Ruigomez A, Plana E, Gilsenan A, Fortuny J, Cainzos-Achirica M, Flynn RWV, MacDonald TM, Garcia-Rodriguez L, Ziemiecki R, Andrews EB. Identification and validation of major cardiovascular events in the United Kingdom data sources included in a multi-database post-authorization safety study of prucalopride. Drug Saf. 2021 May;44(5):541-51.


INTRODUCTION: A multinational post-authorization safety study assessed cardiovascular safety in initiators of prucalopride for chronic constipation compared with a matched cohort of polyethylene glycol 3350 initiators. The primary safety outcome was major adverse cardiovascular events (MACE), a composite of hospitalization for acute myocardial infarction, stroke, or in-hospital cardiovascular death. We report the validation process for MACE endpoints in United Kingdom (UK) data sources: Clinical Practice Research Datalink (CPRD GOLD), The Health Improvement Network (THIN), and the Information Services Division (ISD) Scotland.

METHODS: Modified electronic algorithms from prior research identified potential MACE cases. Validation followed a common protocol, adapted for each database, with all information anonymized: (1) direct confirmation via linkage to hospital records (CPRD GOLD); (2) requests for additional clinical information through questionnaires (CPRD GOLD), free-text (THIN), or abstraction of hospital records (ISD); (3) manual review of electronic records of potential events retrieved by the algorithm (CPRD GOLD/THIN); and (4) event adjudication by three clinicians, blinded to exposure, for all remaining events.

RESULTS: Electronic algorithms identified 260 potential MACE cases: 38 confirmed via linkage to hospital records (CPRD GOLD), 56 ruled out as non-cardiovascular death cases (THIN), and three unavailable for review (ISD), leaving 163 potential cases. After manual review with additional information (steps 2 and 3), 45 were considered noncases (CPRD GOLD/THIN). Upon final adjudication (step 4), remaining potential events were adjudicated as definite (n = 62), probable (n = 10), possible (n = 13), or noncases (n = 33).

CONCLUSIONS: Given the limitations of relying solely on computer algorithms to identify cardiovascular outcomes, validation with clinical review is essential to guide interpretation.

Share on: