Anesth Analg 2005;100:87-93
© 2005 International Anesthesia Research Society
doi: 10.1213/01.ANE.0000140782.04973.D9
AMBULATORY ANESTHESIA
Paul F. White Section Editor
A Comparison of Patients’ and Health Care Professionals’ Preferences for Symptoms During Immediate Postoperative Recovery and the Management of Postoperative Nausea and Vomiting
Anna Lee, MPH, PhD,
Tony Gin, MBChB, MD, FANZCA, FRCA,
Angel S. C. Lau, BSN(Hons), Dip Epid Biostat, RN, and
Floria F. Ng, BASc, RN
Department of Anaesthesia and Intensive Care, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, New Territories, Hong Kong, China
Address correspondence to Anna Lee, MPH, PhD, Department of Anaesthesia and Intensive Care, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, NT, Hong Kong, China. Address e-mail to annalee{at}cuhk.edu.hk Reprints will not be available.
 |
Abstract
|
|---|
In this study we sought to examine the differences in patients’ and health care professionals’ preferences for symptoms during immediate postoperative recovery and the management of postoperative nausea and vomiting (PONV). The key differences between symptoms during immediate postoperative recovery (PONV, sedation, and pain) and management of PONV (prophylaxis, efficacy of antiemetic, and extra cost) were used to develop 14 scenarios in a questionnaire. Fifty-two health care professionals (anesthesiologists and recovery room nurses) and 200 women undergoing elective gynecological surgery were recruited (overall response rate, 97%). From patients’ and health care professionals’ perspectives, conjoint analysis showed that the most important attribute for immediate postoperative recovery was a reduction in the risk of PONV. Health care professionals placed more importance on postoperative sedation than patients did. They were more concerned about the cost of the antiemetic to the patient than the patients were themselves. There was no preference for a policy of effective treatment versus routine prophylaxis. This study shows that there were small differences in the importance of pain, sedation, efficacy of the antiemetic, and extra cost of treatment between patients and health care professionals.
IMPLICATIONS: Patients’ and health care professionals’ preferences for recovery and antiemetic treatment were compared. There were small differences between groups in the importance of pain, sedation, efficacy of the antiemetic, and extra cost. There was no preference in either group for a policy of effective treatment versus routine prophylaxis.
|
Introduction
|
|---|
An important component of improving the quality of health care is that patient preferences and expectations are incorporated into clinical-care decisions (1). Previous studies examining patients’ preferences for symptoms during immediate postoperative recovery have used a number of different methods, including rating willingness to pay (WTP) (2–4) or rating virtual scenarios (1,5–7). These previous studies have been conducted in the United States (US) (1–4,7), Canada (6), and Germany (5). However, none of these studies has directly compared patients’ and health care professionals’ preferences. Because anesthesiologists and patients may have differing perspectives on the relative importance of postoperative outcomes, the choice of treatment may be different. In other areas of medicine, there are different treatment preferences between patients and health care professionals (8). This suggests that it cannot be assumed that clinicians are in a position to advise patients about treatment options based on what they would theoretically accept for themselves (8).
This study used a conjoint analysis (CA) to assess subject preferences. CA is one of a number of "stated preference" techniques used to determine individual preferences in hypothetical controlled experimental conditions (9,10). The CA allows both strength of preference and an indirect WTP for individual treatment attributes to be estimated (9–11). Our hypotheses were 1) that there would be differences between patients’ and health care professionals’ ranking of the importance of symptoms during immediate postoperative recovery and 2) that there would be differences between patients’ and health care professionals’ choice of management of postoperative nausea and vomiting (PONV). Therefore, the main objectives of this study were to estimate and compare the relative importance of factors associated with symptoms during immediate postoperative recovery and management of PONV from patients’ and health care professionals’ perspectives. The secondary objective was to examine the patients’ monetary valuation (WTP) for an effective antiemetic treatment.
