Secondary Logo

Journal Logo

Original Article

Survival analysis of postoperative nausea and vomiting in patients receiving patient-controlled epidural analgesia

Lee, Shang-Yia; Hung, Chih-Jena,b; Chen, Chih-Chieha; Wu, Chih-Chenga,c,*

Author Information
Journal of the Chinese Medical Association: November 2014 - Volume 77 - Issue 11 - p 589-593
doi: 10.1016/j.jcma.2014.08.008

    Abstract

    1. Introduction

    Postoperative nausea and vomiting and postoperative pain are two major concerns in patients receiving surgeries and anesthetics.1 Frequently, postoperative nausea and vomiting occurs in patients receiving general, regional, or local anesthesia and causes significant suffering.2–4 Sometimes, patients prefer to endure postoperative pain rather than receive opioids that may result in postoperative nausea and vomiting,5 and may even voluntarily pay more to obtain effective antiemetic therapies.6

    Previous studies attempted to investigate the incidence and risk factors of postoperative nausea and vomiting using different predictive models.7–10 However, the incidences of postoperative nausea and vomiting explored in previous studies ranged from 10% to 30% in general anesthesia2,11–14 and from 3.2% to 34% in patients receiving patient-controlled epidural analgesia.15–21 This disparity may be attributable to the different time points used when the patients were visited. Postoperative nausea and vomiting could occur several minutes, hours, or even days after anesthesia. If the investigation of postoperative nausea and vomiting is limited to a specified time point, the analysis will be confounded.

    In this study, we analyzed the information obtained by repeatedly visiting the patients to explore the risk factors and incidence of postoperative nausea and vomiting, and assessed the effects of time points, which have rarely been discussed in the literature. When considering the effects of time, survival analysis was applied to analyze the occurrence of postoperative nausea and vomiting. We compared the results obtained by logistic regression to those determined by survival analysis to elucidate the differences between the two statistical methods. This research may contribute to a clearer understanding of the incidence of postoperative nausea and vomiting.

    2. Methods

    In this retrospective, observational, uni-institutional (a medical center in central Taiwan) study, we analyzed the results of 229 patients receiving postsurgical patient-controlled epidural analgesia from June 2007 to December 2007. Ethical approval was obtained from the Institutional Review Board of Taichung Veterans General Hospital. The data were reviewed for patients receiving general anesthesia for the following surgeries: (1) upper abdomen surgery; (2) lower abdomen surgery; and (3) chest surgery. The composition prescription in patient-controlled epidural analgesia consists of bupivacaine (0.1%) and fentanyl (1.5 μg/mL) in normal saline (500 mL). We collected data including age, sex, body mass index, types of surgery, the setting of patient-controlled analgesia pumps, total dosage (mL) and average dosage (mL/hour) of patient-controlled epidural analgesia, and the time point of postoperative nausea and vomiting occurrence.

    In our institution, patients were visited five times over a 3-day period of patient-controlled epidural analgesia use, and the five time points were as follows: (1) the first visit on operation day (OPD-1); (2) the second visit on operation day (OPD-2); (3) the third visit on the 1st postoperative day (POD-1-1); (4) the fourth visit on the 1st postoperative day (POD-1-2); and (5) the fifth and final visit on the 2nd postoperative day (POD-2). If the patient had used patient-controlled epidural analgesia on the first visit, the zero time point was backtracked by calculating the average dosage on the first visit.

    When patients were receiving patient-controlled epidural analgesia, we followed up the patients and adjusted the depth of the epidural catheter or the dosage of medications according to the patients' pain intensity at rest and in motion after surgeries. Our goal was to keep the patients' pain intensity below 3 at rest and below 5 in motion by using the numeric rating scale (where 0 = no pain and 10 = most intense pain imaginable). When analgesia in the required dermatome failed or was found to be inadequate, the type of pain management would be changed to other regimens such as intravenous patient-controlled analgesia. Consequently, these patients were not included in this study.

