Encinosa William E., Hellinger Fred J.
Both authors are affiliated with the Agency for Healthcare Research and Quality, Center for Delivery, Organization and Markets
To estimate the impact of potentially preventable adverse events on health care costs and outcomes. We used inpatient, outpatient, and drug claims data for elderly Medicare enrollees with secondary employer coverage from 41 large firms located throughout the Nation in 1999 and 2000. These enrollees underwent 22,477 major surgeries in 1,725 hospitals. The Patient Safety Indicators (2003) of the Agency for Healthcare Research and Quality (AHRQ) were used to identify 14 types of potentially preventable adverse events among the major surgeries. We then conducted multivariate regression analyses—controlling for market characteristics, hospital characteristics, and the patient's risk of adverse outcomes—to predict the expenditures attributable to the potentially preventable adverse medical event, and to predict the probability of death, readmission, and long-term care use after such an event. The average difference in total 90-day expenditures between those who had a potentially preventable adverse medical event and those who did not was $35,617. We estimate that 20 percent ($6,998) of this difference was attributable to the actual adverse event. Patients who experienced a potentially preventable adverse medical event had 52 percent higher inpatient hospital expenditures ( < 0.001), 21 percent higher inpatient physician expenditures ( < 0.001), and 11 percent higher outpatient expenditures ( < 0.05) attributable to the event. Medicare paid 86.8 percent of the 90-day costs for patients with patient safety events, compared to 82.9 percent for those without events. The employer paid 11.4 percent of the total 90-day costs for those with patient safety events, compared to 15.0 percent for those without events. Thus, we see that Medicare paid a greater proportion of the extra costs due to the patient safety event. Patients who experienced potentially preventable adverse medical events were 64 percent more likely to use long-term care ( < 0.001) and were 2.8 times more likely to die within 90 days than those without events ( < 0.001). The death rate attributable to potentially preventable adverse medical events was 4.5 percent. Patients who did not die were 30 percent more likely to be readmitted ( < 0.05) within 90 days if they had a potentially preventable adverse medical event. Extrapolating our findings to the Nation, we calculate that the 14 potentially preventable adverse medical events identified using AHRQ's Patient Safety Indicator methodology were responsible for 1.1 percent of readmissions, 9.4 percent of deaths, and 1.6 percent ($224 million) of the total 90-day expenditures for all elderly Medicare major surgery patients with secondary employer coverage in 2000.