|
Methods
|
|---|
After approval from the Chinese University Clinical Research Ethics Committee, 200 consecutive women undergoing elective gynecological surgery who were assumed to be at high risk of PONV were recruited into the study (April to August 2002). During the same period, we recruited all health care professionals working in our department (n = 60). Informed con-sent was obtained from all subjects. Patients were interviewed with a standardized questionnaire by a research nurse before surgery. In contrast, a self-administered questionnaire was sent to health care professionals. The questionnaire comprised three main sections. In the first section (Section A), respondents were presented with seven pairwise scenarios to assess their preferences for symptoms during immediate postoperative recovery. In Section B, respondents were randomized to 1 of 3 fictitious risk groups (Group 1, baseline risk of PONV <10%; Group 2, baseline risk of PONV 10%–30%; Group 3, baseline risk of PONV >30%) by computer-generated random numbers and were presented with 7 pairwise scenarios to assess their preferences for management of PONV. Group 1 was described as "low risk," Group 2 as "moderate risk," and Group 3 as "high risk." This was done because we expected the WTP value for reducing PONV to be higher in patients with a history of PONV or motion sickness than in patients without these risk factors. Therefore, the randomization would balance the subjects’ known and unknown risk factors for PONV across groups. In the last section, sociodemographic details were requested. The same pairwise scenarios were used for both patient and health care professional groups.
In CA, several attributes (features) of treatment were selected, and a range of possible values (levels) were defined for each attribute. These were used to create the scenarios. The three attributes assigned to measure subject preferences for symptoms during immediate postoperative recovery were risk of PONV, level of pain, and level of sedation (Section A). The attributes assigned to measure subject preferences for management of PONV were type of regimen, efficacy of the antiemetic, and extra cost to the patient (Section B). The levels chosen for each of the attributes and their coding are detailed in Table 1. The levels and definition were chosen by the authors to reflect a realistic occurrence of symptoms during the immediate recovery period and important aspects of an antiemetic management plan in the local setting. Examples of pairwise scenarios for symptoms during the immediate postoperative recovery and management of PONV are shown in Appendix 1 and Appendix 2, respectively.
The SPSS Version 11 (SPSS Inc., Chicago, IL) procedure Orthoplan was used to reduce the number of scenarios to a manageable level while still leaving the ability to infer utilities (relative importance) for all possible scenarios. The technique resulted in an orthogonal design and gave 8 scenarios from the original 12. The orthogonal design of scenarios ensured that there was no multicollinearity between the independent variables (attributes). Therefore, the effects of each attribute and attribute level were well balanced so that only main effects were measured, and interactions were expected to be negligible. Of the eight scenarios, one scenario was chosen to reflect the current situation, and each of the remaining scenarios was compared with this scenario. To avoid potential bias in scenario ordering, the seven scenarios were randomly ordered.
A random-effects probit model was used to analyze the data because of the repeated measurements of the data (multiple responses from the same respondent). Separate random-effects probit models were estimated for patients and health care professional groups.
To assess preferences for symptoms during immediate postoperative recovery, the general model was specified as follows:
|
where 
B is the change in benefit in moving from the current situation to the alternative, PONV is the difference in the risk of PONV, PAIN is the difference in the level of pain, and SEDATION is the difference in the level of sedation. The unobservable error terms are represented by e and u, where e is the error term due to differences among observations and u is the error term due to differences between subjects. The estimated coefficients for each attribute indicate the utility, a measure of preference. The utility measure not only expresses which attribute is most preferred, but also gives a measure of importance. To estimate the relative importance, the range of utility was obtained within each attribute and then summed with all the other attributes. Then the importance weight for a single attribute was computed as the range of utility divided by the sum of ranges for all the attributes.
To assess preference for management of PONV, the general model was specified as follows:
where B is the change in benefit in moving from the current management of PONV to an alternative, the explanatory variables are the difference in the attributes as defined in Table 1, and ßj (j = 1, 2, 3) are the coefficients of the model to be estimated. The unobservable error terms are represented by e and u, where e is the error term due to differences among observations and u is the error term due to differences between subjects. From these coefficients, the utility, a measure of preference, can be estimated, with associated P statistics to indicate whether the attribute had a statistically significant effect on the subject’s choice of management. The relative importance of the attributes was estimated with the method outlined above. This model was stratified by the three baseline risk groups (low, moderate, and high) in the analysis.
To estimate the indirect WTP for an improvement in an individual attribute, the coefficient of interest was divided by the coefficient attached to cost. For example, to estimate WTP for a more effective antiemetic, this is calculated by ß2/ß3 from Equation 2.
Equations 1 and 2 did not have a constant term. This is because subjects were told to assume that all aspects of the proposed management, other than those specified in the questionnaire, were identical between the treatments being compared.
