Risk factors of postoperative vomiting in children receiving continuous intravenous morphine.

BACKGROUND

Risk factors of postoperative vomiting (POV) have been less extensively explored in children compared to adults. We analyzed the risk factors of POV in children receiving continuous intravenous (i.v.) morphine in a standardized manner without POV prophylaxis after major surgery.

METHODS

This observational retrospective study included 235 children aged from 2 to 216 months (91 F:144 M, 11.5% <6 months, 31.5% 6-11 months). The primary end point was the occurrence of at least one episode of POV recorded on the nursing chart. The independent predictors of POV were determined by univariate analysis followed by a multivariate analysis by logistic regression. The data are presented as either medians (25th-75th percentile) or as values with a 95% confidence interval.

RESULTS

Continuous i.v. morphine was administered over 42 (22-60) h with an initial infusion rate of 20 µg x kg(-1) x h(-1) in 63% of cases, which was increased in 31.5% of cases and was accompanied by an additional bolus in 39.2% of children. At least one episode of POV occurred in 22.6% of children. The following three independent factors were associated with POV: female gender (OR 3.324 [1.695-6.519], P=0.0005), urological surgery (OR 5.605 [1.291-24.340], P=0.0214) and age (OR 1.012 [1.006-1.018], P<0.0001). The discriminating characteristics of the model were good with an ROC curve AUC of 0.778, sensitivity of 71.7% and specificity of 71.4% for a 0.22 cut-off value of POV incidence. The positive predictive value was 42.2%, and the negative predictive value was 89.6%.

CONCLUSION

Female gender, which is usually considered a risk factor after puberty, should be taken into account independent of age to guide the POV prophylaxis in children receiving a postoperative continuous i.v. morphine infusion.