Naslov
Prediction of postoperative vomiting in laparoscopic gynecological surgery

Autori
Šimurina, Tatjana ; Mraović, Boris ; Sonicki, Zdenko

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, znanstveni

Izvornik
British Journal of Anaesthesia, Volume 108 Supplement 2 / - , 2012

Skup
15th World Congress of Anaesthesiologists

Mjesto i datum
Buenos Aires, Argentina, 25.03.2012. - 30.03.2012

Vrsta sudjelovanja
Poster

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
Predictive model ; Postoperative vomiting ; Laparoscopy ; Gynecological surgery

Sažetak
Introduction: Postoperative vomiting (POV) is the second most common postoperative patient complaint. Patients undergoing laparoscopic gynecological surgery have a higher incidence of POV. Assessment of POV risk factors helps clinicians to use appropriate POV prophylaxis. The most used predictive models for POV in clinical practice are Apfel's and Koivuranta's simplified risk scores.(1, 2) Objectives: We analyzed multiple predictive factors for POV in laparoscopic gynecological surgery and proposed a new predictive model for POV. Additionally we compared Apfel's and Koivuranta's risk score with our new score model in this clinical setting. Methods: After obtaining IRB approval and informed consent, 421 women (ASA PS I-II) undergoing laparoscopic gynecological surgery were enrolled in a prospective study. Of these women, 47 were excluded and 374 completed the study. No POV prophylaxis was given. Thiopental was used for induction and isoflurane or sevoflurane for maintenance of general anesthesia. POV and pain scores were measured at 2 and 24 hours postoperatively. Diclofenac and meperidine were used for postoperative pain and metoclopramide for POV. We analyzed 21 patient, 11 anesthesia, and 2 surgery related factors. Multivariate logistic regression was used for predictive modeling. Initially all predictors with p > 0.2 significance and then iterative predictors with p > 0.05 were excluded. All excluded predictors were then individually tested for possible interaction with the final model looking for influence of the predictors' significance on the model for more than 20% of the initial significance.(3) Results: Incidence of POV was 32.3%. Predictive modeling showed 4 predictive factors in the final model: type of surgery (OR = 3.54), history of POV (OR = 1.92), non-smoking (OR = 1.77) and early postoperative pain (OR = 1.033). Our model showed better absolute and relative predictive accuracy (70.86% and 68.97%, respectively) compared with Apfel's (62.03% and 61.16%) and Koivuranta's (66.84% and 54.15%). Also, our model had higher sensitivity and specificity (0.743 and 0.636, respectively) compared with Apfel's (0.636 and 0.586). Koivuranta's model had higher sensitivity (0.901) but poor specificity (0.181). Conclusions. A new predictive model for POV with four predictors (history of PONV, nonsmoking status, early postoperative pain, and type of surgery) compared with two commonly used models was a better predictor for POV in patients undergoing laparoscopic gynecologic surgery. Further validation of our model on a new data set is needed. References 1. Apfel CC, et al. Anesthesiology 1999 ; 91: 693–700. 2. Koivuranta M, et al. Anaesthesia 1997 ; 52: 443–9. 3. Simurina T. 2011, PhD thesis. http://medlib.mef.hr/993.

Izvorni jezik
Engleski

Znanstvena područja
Kliničke medicinske znanosti, Javno zdravstvo i zdravstvena zaštita

POVEZANOST RADA