Multi-Year Retrospective Analysis of Mortality and Readmissions Correlated with STOPP/START and the American Geriatric Society Beers Criteria Applied to Calgary Hospital Admissions

Abstract

Introduction: The goals of this retrospective cohort study of 129,443 persons admitted to Calgary acute care hospitals from 2013 to 2021 were to ascertain correlations of “potentially inappropriate medications” (PIMs), “potential prescribing omissions” (PPOs), and other risk factors with readmissions and mortality. Methods: Processing and analysis codes were built in Oracle Database 19c (PL/SQL), R, and Excel. Results: The percentage of patients dying during their hospital stay rose from 3.03% during the first admission to 7.2% during the sixth admission. The percentage of patients dying within 6 months of discharge rose from 9.4% after the first admission to 24.9% after the sixth admission. Odds ratios were adjusted for age, gender, and comorbidities, and for readmission, they were the post-admission number of medications (1.16; 1.12–1.12), STOPP PIMs (1.16; 1.15–1.16), AGS Beers PIMs (1.11; 1.11–1.11), and START omissions not corrected with a prescription (1.39; 1.35–1.42). The odds ratios for readmissions for the second to thirty-ninth admission were consistently higher if START PPOs were not corrected for the second (1.41; 1.36–1.46), third (1.41;1.35–1.48), fourth (1.35; 1.28–1.44), fifth (1.38; 1.28–1.49), sixth (1.47; 1.34–1.62), and seventh admission to thirty-ninth admission (1.23; 1.14–1.34). The odds ratios for mortality were post-admission number of medications (1.04; 1.04–1.05), STOPP PIMs (0.99; 0.96–1.00), AGS Beers PIMs (1.08; 1.07–1.08), and START omissions not corrected with a prescription (1.56; 1.50–1.63). START omissions for all admissions corrected with a prescription by a hospital physician correlated with a dramatic reduction in mortality (0.51; 0.49–0.53) within six months of discharge. This was also true for the second (0.52; 0.50–0.55), fourth (0.56; 0.52–0.61), fifth (0.63; 0.57–0.68), sixth (0.68; 0.61–0.76), and seventh admission to thirty-ninth admission (0.71; 0.65–0.78). Conclusions: “Potential prescribing omissions” (PPOs) consisted mostly of needed cardiac medications. These omissions occurred before the first admission of this cohort, and many persisted through their readmissions and discharges. Therefore, these omissions should be corrected in the community before admission by family physicians, in the hospital by hospital physicians, and if they continue after discharge by teams of family physicians, pharmacists, and nurses. These community teams should also meet with patients and focus on patients’ understanding of their illnesses, medications, PPOs, and ability for self-care.