Date:
July 7, 2006
Promotors:
Prof. C.J. Kalkman, MD, PhD
Prof. K.G.M. Moons, PhD
Co-promotores:
W.A. van Klei, MD, PhD
Y. Vergouwe, MD, PhD
About 30 to 50 percent of patients who undergo surgery suffer from nausea or vomiting (PONV) after the operation. These side-effects of anesthesia are unpleasant for patients. Therefore, prevention of PONV is desirable. This may be achieved by selective administration of prophylactic anti-emetics or use of less emetogenic anesthetic agents with patients at high risk of PONV. Such high-risk patients can be identified preoperatively using a prediction model for PONV. Such models predict the risk of PONV based on characteristics such as gender, age, smoking, nausea/vomiting after previous surgery, motion sickness and the type of surgery.
Previously, two prediction models for PONV have been developed in Finland and Germany. The research described in this thesis was conducted to test whether these models predict the risk of PONV accurately in other patient groups than the patient groups in which these models were developed (development samples). The predictive accuracy of these models was tested in a group of about 1400 Dutch inpatients who underwent surgery in the Academic Medical Center (AMC) in Amsterdam, The Netherlands. In this group of patients, the predictive accuracy of the models was less than was expected based on the results from the development samples and from previous validation studies. Therefore, a new model was developed using the data from the Dutch validation sample. The predictive accuracy of this new model was subsequently tested in a group of about 1200 inpatients who were operated in the University Medical Center in Utrecht (UMCU). We also tested whether the three available prediction models, developed in inpatients, predict the risk of PONV accurately in outpatients.