    IBM SPSS Statistics 20 (IBM SPSS Inc. Chicago, IL, USA) was used to analyze these data. The Cox proportional hazards model (Cox model) was used to determine the correlations between variables and survival time to calculate the odds ratio of the risk factors. When patients suffered from postoperative nausea and vomiting in the following periods, it was defined as the occurrence of an event. Survival time was outlined as the time duration until the occurrence of postoperative nausea and vomiting. If an event did not occur in the following period, then the patient was assumed to be censored. We also applied logistic regression to analyze the correlations between variables and the incidences of postoperative nausea and vomiting at specific time points, and compared the results with those obtained using the Cox model. Both in the Cox model and logistic regression, the forward likelihood ratio was used to identify significant variables. A p value <0.05 was considered significant.

    3. Results

    A total of 229 patients received patient-controlled epidural analgesia for postoperative pain control between June 2007 and December 2007. The demographic data of the patients are shown in Table 1. The settings of patient-controlled analgesia pumps were as follows: bolus dosage, 4.38 ± 0.94 mL; continuous infusion dosage, 4.25 ± 1.18 mL; lockout time, 17.9 ± 4.6 minutes; and 4-hour upper limit dosage, 46.8 ± 11.25 mL. In patients with and without postoperative nausea and vomiting, the average total dosage of patient-controlled analgesia was 106.5 ± 91.73 mL versus 279.2 ± 98.85 mL and the duration of patient-controlled analgesia use was 16 ± 14.4 hours versus 46.7 ± 6 hours, respectively. The overall incidence of postoperative nausea and vomiting was 35.4% (Table 1), and ranged from 5.5% to 17.6% at five different time points (Table 2). The highest incidence of postoperative nausea and vomiting was noted at the third visit on postoperative Day 1 (POD-1-1); the lowest was found at OPD-1 (Table 2).

    T1-8
    Table 1:
    Demographic data.
    T2-8
    Table 2:
    Incidences of postoperative nausea and vomiting at different time points.

    Before logistic regression was applied to explore the risk factors for incidence of postoperative nausea and vomiting in patients with patient-controlled epidural analgesia, univariate analysis of candidate variables was performed. These results, as provided in Table 3, revealed that age, sex, height, and surgical sites may play significant roles in predicting the occurrence of postoperative nausea and vomiting in POD-1-1, POD-1-2, or POD-2.

    T3-8
    Table 3:
    Univariate analysis of variables predicting the occurrence of postoperative nausea and vomiting by logistic regression.

    At three time points (POD-1-1, POD-1-2, and POD-2), sex was found to have a significant influence on the occurrence of postoperative nausea and vomiting (Table 4). Female patients appeared to be more likely (3.013, 2.790, and 3.015 times) to suffer from postoperative nausea and vomiting than male patients at these three time points. In addition, the surgical sites of the patients were noted to have an effect on the occurrence of postoperative nausea and vomiting at POD-1-2 and POD-2 (Table 4). Patients undergoing surgeries in the upper abdomen may have a higher risk of developing postoperative nausea and vomiting than those receiving surgeries in the chest, and the odds ratios were 4.490 and 4.834, respectively, at the above mentioned time points.

    T4-8
    Table 4:
    Risk factors of postoperative nausea and vomiting in patients with patient-controlled epidural analgesia analyzed by logistic regression.

    Cox regression was used to analyze the risk factors of postoperative nausea and vomiting incidence in patients with patient-controlled epidural analgesia based on the time to occurrence of events (Table 5). The result in step 1 shows that the greater the total dosage of opioids, the lower the risk of postoperative nausea and vomiting with a hazard ratio of 0.988. In the final steps, a higher average dose of opioids consumed was correlated with a higher incidence of postoperative nausea and vomiting, with a hazard ratio of 1.286. Furthermore, female patients were more likely (a hazard ratio of 2.501) to suffer from postoperative nausea and vomiting than male patients, and patients with a higher body mass index were at a higher risk of suffering from postoperative nausea and vomiting, with a hazard ratio of 1.103.

    T5-8
    Table 5:
    Risk factors of postoperative nausea and vomiting in patients with patient-controlled epidural analgesia analyzed by Cox regression.

