide additional evidence that patients tended to stay within the same dosage category in the year after discharge (Table II). When receiving a prescription for AMI, patients showed high rates of compliance and persistence for these drugs (Table III). The one-year compliance rates
American Heart Journal March 2003
442 Simpson et al
Table III. Compliance and persistence with cardiac prescriptions at one year, according to drug class
Drug class Compliance (%) High persistence (%)
LipidACE lowering Aspirin -Blockers inhibitors drugs 74.4 71.4
73.7 71.6
69.6 69.4
84.0 79.7
Compliance was defined as the proportion of patients who filled a discharge prescription and who filled at least 1 prescription within 305 to 365 days after discharge. High persistence was defined as the proportion of patients who filled a discharge prescription and whose prescriptions covered at least 80% of the days within the year after discharge.
were high for each of the 4 drug categories. The highest rate of compliance was associated with the use of lipid-lowering medications (84.0%, 95% CI 82.6%85.4%), followed by aspirin (74.4%, 95% CI 73.5%75.4%), -blockers (73.7%, 95% CI 72.6%-74.7%), and ACE inhibitors (69.6%, 95% CI 68.3%-70.8%). Measurements of patient persistence showed that there were high proportions of patients who maintained their drug prescriptions for at least 80% of the one-year study period (high persistence) (Table III). A total of 79.7% of patients (95% CI 78.2%-81.2%) who had lipid-lowering drug prescriptions filled at discharge had high persistence rates, the highest rate for any of the 4 drug classes. For the remaining drug categories, 71.6% of patients (95% CI 70.6%-72.6%) with -blocker prescriptions, 71.4% of patients (95% CI 70.5%-72.3%) with aspirin prescriptions, and 69.4% of patients (95% CI 68.2%-70.5%) with prescriptions for ACE inhibitors had high persistence rates.
Discussion In this study, we describe the trends for post-AMI drug use in Quebec in the 3 years from 1996 to 1998. Our results suggest that recommended cardiac drugs are prescribed at suboptimal rates and dosages for patients with AMI. However, once prescribed, patients showed high rates of compliance and persistence for these medications in the year after discharge. To date, there are limited data about patient compliance and persistence for cardiac drug use after AMI. Unlike in earlier studies of persistence of lipid-lowering medication use after AMI,10 in our study patients showed high persistence for lipid-lowering drugs and for aspirin, -blockers, and ACE inhibitors. Their high levels of persistence suggest that, to maximize the benefits of cardiac medications, physicians should ensure that these drugs are prescribed to all eligible patients at discharge. However, although the rates of patient compliance were high, there remained approximately 25% of patients to whom aspirin, -blockers, or ACE inhibitors were prescribed who did not continue to renew these prescrip-
tions at one year after discharge. These findings suggest that new strategies must be found to achieve higher compliance rates. Even with evidence of benefit, the annual prescription rates for aspirin, -blockers, and ACE inhibitors did not change substantially with time. However, there was a marked increase in the rate of lipid-lowering drug prescriptions during the study period. This increase might be explained by the advent of new evidence demonstrating the efficacy of lipid-lowering treatment within 3 to 6 months of hospital discharge after AMI and by the availability of novel lipid-lowering medications.32 However, room for improvement remains. On the basis of the general lack of contraindications to lipid-lowering medications, especially statins, most patients who survive an AMI should have lipid-lowering medications prescribed at discharge.33 There was also a modest decline in aspirin prescription rates. This change may be related to the drug co-payment policies initiated by the Quebec government in 1996, which could have increased the likelihood that patients would obtain over-the-counter aspirin preparations rather than prescribed aspirin. Of the 4 drug classes, -blockers were the least often prescribed at the same dosages as those administered in clinical trials. These results are similar to those found in an earlier observational study, in which almost 50% of patients receiving prescriptions for -blockers were treated with dosages that were less than half the dosages administered in randomized clinical trials.34 In addition, the administered dosages for aspirin, ACE inhibitors, -blockers, and—to a lesser extent—lipid-lowering medications did not change in the year after discharge, suggesting that physicians did not increase dosages on the basis of tolerance. However, whether lower dosages of aspirin, -blockers, ACE inhibitors, or lipid-lowering drugs would be equally effective at reducing post-AMI mortality rates has yet to be determined. It is also possible that elderly patients require lower than trial dosages of medications, in comparison with younger patients. The assessment of patient characteristics indicated that the decision to prescribe certain cardiac drug classes instead of others did not depend solely on disease severity. Other unexplained factors likely play a role in the suboptimal use of cardiac prescriptions after AMI. Nonetheless, comparisons of the clinical characteristics for patients with no cardiac prescriptions versus patients with at least one prescription indicated that the physician choice to either prescribe or forgo cardiac medications altogether was influenced by the presence of associated comorbidities. It is possible that the higher rates of comorbidities among patients without prescriptions may have led their physicians to forgo the use of preventive therapies in view of a perceived poor prognosis.
