Page 1 of 1

Chapter 37. Pain Management (continued)

 Subchapter 37.3. Prophylactic Antiemetics During Patient-controlled Analgesia Therapy


Nausea and vomiting are common side effects of patient-controlled analgesia (PCA) with opioids. When severe, nausea and vomiting may limit a patient's tolerance of PCA. Giving antiemetics prophylactically has been suggested as a way to prevent PCA-associated nausea and vomiting.

Practice Description

No study has definitively determined the best prophylactic antiemetic in patient-controlled analgesia. Prophylactic treatments that have been used include droperidol, 5-HT3 receptor antagonists (ondansetron, tropisetron), clonidine, promethazine, hyoscine, propofol and metoclopramide. Prophylactic antiemetics have been given both at the induction of anesthesia and at the end of surgery. A relatively new approach to minimizing nausea and vomiting associated with PCA involves the addition of an antiemetic to PCA, so that the patient receives both analgesic and antiemetic with each PCA demand dose.

Prevalence and Severity of the Target Safety Problem

The incidence of postoperative nausea and vomiting has varied widely (from 8 to 92%) among studies.1 There is no evidence that the incidence of nausea and vomiting with PCA is any different than that with intramuscular opioids. In patients receiving PCA, 57-90% report nausea and 27-40% report vomiting. There appears to be no clear advantage to using one opioid over any others in PCA in terms of postoperative emesis.

Opportunities for Impact

The degree to which prophylactic antiemetics are used in routine practice with PCA is unknown.

Study Designs

Tramer and Walder2 conducted a systematic search for randomized trials (MEDLINE, Embase, Cochrane library, reference lists, hand-searching, no language restriction) that compared prophylactic antiemetics with placebo or no treatment in patients receiving postoperative PCA with opioids. Their review identified 14 placebo-controlled trials with different regimens of droperidol, ondansetron, hyoscine transdermal therapeutic system, tropisetron, metoclopramide, propofol and promethazine. One PCA delivered tramadol; all others delivered morphine.

Both relative risk and number needed to treat were calculated. To estimate the frequency of drug-related adverse effects, the relative risk and the number needed to harm were calculated.

Study Outcomes

The main end point for efficacy was prevention of emetic events (Level 1). The incidence of emetic events with active treatments (experimental event rate) and with placebo or no treatment (control event rate) was extracted. Nausea, vomiting, and "any emetic event" (nausea, vomiting, or nausea and vomiting) were extracted from each trial. Data on drug-related adverse effects (Level 2) were extracted as well.

Evidence for Effectiveness of the Practice

Without antiemetic drugs, the incidence of nausea averaged 43% (range 22-80%), vomiting 55% (45-71%), and any emetic event 67% (54-87%). At 24 hours postoperatively, the cumulative incidence of nausea and vomiting in patients not receiving any antiemetic treatment added to their PCA-morphine was approximately 50%.

The best-studied antiemetic was droperidol. It was added to morphine-PCA in 6 placebo-controlled trials involving 642 patients. Droperidol 0.017-0.17 mg/mg of morphine (0.5-11 mg/d droperidol) was significantly more effective (p=0.04) in preventing nausea than placebo, without evidence of dose-responsiveness. Compared with placebo, the number needed to treat with droperidol to prevent nausea was 2.7 (95% CI: 1.8-5.2) and to prevent vomiting was 3.2 (95% CI: 2.3-4.8).

The second most frequently reported drugs were 5-HT3 receptor antagonists (ondansetron, tropisetron). Their effect on vomiting was satisfactory, with numbers needed to treat of approximately 5 compared with placebo. There was no evidence of any antinausea effect.

Promising results were shown with some of the other interventions (clonidine, promethazine). However, the limited numbers of patients studied did not generate sufficient evidence to make a recommendation.

Potential for Harm

With placebo, the absolute risk of minor adverse events (sedation, drowsiness and dizziness or anxiety, restlessness and agitation) was 0-20%. In those trials that reported adverse events with droperidol, doses of the antiemetic ranged from 1.2 mg to cumulative doses of 7.4 mg. There were no obvious differences in the incidence of minor adverse effects compared with placebo with droperidol doses <4 mg. In all droperidol trials, 2 adverse effect-related study withdrawals were documented. No extrapyramidal symptoms were documented in any trial.

Costs and Implementation

Costs of prophylactic antiemetics during patient-controlled analgesia have not been evaluated. However, 1-4 doses of droperidol per day (0.25-0.5 mL q 4 h) costs $2.83-$11.32 while 1-3 doses of ondansetron per day costs $22.50-$67.50. Implementing prophylactic antiemetics during PCA seems neither difficult nor time-consuming.