    4. Discussion

    In this study, the overall incidence of postoperative nausea and vomiting was 35.4%, which was higher than the results found by previous studies.2,11–14 The incidence of postoperative nausea and vomiting varied with time and ranged from 5.5% to 17.6% at five different time points (Table 2). We explored the total occurrence of events through five time points, which would make the investigation more comprehensive than by only following up the patients at a single time point, and thus a higher incidence of postoperative nausea and vomiting was obtained. The disparity between the incidences of postoperative nausea and vomiting at five time points may reflect the inconsistent results from previous investigations owing to the different visiting time points used in each study.2,11–14

    Results of analyses by logistic regression showed that female patients were more likely (approximately 3 times more) to suffer from postoperative nausea and vomiting than male patients at three time points (Table 4). The total doses and the average dose of opioids were not found to have a significant effect on the occurrence of postoperative nausea and vomiting. The results support the notion that the risk factors of postoperative nausea and vomiting identified by logistic regression could be affected by different time points and may therefore result in conflicting findings.7–10,22

    Although other statistical techniques such as generalized estimating equation or mixed model could be applied to analyze the data collected repeatedly, our focus was on the occurrence of postoperative nausea and vomiting and the duration until the occurrence of the first event. Therefore, we applied the Cox proportional hazards model to analyze the factors with respect to time to occurrence of postoperative nausea and vomiting. The results showed that in addition to sex, which was found by logistic regression, body mass index and the average and total dose of opioids were also found to play significant roles in the occurrence of postoperative nausea and vomiting (Table 5).

    Several studies have shown that consumption of postoperative opioids was one of the risk factors related to the occurrence of postoperative nausea and vomiting23,24; however, other studies did not support this conclusion.7,25,26 Very few studies have investigated the relationship between the dosage of opioids and the occurrence of postoperative nausea and vomiting, factoring in the time effects.27 The inconsistent findings may be attributable to the limitations of logistic regression and the use of different visiting time points. The results showed that patients requiring larger doses of opioids were less likely to suffer from postoperative nausea and vomiting (Table 5), which may be attributable to the time factor. Therefore, the average dose of opioids instead of the total dosage of opioids consumed by patients could account for this phenomenon.

    The above results indicate that investigations of risk factors and the incidence of postoperative nausea and vomiting may be influenced by the visiting time points and the use of statistical methods. Analysis of the data to predict postoperative nausea and vomiting by logistic regression focuses on the finding of the risk factors at a definite time point. In this study, we found that the risk factors and incidence of postoperative nausea and vomiting investigated by logistic regression changed between different time points. Using the predictive models established by logistic regression, we may suggest to a patient that there is no need for prophylaxis for postoperative nausea and vomiting at a specific time point such as the 1st postoperative day. However, the patient may experience postoperative nausea and vomiting on the 2nd postoperative day because the condition of the predictive models changed.

    If we hope to survey the risk factors of postoperative nausea and vomiting during the entire course of follow up, and not only at any one time point, application of the Cox regression model may be more optimal than the conventional logistic regression. By using this approach, we may achieve a more comprehensive prophylaxis of the postoperative nausea and vomiting. Cox regression has seldom been used to investigate the occurrence of postoperative nausea and vomiting, but it has presented a different viewpoint that did not conflict with the results found in previous studies analyzed with logistic regression.23,28

    The average consumption of opioids was proven to be an important factor in the occurrence of postoperative nausea and vomiting during the following course of treatment, and this finding may further support the use of multimodal pain control. Use of gabapentin or nonsteroidal anti-inflammatory drugs would reduce the consumption of opioids and may lead to a decreased occurrence or severity of postoperative nausea and vomiting.29,30 There is reason to believe that these strategies would offer better postoperative care.

    Standl et al3 and Pusch et al4 discovered that patients receiving general anesthesia were more likely to suffer from postoperative nausea and vomiting than patients receiving regional anesthesia. Wilhelm et al14 concluded that the sites and types of surgeries may influence the incidence of postoperative nausea and vomiting. In order to decrease the effect of confounding factors, we surveyed the patients receiving general anesthesia for just three types of surgery: upper abdomen, lower abdomen, and chest surgeries. Patients with different surgeries may suffer from various intensities of postoperative pain, but they would optimize their postoperative analgesia by using epidural patient-controlled analgesia.31,32 Therefore, the occurrence of postoperative nausea and vomiting may be analyzed more discriminatively and predictably by investigating the individualized doses of analgesia regimen consumed than by exploring the effects of different surgeries. Most studies23,24 that sought to predict the occurrence of postoperative nausea and vomiting comprised only the types of surgeries involved. They frequently lacked information about the detailed surgical methods that would make the analysis results relevant not only for a specific surgery, but also more broadly applicable to more generalized procedures.