Possible confounding factors, such as age, education, employment, income, smoking status, history of motion sickness, and history of PONV, were incorporated in Equations 1 and 2. Dummy variable interaction terms were created between cost and income levels. The Wald statistic was calculated to test for statistically significant differences on the coefficients according to income level. STATA Version 7.0 software (Stata Corporation, College Station, TX) was used for the analysis. Mean ± SD and 95% confidence intervals (95% CI) are reported. The level of significance was set at P < 0.05.
|
Results
|
|---|
The response rates in patients and health care professionals were 100% and 87%, respectively. The health care professional group included 18 specialist anesthesiologists, 22 anesthesiology trainees, and 12 recovery room nurses. Sociodemographic data and PONV risk factors of the two groups are shown in Table 2. The health care professional group was younger, and most did not smoke, compared with the patient group (Table 2). There was no significant difference between groups for prevalence of history of PONV or history of motion sickness.
Immediate Postoperative Recovery
Patients’ Preferences.
The results of the random effects probit regression model are shown in Table 3. Sedation was not a significant attribute in influencing preferences for symptoms during immediate postoperative recovery. The negative and significant coefficients for PONV and postoperative pain suggest that patients would be less satisfied with a higher risk of PONV and slight pain, as expected. The relative importance of PONV, pain, and sedation were 74%, 23%, and 3%, respectively. Because no significant sociodemographic factors were associated with patients’ preferences, these factors were left out in the final model (Table 3). The model correctly predicted 89% (95% CI, 88%–91%) of patients’ choices.
Health Care Professionals’ Preferences.
The level of pain was not a significant attribute in influencing preferences for symptoms during immediate postoperative recovery (Table 4). The relative importance of PONV, pain, and sedation was 76%, 5%, and 19%, respectively. PONV was considered more important than the level of sedation after adjusting for significant differences in preferences among health care professional groups (
2 = 26.86; df = 2; P < 0.001). The model correctly predicted 88% (95% CI, 85%–91%) of health care professionals’ choices.
View this table:
[in this window]
[in a new window]
|
Table 4. Results from a Random-Effects Probit Model for Assessing Health Care Professionals’ Preferences (52 Subjects)
|
|
Antiemetic Management
Patients’ Preferences.
The efficacy of the antiemetic and the extra cost to the patient were significant attributes for the management of PONV. Whether prophylaxis was given was not an important factor in this decision (Table 3). The relative importance of the type of antiemetic management, efficacy, and extra cost was 5%, 50%, and 45%, respectively. No significant sociodemographic or PONV risk factors were associated with patients’ preferences. The interaction between cost and income level was not significant (P = 0.77) and was not included in the final model. The mean incremental WTP for a more effective antiemetic was Hong Kong (HK)$227 (95% CI, $169–$311; equivalent to US$29: 95% CI, $22–$40).
Health Care Professionals’ Preferences.
The significant attributes were efficacy of the antiemetic and extra cost to the patient (Table 4). The relative importance of the type of antiemetic management, efficacy, and extra cost was 6%, 37%, and 57%, respectively. No significant sociodemographic or PONV risk factors were associated with health care professionals’ preferences. The mean incremental WTP for a more effective antiemetic was HK$129 (95% CI, $59–$274; equivalent to US$17: 95% CI, $8–$35).
|
Discussion
|
|---|
The results of patients’ preferences for symptoms during immediate postoperative recovery were consistent with those from another study that used CA (5). Avoiding PONV and reducing pain were significant factors in influencing patients’ choices for postoperative recovery. This study also showed that from both the patient’s and anesthesiologist’s perspective, avoiding PONV is one of the most important clinical outcomes, supporting previous studies (1,5,6,12).
Interestingly, health care professionals rated the level of sedation to be more important than the level of postoperative pain. In this study, postoperative pain levels were defined as "no pain" and "slight pain." We believe that health care professionals would have placed more importance on pain if the pain levels were defined as "no pain" and "moderate pain." Our results suggest that our health care professionals perceived the possible danger of sedation in the postanesthesia care unit to be more important than the level of pain. Alternatively, there was an underappreciation of postoperative pain assessment. A previous study showed that nurses and anesthesiologists underestimated the high levels of postoperative pain indicated by patients (13). In the study of Macario et al. (12), incisional pain was ranked considerably higher than sedation as a common and important outcome to avoid by a panel of anesthesiologists.