American Heart Journal Volume 145, Number 3
Our study was limited because, due to the nature of the drugs claims database, we did not have sufficient inhospital data pertaining to clinical characteristics of the study patients. As a result, we were unable to accurately exclude patients with contraindications from our analyses. Consequently, we may have underestimated the prescription rates for the 4 drug classes studied. However, because there is no single set of criteria that define contraindications to each of the studied drug classes, contraindication assessment tends to be very subjective. As a result, certain contraindications may be completely preclusive to drug therapy (“absolute contraindications”), whereas other contraindications may be deemed as permissive to therapy with certain conditions (“relative contraindications”). The looseness of this definition made it very difficult to estimate accurately patient contraindications to cardiac drug treatment, especially from an administrative database. Additionally, we could not identify the reasons for suboptimal prescription practices from our administrative databases. Finally, we cannot determine from our databases whether all the patients who continued to renew their cardiac prescriptions were actually taking their medication as it was prescribed. However, the assessment of prescription refill patterns does provide an accurate and practical means of approximating medication use in large study populations.10 Despite the limitations of using administrative databases, previous studies have clearly demonstrated that these data provide an accurate standardized tool for assessing patterns of medication use in large patient populations.35-38 Specifically for the Quebec hospital summary and prescription claims databases, recent reports have demonstrated that their predictive value in identifying patients with AMI is high and that this information gives an accurate summary of prescriptions dispensed to patients aged ⱖ65 years.39,40 The results of this study suggest that, if physicians prescribe medications according to evidence-based guidelines, patients are likely to adhere to these prescriptions in the long term. Thus, although we are not certain of all factors that are responsible for long-term prescription patterns after AMI, these results suggest that the suboptimal adherence to guidelines evident in Quebec from 1996 to 1998 arose primarily because of low rates of discharge prescriptions on the part of physicians.
References 1. Jamali AH, Wilson Tang WH, Khot UN, et al. The role of angiotensin receptor blockers in the management if chronic heart failure. Arch Int Med 2001;161:667-72. 2. Collaborative overview of randomized trials of antiplatelet therapy—I: prevention of death, myocardial infarction, and stroke by prolonged antiplatelet therapy in various categories of patients. Antiplatelet Trials Collaboration. BMJ 1994;308:81-106. 3. Gottlieb SS, McCarter RJ, Vogel RA. Effect of beta-blockade on mortality among high-risk and low-risk patients after myocardial infarction. New Engl J Med 1998;339:489-97.
