Abstract 260: Risk Factors for Readmission from Medical Cardiology Services. A Medication Reconciliation Perspective.

Abstract

BACKGROUND: Hospital readmissions for cardiac diseases have a strong socio-economic impact. Medication reconciliation is one of the current interventions intended to improve quality of care, in which pharmacists play an important role. We aimed to determine risk factors and identify patients who are at risk of re-hospitalization by gathering data from the medication reconciliation process. METHODS: We included adults from our local community admitted to a medical cardiology service from September 2011 to August 2012. Two pharmacists led interviews within 48 hours of presentation. In addition to a rigorous medication reconciliation evaluation at admission and discharge, we assessed medication handling habits, adherence, familiarity with prescriptions, health literacy and reading impairment. We determined Potential Adverse Drug Events (PADE) by assessing discrepancies within medication records. Multivariate analyses were performed to look for association among factors and outcomes. RESULTS: We included 258 patients, mean age 70.5, 52 % male, 19% received help with medications, 83 % were adherent, 14% reported reading impairment, 40% heart failure (HF). The mean number of prescriptions was 7.5 and the mean Charlson comorbidity index 3.5. Medication errors reached 39% of patients on admission and 20% at discharge; no clinical complications were detected. The number of prescriptions was associated with PADEs at admission (p=0.0001) and discharge (p=0.03). Presence of PADE on admission or discharge did not affect the readmission rate. At 30 days 18% of 258 patients were readmitted, 9% visited the Emergency Department (ED) and 2.3% died; meanwhile at 6 months 38% were hospitalized, 46% went to the ED and 5.8% died. We identified a diagnosis of HF and the need for help handling medications at home as the strongest predictors of readmission at both 1 and 6 months. Other factors associated with higher readmissions rates and ED visits were the number of prescriptions, number of hospitalizations within the prior year, lack of familiarity with medical regimen and reading impairment. HF was associated with lower adherence (p=0.025); more admissions in the previous year (p=0.002), higher number of medications (p=0.001), reading impairment (p=0.02), higher readmissions at 1 and 6 months (p=0.003 and 0.0008 respectively) as well as ED visits (p=0.016) at 6 months compared with non-HF subjects. CONCLUSION: We found several characteristics, key examples being presence of HF and receipt of help with medications, that may warrant formal assessment during medication reconciliation as they seem associated with re-hospitalization. These patients may benefit from close follow-up during and post hospitalization.