Determining the most frequent combinations of drugs prescribed prior to bullous pemphigoid diagnosis using association rule mining: a UK population-based case–control study

Le résumé fourni par la source

BACKGROUND: Bullous pemphigoid (BP) is a serious autoimmune skin disease affecting older people. Given the highly prevalent polypharmacy in older people, identifying drug combinations associated with high BP risk is needed. OBJECTIVES: To identify drug combinations frequently prescribed for older people prior to BP using association rule mining (ARM) and to estimate their risk of developing BP. METHODS: We conducted a nested case-control study using UK Clinical Practice Research Datalink primary care records (1998-2021). ARM was used to identify the 10 most common drug class or active substance pairs most likely coprescribed (lift) on the same day and within 30 days of one another (sensitivity analysis). Fold change (FC; case-to-control lift ratio) was used to identify combinations with greater lift in cases. Multivariable conditional logistic regression was performed to estimate BP risk following combinations and their constituent drugs. RESULTS: Among 16 844 cases and 79 493 controls, the most frequent drug class combinations associated with increased BP risk with greater lift in cases were angiotensin-converting enzyme inhibitors (ACEIs)-statins (FC 1.31; FC in sensitivity analysis 1.18), antiplatelets-statins (1.23; 1.11), proton pump inhibitors (PPIs)-antiplatelets (1.22; 1.14), PPIs-statins (1.22; 1.14) and ACEIs-antiplatelets (1.20; 1.09). Substance combinations with a greater lift in cases included ramipril-simvastatin (FC 1.30), simvastatin-aspirin (1.21), ramipril-aspirin (1.19) and bendroflumethiazide-aspirin (1.16). The results of sensitivity analyses for substances were similar. BP risk following drug combinations was lower than after single drugs. CONCLUSIONS: Using ARM we found that the most frequently prescribed drug combinations associated with increased BP risk were used to treat cardiovascular diseases and hypertension. Patients should be vigilant for skin changes following these drugs rather than avoid them.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.