Prevention of Adverse Drug Reactions in Hospitalised Older Patients Using a Software-Supported Structured Pharmacist Intervention: A Cluster Randomised Controlled Trial

David O’Sullivan, Denis O’Mahony, Marie N. O’Connor, Paul Gallagher, James Gallagher, Shane Cullinan, Richard O’Sullivan, Joseph Eustace, Stephen Byrne

Proven interventions to reduce adverse drug reactions (ADRs) in older hospitalised patients are lacking. Previous randomised controlled trial (RCT) data indicate that a structured pharmacist review of medication (SPRM) can reduce inappropriate prescribing in older hospitalised patients. However, no RCT data show that an SPRM reduces ADRs in this population. We performed a cluster RCT comparing a clinical decision support software (CDSS)-supported SPRM intervention with standard pharmaceutical care in older patients hospitalised with an acute unselected illness. Over 13 months, we screened 1833 patients aged ≥65 years admitted to specialist services other than geriatric medicine for study inclusion. We randomised 361 patients to the trial intervention arm and 376 patients to the control arm, applying the intervention at a single timepoint within 48 h of admission. The primary endpoint (ADR incidence) was assessed at 7-10 days post-admission or at discharge (whichever came first). The secondary endpoints were the median hospital length of stay (LOS) and hospital mortality rate. Attending clinicians in the intervention group implemented 54.8 % of SPRM/CDSS prescribing recommendations. Ninety-one ADRs occurred in 78 control patients (20.7 %) compared with 61 ADRs in 50 intervention patients (13.9 %), i.e., an absolute risk reduction of 6.8 %. The number needed to treat (NNT) to prevent one patient having one ADR was 15; the total NNT to prevent one ADR was 14. The median LOS and hospital mortality were not significantly different. An SPRM delivered on a CDSS platform significantly reduces ADR incidence in acutely hospitalised older people.