The incidence and severity of adverse events affecting patients after discharge from the hospital

Background: Studies of hospitalized patients identify safety as a significant problem, but few data are available regarding injuries occurring after discharge. Patients may be vulnerable during this transition period. Objective: To describe the incidence, severity, preventability, and "ameliorability" of adverse events affecting patients after discharge from the hospital and to develop strategies for improving patient safety during this interval. Design: Prospective cohort study. Setting: A tertiary care academic hospital. Patients: 400 consecutive patients discharged home from the general medical service. Measurements: The three main outcomes were adverse events, defined as injuries occurring as a result of medical management; preventable adverse events, defined as adverse events judged to have been caused by an error; and ameliorable adverse events, defined as adverse events whose severity could have been decreased. Posthospital course was determined by performing a medical record review and a structured telephone interview approximately 3 weeks after each patient's discharge. Outcomes were determined by independent physician reviews. Results: Seventy-six patients had adverse events after discharge (19% [95% Cl, 15% to 23%]). Of these, 23 had preventable adverse events (6% [Cl, 4% to 9%]) and 24 held ameliorable adverse events (6% [Cl, 4% to 9%]). Three percent of injuries were serious laboratory abnormalities, 65% were symptoms, 30% were symptoms associated with a nonpermanent disability, and 3% were permanent disabilities. Adverse drug events were the most common type of adverse event (66% [Cl, 55% to 76%]), followed by procedure-related injuries (17% [Cl, 86% to 26%]). Of the 25 adverse events resulting in at least a nonpermanent disability, 12 were preventable (48% [Cl, 28% to 68%]) and 6 were ameliorable (24% [Cl, 7% to 41%]). Conclusion: Adverse events occurred frequently in the peridischarge period, and many could potentially have been prevented or ameliorated with simple strategies.