    The present study was a retrospective investigation, and the patient-controlled epidural analgesia data were collected from the patients' medical records. Previous studies revealed various risk factors for postoperative nausea and vomiting, including a history of smoking, motion sickness, postoperative nausea, and vomiting7–10 which could not be collected by reviewing the patient charts in the present study. Another limitation in the present retrospective study was that we did not include in the analyses the opioids used in anesthesia and the rescue or extra opioids administered in the ward.

    In conclusion, a wide variation of the risk factors and the incidence of postoperative nausea and vomiting may exist because patients were assessed at different time points, coupled with the overall limitation of statistical methods. Visiting patients at different time points and using Cox regression that could factor in the time effects would allow researchers to explore the investigation of postoperative nausea and vomiting more comprehensively. Multimodal pain management and prophylaxis treatment of postoperative nausea and vomiting could offer better postoperative care.

    Acknowledgments

    Our article has been revised by Mr Peter Wilds, who obtained the degree of Combined Studies (English and Psychology) from the University of Newcastle-upon-Tyne (UK) and who is currently a language editor at Taichung Veterans General Hospital.

    References

    1. Chandrakantan A, Glass PS. Multimodal therapies for postoperative nausea and vomiting, and pain. Br J Anaesth. 2011;107:i27-40.
    2. Watcha MF, White PF. Postoperative nausea and vomiting. Its etiology, treatment, and prevention. Anesthesiology. 1992;77:162-184.
    3. Standl T, Eckert S, Schulteam Esch J. Postoperative complaints after spinal and thiopentone–isoflurane anaesthesia in patients undergoing orthopaedic surgery. Spinal versus general anaesthesia. Acta Anaesthesiol Scand. 1996;40:222-226.
    4. Pusch F, Freitag H, Weinstabl C, Obwegeser R, Huber E, Wildling E. Single-injection paravertebral block compared to general anaesthesia in breast surgery. Acta Anaesthesiol Scand. 1999;43:770-774.
    5. van Wijk MG, Smalhout B. A postoperative analysis of the patient's view of anaesthesia in a Netherlands' teaching hospital. Anaesthesia. 1990;45:679-682.
    6. Gan T, Sloan F, Dear Gde L, El-Moalem HE, Lubarsky DA. How much are patients willing to pay to avoid postoperative nausea and vomiting? Anesth Analg. 2001;92:393-400.
    7. Apfel CC, Kranke P, Eberhart LH, Roos A, Roewer N. Comparison of predictive models for postoperative nausea and vomiting. Br J Anaesth. 2002;88:234-240.
    8. Eberhart LH, Hogel J, Seeling W, Staack AM, Geldner G, Georgieff M. Evaluation of three risk scores to predict postoperative nausea and vomiting. Acta Anaesthesiol Scand. 2000;44:480-488.
    9. Thomas R, Jones NA, Strike P. The value of risks scores for predicting postoperative nausea and vomiting when used to compare patient group in a randomised controlled trial. Anaesthesia. 2002;57:1119-1128.
    10. Tramer MR. A rational approach to the control of postoperative nausea and vomiting: evidence from systematic reviews: Part I. Efficacy and harm of antiemetic interventions, and methodological issues. Acta Anaesthesiol Scand. 2001;45:4-13.
    11. Cohen MM, Duncan PG, DeBoer DP, Tweed WA. The postoperative interview: assessing risk factors for nausea and vomiting. Anesth Analg. 1994;78:7-16.
    12. Lerman J. Surgical and patient factors involved in postoperative nausea and vomiting. Br J Anaesth. 1992;69:S24-32.
    13. Quinn AC, Brown JH, Wallace PG, Asbury AJ. Studies in postoperative sequelae. Nausea and vomiting—still a problem. Anaesthesia. 1994;49:62-65.