Both patients and health care professionals agreed that improved efficacy of an antiemetic and extra costs were significant factors in an optimal antiemetic regimen. The relative importance of the antiemetic efficacy and extra cost were rated similarly by patients. In contrast, health care professionals believed that the extra cost for patients’ antiemetic management was more important than the antiemetic efficacy. There was no evidence that sociodemographic or PONV risk factors were significantly associated with subjects’ choice of antiemetic regimen. This supports a previous finding in which patients who had experienced nausea ranked nausea as an outcome to avoid similarly to those who had not had experienced nausea (1).
There was no preference by patients or health care professionals for routine prophylaxis over symptomatic treatment. This may be due to the infrequent prevalence of a history of PONV in our subjects. In another study, there was no clinically significant difference in the level of patient satisfaction between routine prophylaxis and treatment of PONV (14). The estimated number needed to treat to increase the level of patient satisfaction with control of PONV associated with routine prophylaxis with ondansetron 4 mg was 25, or approximately US$400 (14). Therefore, our study suggests that a highly effective and cost-effective antiemetic regimen to treat PONV would be acceptable as routine prophylaxis.
The WTP value in this study was similar to that reported by Macario et al. (1), with patients allocating US$30 out of US$100 to avoid PONV. However, it was less than that reported by Gan et al. (2), in which patients were willing to pay between US$56 to US$100 for a totally effective antiemetic. The differences between the studies may be related to differences in study methods. First, instead of using a hypothetical antiemetic that would totally prevent PONV in the questionnaire (2), we defined moderate and high efficacy as a reduction in the risk of PONV by 50% and 75%, respectively—a more realistic scenario. Second, we administered the questionnaire to patients with the assistance of a research nurse before surgery instead of in the postanesthesia care unit (2). However, a recent study on patient preferences for acute pain treatment suggests that patient preferences are stable over time (7).
A limitation in this study was that the patients knew that the extra cost (HK$0, $100, or $200) was a virtual amount of money and that they were not charged regardless of their answer. In the Hong Kong health care system, patients do pay for hospital charges and pay extra for patient-choice items. Therefore, our patients have a reference point and are not "willing" to pay huge amounts of money even if they know that it is a virtual amount of money. This may explain why our WTP estimates are less than those of other studies (2,4). The use of virtual scenarios is a common practice when CA is performed for policy and economic evaluations, because the researcher has complete control over the experimental design: this ensures robustness of the results (15). CA is a rigorous method of eliciting patients’ preferences (10). The WTP concept is increasingly being used and is based on the premise that the maximum amount of money an individual is willing to pay is an indication of his or her satisfaction with a health care intervention (16). A multimodal antiemetic prophylaxis for high-risk patients was associated with a high WTP compared with a treatment-of-PONV regimen in low- to moderate-risk patients (17). Another limitation was that we did not examine the comparative WTP value to avoid postoperative pain and sedation with PONV. A recent study showed that patients were willing to pay US$50, $34, $33, and $20 to avoid postoperative pain, intraoperative awareness, PONV, and sedation, respectively (4).
There was minimal selection bias in this study, because we surveyed women from a wide range of ages, incomes, and education levels, as well as recovery nurses and anesthesiologists with different levels of experience. However, there are some issues to consider. The results of this study are limited to women undergoing elective gynecological surgery with a frequent incidence of a history of motion sickness and risk of PONV. Therefore, the results may not be generalizable to a larger patient population with less risk of PONV. The results would not be applicable to patients with severe comorbidities or those undergoing major surgery, because mortality would be an important outcome to consider. To overcome some of these limitations, a multicenter study involving a diverse patient population that indicates patient preferences about PONV and treatment should be undertaken both before and after surgery.
In contrast to previous patient preference studies in anesthesia (1–7), this study is the first to compare patients’ and health care professionals’ preferences directly. There were small differences in preferences. Patients want to avoid PONV and pain and ranked the efficacy and cost of the antiemetic equally. In contrast, from the health care professionals’ perspective, the important factors were avoidance of PONV and sedation. They were more concerned about the cost of the antiemetic to the patient than were the patients themselves. There was no preference for a policy of effective treatment versus routine prophylaxis in either group. Recognition of these small differences is an important step to developing a consensus of appropriate treatment delivery to improve the quality of anesthesia care. Inclusion of patient preferences in consensus guidelines is poor, and has been identified as an important area to be addressed (18). Furthermore, whether anesthesiologists can improve the quality of anesthesia care by tailoring anesthetic regimens to individual patient’s needs requires further study (1).
