Prevention of adverse drug reactions in hospitalized older patients with multi-morbidity and polypharmacy: the SENATOR* randomized controlled clinical trial
Denis O’Mahony, Aðalsteinn Guðmundsson, Roy L. Soiza, Mirko Petrović, Alfonso J. Cruz‐Jentoft, Antonio Cherubini, Richard Fordham, Stephen Byrne, Darren Dahly, Paul Gallagher, Amanda Lavan, Denis Curtin, Kieran Dalton, Shane Cullinan, Evelyn Flanagan, Frances Shiely, Ólafur Samúelsson, Ástrós Sverrisdóttir, Selvarani Subbarayan, Lore Vandaele, Eline Meireson, Beatriz Montero‐Errasquín, Aurora Rexach-Cano, Andrea Correa Perez, Isabel Lozano-Montoya, Manuel Vélez‐Díaz‐Pallarés, Annarita Cerenzia, Samanta Corradi, Maria Soledad Ferreiro, Federica Dimitri, Paolo Marinelli, Gaia Martelli, Rebekah Fong Soe Khioe, Joseph A. Eustace
Abstract
BACKGROUND: Multi-morbidity and polypharmacy increase the risk of non-trivial adverse drug reactions (ADRs) in older people during hospitalization. Despite this, there are no established interventions for hospital-acquired ADR prevention. METHODS: We undertook a pragmatic, multi-national, parallel arm prospective randomized open-label, blinded endpoint (PROBE) controlled trial enrolling patients at six European medical centres. We randomized 1,537 older medical and surgical patients with multi-morbidity and polypharmacy on admission in a 1:1 ratio to SENATOR software-guided medication optimization plus standard care (intervention, n = 772, mean number of daily medications = 9.34) or standard care alone (control, n = 765, mean number of daily medications = 9.23) using block randomization stratified by site and admission type. Attending clinicians in the intervention arm received SENATOR-generated advice at a single time point with recommendations they could choose to adopt or not. The primary endpoint was occurrence of probable or certain ADRs within 14 days of randomization. Secondary endpoints were primary endpoint derivatives; tertiary endpoints included all-cause mortality, re-hospitalization, composite healthcare utilization and health-related quality of life. RESULTS: For the primary endpoint, there was no difference between the intervention and control groups (24.5 vs. 24.8%; OR 0.98; 95% CI 0.77-1.24; P = 0.88). Similarly, with secondary and tertiary endpoints, there were no significant differences. Among attending clinicians in the intervention group, implementation of SENATOR software-generated medication advice points was poor (~15%). CONCLUSIONS: In this trial, uptake of software-generated medication advice to minimize ADRs was poor and did not reduce ADR incidence during index hospitalization.