    14. Wilhelm SM, Dehoorne-Smith ML, Kale-Pradhan PB. Prevention of postoperative nausea and vomiting. Ann Pharmacother. 2007;41:68-78.
    15. Tsui SL, Irwin MG, Wong CM, Fung SK, Hui TW, Ng KF, et al. An audit of the safety of an acute pain service. Anaesthesia. 1997;52:1042-1047.
    16. Stenseth R, Sellevold O, Breivik H. Epidural morphine for postoperative pain: experience with 1085 patients. Acta Anaesthesiol Scand. 1985;29:148-156.
    17. Lo PH, Chiou CS, Tsou MY, Chan KH, Chang KY. Factors associated with vomiting in orthopedic patients receiving patient-controlled epidural analgesia. Acta Anaesthesiol Taiwan. 2008;46:25-29.
    18. Flisberg P, Rudin A, Linner R, Lundberg CJ. Pain relief and safety after major surgery. A prospective study of epidural and intravenous analgesia in 2696 patients. Acta Anaesthesiol Scand. 2003;47:457-465.
    19. de Leon-Casasola OA, Parker B, Lema MJ, Harrison P, Massey J. Postoperative epidural bupivacaine–morphine therapy. Experience with 4,227 surgical cancer patients. Anesthesiology. 1994;81:368-375.
    20. Chen YJ, Chang KY, Tsou MY, Lin SP, Chan KH, Ting CK. Risk factors of vomiting among females on patient-controlled epidural analgesia. J Chin Med Assoc. 2009;72:183-187.
    21. Dolin SJ, Cashman JN. Tolerability of acute postoperative pain management: nausea, vomiting, sedation, pruritus, and urinary retention. Evidence from published data. Br J Anaesth. 2005;95:584-591.
    22. Apfel CC, Kranke P, Greim CA, Roewer N. What can be expected from risk scores for predicting postoperative nausea and vomiting? Br J Anaesth. 2001;86:822-827.
    23. Apfel CC, Korttila K, Abdalla M, Kerger H, Turan A, Vedder I, et al. A factorial trial of six interventions for the prevention of postoperative nausea and vomiting. N Engl J Med. 2004;350:2441-2451.
    24. Apfel CC, Kranke P, Eberhart LH. Comparison of surgical site and patient's history with a simplified risk score for the prediction of postoperative nausea and vomiting. Anaesthesia. 2004;59:1078-1082.
    25. Koivuranta M, Laara E, Snare L, Alahuhta S. A survey of postoperative nausea and vomiting. Anaesthesia. 1997;52:443-449.
    26. Sinclair DR, Chung F, Mezei G. Can postoperative nausea and vomiting be predicted? Anesthesiology. 1999;91:109-118.
    27. Roberts GW, Bekker TB, Carlsen HH, Moffatt CH, Slattery PJ, McClure AF. Postoperative nausea and vomiting are strongly influenced by postoperative opioid use in a dose-related manner. Anesth Analg. 2005;101:1343-1348.
    28. Apfel CC, Läärä E, Koivuranta M, Greim CA, Roewer N. A simplified risk score for predicting postoperative nausea and vomiting. Anesthesiology. 1999;91:693-700.
    29. Marret E, Kurdi O, Zufferey P, Bonnet F. Effects of nonsteroidal antiinflammatory drugs on patient-controlled analgesia morphine side effects: meta-analysis of randomized controlled trials. Anesthesiology. 2005;102:1249-1260.
    30. Mathiesen O, Dahl B, Thomsen BA, Kitter B, Sonne N, Dahl JB, et al. A comprehensive multimodal pain treatment reduces opioid consumption after multilevel spine surgery. Eur Spine J. 2013;22:2089-2096.
    31. Hu KH, Tsou MY, Chan KH, Chang KY. An investigation on influential factors of patient-controlled epidural analgesic requirement over time for upper abdominal surgeries. J Chin Med Assoc. 2013;76:446-451.
    32. Teng WN, Su YP, Kuo IT, Lin SM, Tsou MY, Chan KH, et al. Patient controlled epidural analgesia for bilateral versus unilateral total knee arthroplasty: a retrospective study of pain control. J Chin Med Assoc. 2012;75:114-120.
    Keywords:

    epidural analgesia; logistic models; patient-controlled analgesia; postoperative nausea and vomiting; survival analysis

    © 2014 by Lippincott Williams & Wilkins, Inc.