Impact of chronic kidney disease on case ascertainment for hospitalised acute myocardial infarction: an English cohort study

Objectives Acute myocardial infarction (AMI) case ascertainment improves for the UK general population using linked health data sets. Because care pathways for people with chronic kidney disease (CKD) change based on disease severity, AMI case ascertainment for these people may differ compared with the general population. We aimed to determine the association between CKD severity and AMI case ascertainment in two secondary care data sets, and the agreement in estimated glomerular filtration rate (eGFR) between the same data sets. Methods We used a cohort study design. Primary care records for people with CKD or risk factors for CKD, identified using the National CKD Audit (2015-2017), were linked to the Myocardial Ischaemia National Audit Project (MINAP, 2007-2017) and Hospital Episode Statistics (HES, 2007-2017) secondary care registries. People with an AMI recorded in either MINAP, HES or both were included in the study cohort. CKD status was defined using eGFR, derived from the most recent serum creatinine value recorded in primary care. Moderate-severe CKD was defined as eGFR <60 mL/min/1.73 m(2), and mild CKD or at risk of CKD was defined as eGFR >= 60 mL/min/1.73 m(2) or eGFR missing. CKD stages were grouped as (1) At risk of CKD and Stages 1-2 (eGFR missing or >= 60 mL/min/1.73 m(2)), (2) Stage 3a (eGFR 45-59 mL/min/1.73 m(2)), (3) Stage 3b (eGFR 30-44 mL/min/1.73 m(2)) and (4) Stages 4-5 (eGFR <30 mL/min/1.73 m(2)). Results We identified 6748 AMIs: 23% were recorded in both MINAP and HES, 66% in HES only and 11% in MINAP only. Compared with people at risk of CKD or with mild CKD, AMIs in people with moderate-severe CKD were more likely to be recorded in both MINAP and HES (42% vs 11%, respectively), or MINAP only (22% vs 5%), and less likely to be recorded in HES only (36% vs 84%). People with AMIs recorded in HES only or MINAP only had increased odds of death during hospitalisation compared with those recorded in both (adjusted OR 1.61, 95% CI 1.32 to 1.96 and OR 1.60, 95% CI 1.26 to 2.04, respectively). Agreement between eGFR at AMI admission (MINAP) and in primary care was poor (kappa (K) 0.42, SE 0.012). Conclusions AMI case ascertainment is incomplete in both MINAP and HES, and is associated with CKD severity.