Alternative shift models and the duality of patient care - An empirical study in surgical intensive care units

On 1 January 1996, the German Arbeitszeitgesetz (working-time regulation) c ame into effect for hospital physicians. It states that working hours must not exceed 8 h per day, even for physician in hospitals. As a consequence, the prevalent two-shift model is legally inadmissible. The intention of thi s law is to protect the physician and to create better conditions for the p atients. However, a systematic evaluation of the postulated benefits is sti ll lacking. Aim: The aim of our study was to analyze the influence of the l ength of daily working hours on the quality of patient care by measuring th e outcome of patients in intensive care units (ICUs), comparing the two-shi ft model (2-SM) - two 12-h shifts - with the three-shift model (3-SM) - thr ee 8-h shifts, ,Materials and methods: In a prospective multicenter study, we compared the outcome of patients in six ICUs (organized by surgeons) wit h different models of working hours. The health status of each patient and the course on ICU [described by hospital mortality, number of complications , readmission to the ICU, reinterventions, duration of the stay in an ICU, and hospital, the course of the Acute Physiology mid Chronic Health Evaluat ion (APACHE) II score] were uni- and multivariately analyzed. In addition, the technical and personnel resources of the ICUs and the hospitals were do cumented. Result: Three hundred and forty seven patients (103 2-SM, 244 3-S M) were included. The epidemiological and the health status on admission to the ICU were comparable. Patients in the 3-SM stayed 1.6 days longer on IC U and 2.3 days longer in the hospital than the 2-SM patients. The frequency of complications, reinterventions, and readmissions to ICU was higher in t he 3-SM. The median nf the APACHE-II score decreased more for 2-SM than for 3-SM patients. This means a significantly quicker recovery of the patients in 2 SM (P <0.05). The multivariate analysis with individual outcome measu res as dependent variables revealed a significant positive effect of the 2- SM on the physicians' assessment of postoperative course, on the relative f requency of therapeutic procedures, and to a lesser extent on the duration of stay in the ICU.