Premium
Definition of patients treated with evidence based drugs in absence of prescribed daily doses: the example of acute myocardial infarction
Author(s) -
Belleudi Valeria,
Fusco Danilo,
Kirchmayer Ursula,
Agabiti Nera,
Di Martino Mirko,
Narduzzi Silvia,
Davoli Marina,
Arcà Massimo,
Perucci Carlo Alberto
Publication year - 2011
Publication title -
pharmacoepidemiology and drug safety
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.023
H-Index - 96
eISSN - 1099-1557
pISSN - 1053-8569
DOI - 10.1002/pds.2079
Subject(s) - medicine , myocardial infarction , pharmacoepidemiology , intensive care medicine , cardiology , emergency medicine , pharmacology , medical prescription
Purpose Define patients treated with evidence‐based drugs in a cohort discharged after acute myocardial infarction (AMI) in absence of prescribed daily doses (PDD). To compare different drug use measures and analyze their impact on the effect estimate of risk factors related to drug use. Methods AMI patients discharged in Rome during 2006–2007 were selected from the Hospital Information System. Drugs claimed during the 12 months after discharge were retrieved. Measures of drug use were defined as: ‘ continuity ’ (one prescription each follow‐up quarter‐year) and the ‘ proportion of days covered ’ calculated by defined daily doses (DDDs) or pill counts (PCs) (≥80% of individual follow‐up). Poly‐therapy was defined through the same drug use measure for all drug groups. Kappa index was calculated to analyze the concordance between measures. For each measure we estimated the effect of age, gender and Percutaneous Transluminal Coronary Angioplasty (PTCA) on poly‐therapy. Results Poly‐therapy rates varied between 11.5 and 37.8% in the cohort and between 17.3 and 56.9% in patients with at least one prescription for all drugs. Concordance between all measures was high for antiplatelets ( k = 0.74) and very low for beta‐blockers ( k = 0.22). According to measures used, gender and older age effects slightly varied, while PTCA remained a strong determinant of drug use. Conclusions Different measures of exposure to drug treatment may affect the estimate of the proportion of treated patients and the effect estimates of risk factors. Drug dispense registries are useful, but it is necessary to develop and validate methodologies in absence of PDD. Copyright © 2010 John Wiley & Sons, Ltd.