Aggregated and individual pharmacy record data. Association between antibiotic and antihypertensive drug use

Abstract

In a previous attempt to relate drug use to morbidity wecompared “sickness numbers ” and dispensed drug volumesin Swedish municipalities and found positive correlationsbetween sickness numbers and dispensed volumes of mostdrug types. Antibiotics were an exception; higher anti-biotics volumes were associated with lower sicknessnumbers [ 1].Aggregated pharmacy record data showed that munici-palities with antibiotics volumes above the national averagehad lower volumes of cardiovascular (CV) drugs, primarilyof antihypertensives [ 1], indicating a possible link betweeninfection and hypertension. However, the employed aggre-gated pharmacy record data gave no information regardingthe relation between the dispensed volumes of antibioticsand CV drugs for individuals. The present study aimed atobtaining such information.The dispensed drug volumes in Sweden (2006) in DDDof antibiotics (ATC code J01) and the CV drugs diuretics(C03), beta-blocking agents (C07), calcium antagonists(C08) and ACE-inhibitors (C09) per 10-year age groupsfor patients aged 40 years were extracted from theSwedish Prescribed Drug Register [ 2]. The number ofindividuals and percentages of the population with anti-biotics or CV drugs were determined. The expectedpercentage with both an tibiotics and CV drugs wascalculated as the product of the proportion of the populationwith antibiotics times the proportion with CV drugs. Theratios of observed percentage/expected percentage werecalculated for each age group.To determine if the use of antibiotics influenced the CVdrugvolumeperindividual,wecomparedforeachagegroupthe average volumes of CV drugs (DDDs of C03-C09) forpatients with both antibiotics and CV drugs with the averagevolumes of CV drugs for patients with no antibiotics.The extracted data comprised 2.35 million individuals orbetween 347,000 and 597,000 individuals per age group.The ratios of observed/expected percentages of patientswith both antibiotics and CV drugs were higher than 1 infour of the five age groups, indicating that more patientsthan expected had acquired both types of drugs. Theaverage volumes of CV drugs (average DDD/patient) were5–6% higher per patient treated with antibiotics than forthose not treated with antibiotics.The present study showed that the proportion of thepopulation with both antibiotics and CV drugs was higherthan expected and that the groups of individuals treatedwith antibiotics on average had 5 –6% higher volumes ofCV drugs than those without antibiotics. Consequently, theuse of antibiotics was not associated with lower volumes ofCV drugs at the individual level as might be expected fromthe previous study using aggregated data [ 1]. This is inaccordance with reported observations where patients withCV disease were likely to be frequent attenders [ 3] of healthcare centres and to receive prescriptions of antibiotics as co-medication [ 4].