Abstract
STUDY OBJECTIVES: Intensive care unit nurses commonly work multiple consecutive 12-hour shifts that leave little time for sleep between work shifts. Working multiple consecutive shifts could compromise vigilance and patient care, especially with respect to managing high-risk medications such as insulin infusions. We hypothesized that as the number of consecutive shifts worked by nurses increases, the rate of hypoglycemia in patients who are receiving an insulin infusion would also increase. METHODS: We identified patients who had hypoglycemia (glucose ≤ 3.5 mmol/L, 63 mg/dL) between December 2008 and December 2009 in 3 intensive care units in Vancouver, British Columbia, Canada. For each hypoglycemic event, we counted the number of shifts worked on consecutive days during the previous 72 hours by the bedside nurse who was caring for the patient at the time of hypoglycemia (case shift). For each case shift, we identified up to 3 control shifts (24, 48, and 72 hours before the hypoglycemic event in the same patient when there were no hypoglycemic events) and counted the number of consecutive shifts worked by those nurses in the previous 72 hours. This analysis allowed us to control for patient-associated confounders. Conditional logistic regression was used to determine the association between number of consecutive shifts worked and occurrence of hypoglycemic events. RESULTS: A total of 282 hypoglycemic events were identified in 259 patients. For 191 events, we were able to identify 1 or more control shifts. Compared with nurses who had not worked a shift in the preceding day, the odds ratio of a hypoglycemic event was 1.68 (95% confidence interval: 1.12-2.52), 2.16 (95% confidence interval:1.25-3.73), and 2.54 (95% confidence interval: 1.28-5.06) for nurses who were working their second, third, or fourth consecutive shift, respectively. CONCLUSIONS: Working multiple consecutive nursing shifts is associated with increased risk of hypoglycemic events in patients in an intensive care unit.