Unexplained deaths in a children's hospital. An epidemiologic assessment.

During a nine-month period, July 1980 through March 1981, the mortality rate for patients on the cardiology ward of a children's hospital was 43.1 deaths per 10,000 patient-days, as compared with 11.0 deaths per 10,000 patient-days during the preceding 54 months. Twenty-five (76 per cent) of 33 infant deaths during this nine-month period occurred between midnight and 6:00 a.m., as compared with 1 of 10 infant deaths during a separate 27-month period (P less than 0.001). Although nearly all deaths occurred in patients with serious congenital heart disease, epidemic-period deaths were more likely to have an unexpected timing and a clinical pattern consistent with digoxin toxicity. In four patients, forensic and clinical digoxin measurements suggested that an intravenous overdose of digoxin had been administered shortly before death. Although a review of nursing schedules revealed a strong association (relative risk, 64.6) between infant deaths and the duty times of a particular nurse, the cause of the epidemic remains unclear. The study led to suggestions that the hospital strengthen central control over procedures for dispensing medicines and implement a system for monitoring the occurrence of deaths by time and place within the hospital.