Table of Contents:
Postoperative nausea and vomiting is a frequent anaesthesia-related complication that should not be neglected in terms of patient satisfaction as well as physical and economic consequences. In order to optimise antiemetic prophylaxis, identification of the patient-specific risk profile is essential.
In the present study, 2347 patients were examined for the dependence of the frequency of PONV on the type of operation, while eliminating further influencing factors as accurately as possible.
In order to calculate a PONV prognosis, an average PONV risk within the intervention groups was first inferred using the Apfel and Koivuranta scores. Drug prophylaxis was calculated into these risks using additive relative risk reduction. Thus three calculations with risk reductions of 30%, 25% and 20% per antiemetic administered were performed for the PONV frequency values of both risk scores. This resulted in PONV predictions for each surgical group that included risk factors as well as the protective effect of antiemetic prophylaxis. The postoperative PONV incidence was then analysed using confidence intervals with focus on deviations from these PONV prediction values and the PONV incidence in the overall collective.
A concordance of PONV prognosis and actual incidence in the overall population as reference group was shown in the calculation with a relative risk reduction of 30% based on both risk scores and with a relative risk reduction of 25% based on the Koivuranta score. Here there were significant deviations in the collectives of cholecystectomies and total and open hysterectomies, whose PONV incidence was higher than predicted. For abdominal surgery, the incidence of PONV was higher than the calculated risk only when calculating a prognosis with a relative risk reduction of 30%. In gynaecological patients in total and laparoscopically operated gynaecological patients, as well as in the laparoscopic procedure as a whole, the incidence was exclusively higher than the PONV prognosis calculated on the basis of the Koivuranta score with a risk reduction of 30%.
When comparing the actual incidences with those of the total collective (23.7%), there were lower incidences of PONV in the groups of urological patients (13.9%), breast surgery (17.0%) and in ENT procedures (17.4%). The incidence rates were higher for orthopaedic (29.0%) and gynaecological procedures (27.5%), cholecystectomies (35.9%), hysterectomies (42.6%), abdominal surgery (34.0%) and laparoscopies overall (28.8%) and in the gynaecological collective (32.1%). The laparoscopic procedure was associated with a higher incidence of PONV than the open procedure; for hysterectomies and gynaecological procedures, however, there was no difference between the two techniques.
Despite these findings, the intensification of antiemetic prophylaxis should not be considered in principle for the procedures identified as risk-associated because of the confounding factors that could not be eliminated. Some of these influencing factors were the younger age structure, especially in laparoscopic procedures, the varying duration of surgery, the dose dependence of the PONV increase of opioids and volatile anaesthetics, the limited informative value of risk scores as well as the generalised equal and additive efficacy of all antiemetic methods. Furthermore, a heterogeneity of the operations included in the individual subgroups and the different perioperative settings with varying intervention spectra in the participating hospitals was evident, which made it difficult to draw conclusions about the risk profile of specific types of operations.
Nevertheless, hysterectomies and cholecystectomies showed a sufficiently robust deviation of the incidence from the lower calculated prognosis as well as a more frequent need for antiemetic therapy compared to the overall population. In combination with other risk-increasing factors, an adjustment of prophylaxis should be considered.