SP189ASSOCIATION BETWEEN COMMUNITY PRESCRIBING OF ANTIHYPERTENSIVE DRUGS AND INCIDENCE RATES OF ACUTE KIDNEY INJURY IN THE UK

Abstract

Introduction and Aims: Patients taking ACE-Inhibitors and Angiotensin Receptor Blockers (ACE/ARB) are thought to be at greater risk of acute kidney injury (AKI) but the population incidence of AKI among patients taking these drugs and the degree of increased risk for different prescribing indications is not known. We therefore sought to estimate the incidence rate of AKI in the UK general population among patients taking these drugs compared to users of alternative antihypertensive drugs and controls with no antihypertensive drug use. Methods: We conducted a cohort study using UK primary care data from the Clinical Practice Research Datalink linked to Hospital Episodes Statistics (HES). The cohort included patients who were new users of 1) ACE/ARB, 2) Calcium channel blockers (CCB), 3) Thiazide diuretics, 4) Beta-blockers (BB) or who were controls, never previously prescribed antihypertensives and matched to each antihypertensive user for age, sex and GP practice. Drug use for each class was time updated; all included patients had known renal function prior to cohort entry. The outcome was a first episode of AKI defined either during a hospital admission using ICD-10 coding in HES (hospital AKI) and separately using the NHS e-alert algorithm applied to outpatient creatinine testing during risk periods in the 2 weeks following acute infections (community AKI: results not described here). We estimated rates of AKI among ACE/ARB users, both overall and within subgroups. Using Poisson regression we estimated hazard ratios associated with use of each antihypertensive drug class, adjusting for potential confounders. In a further analysis we estimated rates and HRs for patients taking ACE/ARB in combination with loop, thiazide and potassium-sparing diuretics compared to those taking ACE/ARB alone. Results: Among a cohort of 427,945 patients with a mean follow-up of 4.1 years there were 6031 cases of AKI associated with hospital admission. The crude rate of AKI among patients taking ACE/ARB was 6.5 cases per 1000 person years at risk (95% CI 6.3-6.8). Rates of AKI among patients taking ACE/ARB varied markedly by baseline renal function (figure), by age and by co-morbidities. After adjustment for age, gender, diabetes, cardiac failure, hypertension, ischaemic heart disease, CKD stage and other antihypertensives the hazard ratios for AKI were 1.7 (1.6-1.8) for those taking ACE/ ARB compared to non-users, 1.5 (1.4-1.6) for BB, 1.1 (1.0-1.2) for CCB and 0.8 (0.8-0.9) for thiazides. Compared to patients taking ACE/ARB alone the fully-adjusted HRs for AKI were 3.0 (2.7-3.3) for those also taking loop diuretics, 5.9 (5.0-7.0) for those taking loop and K-sparing diuretics and 8.0 (3.0-21.4) for those taking loop, potassium-sparing and thiazide diuretics. Conclusions: These results are robust estimates of the UK population incidence of AKI among patients taking different antihypertensives stratified by comorbidities. Rates of AKI among patients taking ACEI/ARB vary markedly according to patient characteristics and the fully-adjusted hazard ratio for Beta-blocker use is close to that of ACE/ARB. These results have important implications for the development of 'sick-day rules'. (Table Presented).