In conclusion, there were small differences in preferences for symptoms during recovery and antiemetic management between patients and health care professionals. Both groups indicated strong preferences for a reduction in the risk of PONV and believed that patients were willing to pay between US$17 and US$29 out of their own pocket for a more effective antiemetic regimen.
|
Acknowledgments
|
|---|
Supported by a grant from the Research Grants Council of the Hong Kong Special Administrative Region, China (Project CUHK 4004/01M).
|
References
|
|---|
- Macario A, Weinger M, Carney S, Kim A. Which clinical anesthesia outcomes are important to avoid? The perspective of patients. Anesth Analg 1999; 89: 652–8.[Abstract/Free Full Text]
- Gan T, Sloan F, Dear GL, et al. How much are patients willing to pay to avoid postoperative nausea and vomiting? Anesth Analg 2001; 92: 393–400.[Abstract/Free Full Text]
- Engloren M, Steffel C. Patient perception of monetary value to avoiding unpleasant side effects of anesthesia and surgery. J Clin Anesth 2000; 12: 388–91.[Medline]
- Gan TJ, Ing RJ, de L Dear G, et al. How much are patients willing to pay to avoid intraoperative awareness? J Clin Anesth 2003; 15: 108–12.[Medline]
- Eberhart LH, Morin AM, Wulf H, Geldner G. Patient preferences for immediate postoperative recovery. Br J Anaesth 2002; 89: 760–1.[Abstract/Free Full Text]
- Jenkins K, Grady D, Wong J, et al. Post-operative recovery: day surgery patients’ preferences. Br J Anaesth 2001; 86: 272–4.[Abstract/Free Full Text]
- Gan TJ, Lubarsky DA, Flood EM, et al. Patient preferences for acute pain treatment. Br J Anaesth 2004; 92: 681–8.[Abstract/Free Full Text]
- Montgomery AA, Fahey T. How do patients’ treatment preferences compare with those of clinicians? Qual Health Care 2001; 10 (Suppl 1): i39–43.[Abstract/Free Full Text]
- Ratcliffe J, Van Haselen R, Buxton M, et al. Assessing patients’ preferences for characteristics associated with homeopathic and conventional treatment of asthma: a conjoint analysis study. Thorax 2002; 57: 503–8.[Abstract/Free Full Text]
- Ryan M, Farrar S. Using conjoint analysis to elicit preferences for health care. BMJ 2000; 320: 1530–3.[Free Full Text]
- Aristides M, Chen J, Schulz M, et al. Conjoint analysis of a new chemotherapy: willingness to pay and preference for the features of raltitrexed versus standard therapy in advanced colorectal cancer. Pharmacoeconomics 2002; 20: 775–84.[Medline]
- Macario A, Weinger M, Truong P, Lee M. Which clinical anesthesia outcomes are both common and important to avoid? The perspective of a panel of expert anesthesiologists. Anesth Analg 1999; 88: 1085–91.[Abstract/Free Full Text]
- Sjostrom B, Haljamae H, Dahlgren LO, Lindstrom B. Assessment of postoperative pain: impact of clinical experience and professional role. Acta Anaesthesiol Scand 1997; 41: 339–44.[Medline]
- Scuderi PE, James RL, Harris L, Mims GR III. Antiemetic prophylaxis does not improve outcomes after outpatient surgery when compared to symptomatic treatment. Anesthesiology 1999; 90: 360–71.[ISI][Medline]
- Ubach C, Scott A, French F, et al. What do hospital consultants value about their jobs? A discrete choice experiment. BMJ 2003; 326: 1432.[Abstract/Free Full Text]
- McIntosh E, Donaldson C, Ryan M. Recent advances in the methods of cost-benefit analysis in healthcare: matching the art to the science. Pharmacoeconomics 1999; 15: 357–67.[ISI][Medline]
- Eberhart LH, Mauch M, Morin AM, et al. Impact of a multimodal anti-emetic prophylaxis on patient satisfaction in high-risk patients for postoperative nausea and vomiting. Anaesthesia 2002; 57: 1022–7.[ISI][Medline]
- Shaneyfelt TM, Mayo-Smith MF, Rothwangl J. Are guidelines following guidelines? The methodological quality of clinical practice guidelines in the peer-reviewed medical literature. JAMA 1999; 281: 1900–5.[Abstract/Free Full Text]
Accepted for publication July 14, 2004.
Top
Abstract
Introduction