Simpson et al 443
4. Crouch MA. Effective use of statins to prevent coronary heart disease. Am Fam Physician 2001;63:309-20. 5. Mendelson G, Aronow WS. Underutilization of angiotensin-converting enzyme inhibitors in older patients with Q-wave anterior myocardial infarction in an academic hospital-based geriatrics practice. Am Geriatr Soc 1998;46:751-2. 6. Heller DA, Ahern FM, Kozak M. Changes in rates of -blocker use between 1994 and 1997 among elderly survivors of acute myocardial infarction. Am Heart J 2000;140:663-71. 7. Krumholz HM, Radford MJ, Ellerbeck EF, et al. Aspirin for secondary prevention after acute myocardial infarction in the elderly: prescribed use and outcomes. Ann Intern Med 1996;124:292-8. 8. Bown Luzier A, Navsarikar A, Wilson MF, et al. Patterns of prescribing ACE inhibitors after myocardial infarction. Pharmacotherapy 1999;19:655-60. 9. Rochon PA, Anderson GM, Tu JV, et al. Use of -blocker therapy in older patients after acute myocardial infarction in Ontario. CMAJ 1999;161:1403-8. 10. Avorn J, Monette J, Lacour A, et al. Persistence of use of lipidlowering medications. JAMA 1998;279:1458-62. 11. Beck CA, Lauzon C, Eisenberg MJ, et al. Discharge prescriptions following admission for acute myocardial infarction at tertiary care and community hospitals in Quebec. Can J Cardiol 2001;17:33-40. 12. Krumholz HM, Radford MJ, Wang Y, et al. National use and effectiveness of -blockers for the treatment if elderly patients after acute myocardial infarction. National Cooperative Cardiovascular Project. JAMA 1998;280:623-9. 13. Canadian Pharmacists Association Group. Bruce LD, editor. Compendium of pharmaceuticals and specialties. 33rd ed. Ottawa: Canadian Pharmacists Association; 1998. 14. ISIS-4 (Fourth International Study of Infarct Survival) Collaborative Group. ISIS-4: a randomised factorial trial assessing early oral captopril, oral mononitrate, and intravenous magnesium sulphate in 58,050 patients with suspected acute myocardial infarction. Lancet 1995;345:669-85. 15. Swedberg K, Held P, Kjekshus J, et al. Effects of the early administration of enalapril on mortality in patients with acute myocardial infarction: results of the Cooperative New Scandinavian Enalapril Survival Study II (CONSENSUS II). New Engl J Med 1992;327:678-84. 16. The SOLVD Investigators. Effect of enalapril on survival in patients with reduced left ventricular ejection fractions and congestive heart failure. New Engl J Med 1991;325:293-302. 17. Borghi C, Marino P, Zardini P, et al. Short- and long-term effects of early fosinopril administration in patients with acute anterior myocardial infarction undergoing intravenous thrombolysis: results from the Fosinopril in Acute Myocardial Infarction Study. FAMIS Working Party. Am Heart J 1998;136:213-25. 18. Gruppo Italiano per lo Studio della Sopravvivenza nell’infarto Miocardico. GISSI-3: effects of lisinopril and transdermal glyceryl trinitrate singly and together on 6-week mortality and ventricular function after acute myocardial infarction. Lancet 1994;343:1115-22. 19. Anonymous. PREAMI: Perindopril and Remodelling in Elderly with Acute Myocardial Infarction: study rationale and design. Cardiovasc Drugs Ther 2000;14:671-9. 20. The Acute Infarction Ramipril Efficacy (AIRE) Study Investigators. Effect of ramipril on mortality and morbidity of survivors of acute myocardial infarction with clinical evidence of heart failure. Lancet 1993;342:821-8. 21. Neil HA, Fowler G, Patel H, et al. An assessment of the efficacy of atorvastatin in achieving LDL cholesterol target levels in patients
American Heart Journal March 2003
444 Simpson et al
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36. 37.
