SCIENCE@home

The Incidence and Severity of Adverse Events Affecting Patients after Discharge from the Hospital

A. Forster,H. Murff,J. Peterson,T. Gandhi,D. Bates

Published 2003 in Annals of Internal Medicine

ABSTRACT

Context Research on patient safety has focused on adverse events in hospitalized patients. Adverse events following hospitalization have received little attention. Contribution Among 400 consecutive patients discharged from the general medical service of an academic hospital, 76 had adverse events during the 2 weeks following hospital discharge. Of the adverse events, the researchers categorized 23 as preventable, 24 as ameliorable, and the remainder as neither. Adverse drug events were the most frequent occurrence. Implications Interventions to maximize patient safety should consider the vulnerable transition from hospital to home. The Editors Patient safety is a public concern that has received substantial attention, especially since the release of two reports from the U.S. Institute of Medicine. The first of these, To Err Is Human, reviewed the literature on adverse events, or injuries resulting from medical care (1). On the basis of data from two large population-based chart review studies, the report estimated that adverse events occur in 2.9% (2) to 3.7% (3) of hospitalizations. The first of these studies (2) found that although most injuries are minor, approximately 1 in 10 results in death. The researchers determined that approximately half of the adverse events were caused by errors (1, 3, 4). The second report, Crossing the Quality Chasm, which proposes strategies for improving the health system, declared patient safety a fundamental component of care quality (5). However, the estimated incidence of adverse events quoted in To Err Is Human may underestimate the overall safety problem, since injuries occurring after discharge were not included in the evaluation. Patients may be especially vulnerable to injuries during this period because they may still have functional impairments and because discontinuities may occur at the interface of acute and ambulatory care (6). These discontinuities may be worsened by the current health care environment, in which patients are leaving the hospital quicker and sicker (7) and may receive care from hospitalists rather than their primary care physicians (8). Despite these theoretical concerns, few studies are available to estimate the extent of the problem. Several studies have determined the risk for postdischarge wound infections in surgical patients (9, 10), but this is a specific type of safety problem and the risk cannot be generalized. Other studies have used readmission rates to identify poor-quality hospital care (11). However, this outcome is an insensitive and nonspecific marker of quality in general (11) and safety in particular. Information on the incidence and type of adverse events is vital for improving postdischarge safety. We wanted to determine the incidence and severity of adverse events affecting patients after discharge from the hospital to home. In addition, we wanted to identify the incidence of preventable and ameliorable adverse events, classify types of adverse events, and identify system improvements to reduce the incidence and severity of adverse events after discharge. Methods Overview To determine the rate of adverse events, we first created a case summary of every patient's posthospital course by performing a medical record review and a telephone interview approximately 3 weeks after discharge. Using this information, we created event summaries, which two board-certified internists independently reviewed to determine whether medical management caused an injury and, if so, whether it was preventable or ameliorable. The institutional review board at the study hospital approved the protocol. Definitions Adverse outcomes were any of the following patient experiences after discharge: new or worsening symptoms, unanticipated visits to health facilities for tests or treatments, or death. An adverse event was an injury resulting from medical management rather than the underlying disease. We evaluated all injuries resulting in symptoms after discharge, regardless of when the events occurred. Thus, we included adverse events that occurred in the hospital and after discharge as long as symptoms persisted until the patient went home. A preventable adverse event was an injury that could have been avoided, that is, an injury judged to probably be the result of an error or a system design flaw. An ameliorable adverse event was an injury whose severity could have been substantially reduced if different actions or procedures had been performed or followed. Patient Sample We used a prospective cohort design to study adult medical patients at an 800-bed urban academic teaching hospital. Patients were included if they went home from the general medical service during a sequential 81-day period, had telephone service, and could complete a telephone interview in English. Patient surrogates were permitted to complete the interview if the patient could not. Telephone Interview A board-certified internist contacted patients by telephone within 2 to 5 weeks of discharge. If the patient was