Previous Abstract | Next Abstract
Printable Version
October 21, 2008
9:00 AM - 11:00 AM
Room Hall E2-Area A,
Implementing a Prediction Rule for Postoperative Nausea and Vomiting: A Cluster-Randomized Trial
Teus H. Kappen, M.D., Yvonne Vergouwe, Ph.D., Wilton A. van Klei, M.D., Ph.D., Leo van Wolfswinkel, M.D., Ph.D., Cor J. Kalkman, M.D., Ph.D.
Perioperative Care and Emergency Medicine, University Medical Center, Utrecht, Netherlands

Clinical prediction rules may guide risk tailored prophylaxis for postoperative nausea and vomiting (PONV). Several rules have been developed to predict the risk of PONV. However, the effects of implementation of such rules on physician behavior and, more importantly, on patient outcome, have not been studied. We performed a cluster-randomized trial to quantify the effects of implementation of a previously developed and validated prediction rule for PONV. (Van den Bosch JE et al. Anesth Analg. 2005 May;100(5):1525-32.)


Attending anesthesiologists and senior residents were randomized to the implementation intervention (n = 41) or to usual care (n = 40). In the intervention group the calculated PONV risk was presented continuously on-screen in the anesthesia record keeping system during the operation. Anesthesiologists were free to act upon the presented risk and to select preventive anti-emetic treatments.

Eligible patients were patients for elective, non-cardiac surgery (n = 12,242). PONV was assessed up to 24 hours postoperatively. Differences in PONV incidence and the number of preventive anti-emetic treatments were analyzed with mixed effects regression analysis to take into account the multilevel structure of the data.


After complete case analysis, the PONV incidence was 46% (2,228/4,840) in the control group and 44% (1,886/4,271) in the intervention group (p=0.08). The implementation of the rule led to a significant increase in the number of preventive anti-emetic treatments (0.66 versus 0.73 per patient, p<0.01), particularly in high-risk patients (1.1 versus 1.7 in patients with a predicted risk > 65%). Results of the regression analysis are shown in figure 1.[figure1][figure2]Conclusion

Implementation of a prediction rule to guide risk tailored PONV prophylaxis changes physician behavior, resulting in an increase in preventive anti-emetic treatments. However, the intervention did not result in improved patient outcome, i.e. in a decrease in PONV incidence.

Anesthesiology 2008; 109 A1266
Figure 1

Figure 2