with coronary heart disease: a general practice study. Int J Clin Pract 1999;53:509-13. The BIP Study Group. Secondary prevention by raising HDL cholesterol and reducing triglycerides in patients with coronary artery disease: the Bezafibrate Infarction Prevention (BIP) study. Circulation 2000;102:21-7. Riegger G, Abletshauser C, Ludwig M, et al. The effect of fluvastatin on cardiac events in patients with symptomatic coronary artery disease during one year of treatment. Artherosclerosis 1999;144:263-7. Downs JR, Clearfield M, Weis S, et al. Primary prevention of acute coronary events with lovastatin in men and women with average cholesterol levels: results of AFCAPS/TexCAPS. JAMA 1998;279: 1615-22. The Long-Term Intervention with Pravastatin in Ischemic Disease (LIPID) Study Group. Prevention of cardiovascular events and death with pravastatin in patients with coronary heart disease and a broad range of initial cholesterol levels. New Engl J Med 1998; 339:1349-57. Scandinavian Simvastatin Survival Study Group. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet 1994;344:1383-9. American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on Management of Acute Myocardial Infarction). 1999 update: ACC/AHA guidelines for the management of patients with acute myocardial infarction: executive summary and recommendations. Circulation 1999;100:1016-30. Wilcox RG, Roland JM, Banks DC, et al. Randomised trial comparing propranolol with atenolol in immediate treatment of suspected myocardial infarction. BMJ 1980;280:885-8. Hjalmarson A, Herlitz J, Holmberg S, et al. The Goteborg metoprolol trial: effects on mortality and morbidity in acute myocardial infarction. Circulation 1983;67(1 Suppl):I26-32. -Blocker Heart Attack Trial Research Group. A randomized trial of propranolol in patients with acute myocardial infarction. JAMA 1982;247:1707-14. Norwegian Multicenter Study Group. Timolol-induced reduction in morbidity and reinfarction in patients surviving acute myocardial infarction. New Engl J Med 1981;304:801-7. Jackevicius CA, Anderson GM, Leiter L, et al. Use of the statins in patients after acute myocardial infarction: does evidence change practice? Arch Intern Med 2001;161:183-8. Expert Panel on Detection Evaluation, and Treatment of High Blood Cholesterol in Adults. Summary of the second report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel II). JAMA 1993;269:3015-23. Barron HV, Viskin S, Lundstrom RJ, et al. -blocker dosages and mortality after acute myocardial infarction. Arch Int Med 1998; 158:449-53. Bright RA, Avorn J, Everitt DE. Medicaid data as a resource for epidemiologic studies: strengths and limitations. J Clin Epidemiol 1989;42:937-45. Ray WA, Griffin MR. Use of Medicaid data for pharmacoepidemiology. Am J Epidemiol 1989;129:837-49. Romano PS, Roos LL, Luft HS, et al. A comparison of administrative versus clinical data: coronary artery bypass surgery as an ex-
ample. Ischaemic Heart Disease Patient Outcomes Research Team. J Clin Epidemiol 1994;47:249-60. 38. D’Hoore W, Bouckaert A, Tilquin C. Practical considerations on the use of the Charlson comorbidity index. J Clin Epidemiol 1996; 49:1429-33. 39. Tamblyn R, Lavoie G, Petrella L, et al. The use of prescription claims databases in pharmacoepidemiological research: the accuracy and comprehensiveness of the prescriptions claims database in Que´ bec. J Clin Epidemiol 1995;48:999-1009. 40. Levy AR, Tamblyn R, Fitchett D, et al. Coding accuracy of hospital discharge data for elderly survivors of myocardial infarction. Can J Cardiol 1999;15:1277-82.
Appendix Classification of clinical trial dosage therapies for all acute myocardial infarction medications from the Quebec physicians and drugs claims database (RAMQ). Classification of clinical trial dosage therapies for all AMI medications from the Quebec physicians and drugs claims database (RAMQ) Drug class ACE inhibitors Benazepril13 Captopril14 Cilazapril13 Enalapril15,16 Fosinopril17 Lisinopril18 Perindopril19 Quinapril13 Ramipril20 Lipid-lowering drugs Atorvastatin21 Bezafibrate22 Cerivastatin13 Cholestyramine13 Clofibrate13 Colestipol (suspension)13 Dextrothyroxine sodium13 Fenofibrate13 Fluvastatin23 Gemfibrozil13 Lovastatin24 Nicotinic acid13 Pravastatin25 Simvastatin26 Aspirin27 -Blockers Acebutolol13 Atenolol28 Metoprolol29 Nadolol13 Oxprenolol13 Pindolol13 Propranolol30 Timolol31
Clinical trial dosage (mg/day)
ⱖ10 ⱖ12.5 ⱖ1.25 ⱖ5 ⱖ5 ⱖ5 ⱖ2 ⱖ10 ⱖ5 ⱖ10 ⱖ400 ⱖ0.4 ⱖ4000 ⱖ2000 ⱖ2000 ⱖ1 ⱖ300 ⱖ40 ⱖ1200 ⱖ20 ⱖ1500 ⱖ40 ⱖ20 ⱖ160 ⱖ200 ⱖ100 ⱖ200 ⱖ80 ⱖ60 ⱖ10 ⱖ20 ⱖ10