eligible and consented, he or she underwent a semi-structured interview over the telephone. If the patient was not available, we made up to 20 attempts to contact him or her at different times of the day and week. If the patient declined to participate or we were unable to contact him or her after 5 weeks, we rated the patient as a nonresponder. Responders underwent a telephone interview that had several components (Appendix Figure). The principal component assessed the patient's condition since discharge by using a full review of organ systems. If patients answered yes to any question, the interviewer asked them to elaborate about the symptom's severity, timing in relation to hospitalization and treatments, and resolution. The survey also determined patients' use of health services since discharge, including home care services, physician services, visits to laboratories, and readmissions to the hospital. Preparation of Case Summaries To create a case summary, the internist combined the information from the telephone interview with information obtained from a review of the patient's electronic medical record. This computerized clinical record contains handover notes from the hospitalization; discharge summaries; previous orders and written instructions; emergency department and clinic notes; operative and procedure notes; and all laboratory results, including results of radiography, electrocardiography, and pathologic examination. If necessary, the paper chart was also reviewed to clarify information that was in the discharge summary or handover notes. If patients returned to the study hospital after discharge, we were able to validate their history by reviewing their records. However, for patients receiving care elsewhere, we had to rely on firsthand accounts. Preparation of Event Summaries The internist reviewed every patient's case summary for adverse outcomes. If any were found, an event summary that consisted of a detailed description of each outcome, including its onset, its severity, the health services used during its treatment, and its resolution, was created for each one. If there were no adverse outcomes, then the event summary consisted of the case summary. Determination of Adverse Events Two other board-certified internists independently rated each event summary by using standard techniques as described elsewhere (3). Each event summary was evaluated separately. Reviewers first rated on a scale of 1 to 6 their confidence that medical management caused an injury. If their rating was 5 or 6, indicating that the injury was probably or definitely caused by management, the event was considered an adverse event. For adverse events, the internists used implicit criteria to determine their preventability and ameliorability. The two reviewers' assessments were then compared for causality, preventability, and ameliorability. If there was disagreement, the two reviewers discussed the case to attempt to come to consensus. If they did not agree, a third board-certified internist rated the event independently. Next, reviewers rated injury severity, health services used, type of injury, and location. Injury severity was categorized as serious laboratory abnormality only, one day of symptoms, several days of symptoms, nonpermanent disability, permanent disability, or death. Differentiation between several days of symptoms and nonpermanent disability relied on evidence that symptoms interfered with patients' activities of daily living. We categorized health services used as none, additional visit to a physician, additional visit for laboratory testing in addition to a physician visit, visit to an emergency department, readmission to the hospital, or death. We were able to determine whether health services were additional because during the interview we asked whether the health service was arranged before discharge or specifically for the particular symptom. We used a standard approach to classify the type of injury (2, 3), categorizing each as an adverse drug event, a procedure-related injury, a nosocomial infection, a fall, or other. We also classified the location of adverse events. Our primary objective was to evaluate adverse events affecting patients after discharge from the hospital to home. We felt that most of these events would occur after the patient went home, but we recognized a priori that some adverse events could occur in the hospital but lead to morbidity after discharge because of the severity of the injury or because health providers did not recognize the problem at discharge. Examples of injuries occurring in the hospital but affecting patients after discharge include pressure ulcers, missed diagnoses, and erroneous drug prescriptions. If it was determined that the adverse event occurred in the hospital, we examined whether the injury was recognized before discharge and, if so, whether definitive management was instituted at that time. To develop strategies to

PUBLICATION RECORD

CITATION MAP

EXTRACTION MAP

CLAIMS

  • No claims are published for this paper.

CONCEPTS

  • No concepts are published for this paper.

REFERENCES

Showing 1-23 of 23 references · Page 1 of 1

CITED BY

Showing 1-100 of 1956 citing papers · Page 1 of 20

123420
Next