Postoperative nausea and vomiting is a common event and patients may refuse to continue PCA because of these side effects. Prophylactic droperidol appears to decrease such side effects. Of 100 patients who have droperidol added to their PCA pump with morphine, 30 who would have vomited or been nauseated had they not received droperidol will not suffer these effects.

The results of this systematic review should be confirmed, as a pooled effect size estimated by meta-analysis must be considered provisional. Additional randomized, placebo-controlled trials assessing droperidol's prophylactic efficacy in the morphine-PCA setting can establish its optimal use in this clinical setting.


1. Woodhouse A, Mather LE. Nausea and vomiting in the postoperative patient-controlled analgesia environment. Anaesth 1997;52:770-775.

2. Tramer MR, Walder B. Efficacy and adverse effects of prophylactic antiemetics during patient-controlled analgesia therapy: a quantitative systematic review. Anesth Analg 1999;88:1354-1361.

Return to Contents

 Subchapter 37.4. Non-pharmacologic Interventions for Postoperative Plan


Over the past decade, increased attention has been paid to non-pharmacologic interventions in conjunction with pharmacologic interventions to treat postoperative pain. The Agency for Health Care Policy and Research's (AHCPR) 1992 Clinical Practice Guideline on Acute Pain Management recommends that preoperative preparation includes educating patients concerning pain control. The Guidelines specifically state that management of postoperative pain may include cognitive-behavioral interventions, such as relaxation, distraction, and imagery.1 Nonetheless, non-pharmacologic interventions can only be considered beneficial if postoperative pain is effectively decreased.

Practice Description

Non-pharmacologic interventions typically fall into 3 categories: healthcare information, skills teaching and psychosocial support.2 Healthcare information includes information on preparation for surgery, timing of procedures, functions and roles of healthcare providers, self-care responsibilities, and pain/discomfort information. Skills teaching includes coughing, breathing and bed exercises, hypnosis, cognitive reappraisal and relaxation exercises. Relaxation strategies include such techniques as the modified Jacobson method, Flaherty and Fitzpatrick jaw relaxation, relaxation tapes, tapes with structured breathing, muscle relaxation and pleasant imagery, and cognitive relaxation. Psychosocial support includes identifying and alleviating concerns, providing reassurance, problem solving with the patient, encouraging questions and increasing frequency of support.

Prevalence and Severity of the Target Safety Problem

The safety problem targeted by the practice of non-pharmacologic interventions is postoperative pain. At least 40-50% of postoperative patients report inadequate pain relief, despite pharmacologic interventions.3 Postoperative pain can have deleterious psychological and physiologic consequences that contribute to patient discomfort and longer recovery periods, and may compromise outcomes.1 It also consumes greater healthcare resources.

Opportunities for Impact

Non-pharmacologic interventions are increasingly used in management of postoperative pain, although the exact proportion of patients receiving such interventions is unknown.

Study Designs

We identified 7 meta-analyses of studies evaluating the effectiveness of non-pharmacologic interventions on postoperative pain. Two of these meta-analyses evaluated various non-pharmacologic interventions for the management of acute pain. Devine2 conducted a meta-analysis of 191 studies looking at psychoeducational care (including all 3 intervention categories above) for adult surgical patients. Sindhu3 conducted a meta-analysis including 49 randomized controlled trials looking at non-pharmacologic nursing interventions (preoperative education/information, relaxation, music, imagery, biofeedback, multidisciplinary approaches and others) for the management of acute pain. AHCPR1 conducted a systematic search of non-drug intervention studies on postoperative pain management. One hundred forty studies were included and formal meta-analysis of 3 was performed. The American Society of Anesthesiologists Task Force on Pain Management, Acute Pain Section4 performed a systematic search for acute pain management in the perioperative setting. Two hundred thirty-three articles were used in the formal meta-analysis; the number of articles regarding education of patients was not reported. Good5 performed a systematic search of trials evaluating the effects of relaxation and music on postoperative pain. Suls and Wan6 performed a systematic search of trials evaluating the effects of sensory and procedural information on coping with stressful medical procedures and pain. Seers and Carroll7 performed a systematic review of various relaxation techniques for postoperative pain relief. Table 37.4.1 lists the 7 articles and briefly describes their salient features.

Study Outcomes

All studies reported postoperative pain (Level 1) as a primary outcome measure. Other outcomes, such as recovery time, psychological distress, and opioid intake were reported in some cases (Level 1).

Evidence for Effectiveness of the Practice

The extensive literature covered by the meta-analyses suggested beneficial effects of psychoeducational care, education and instruction of patients, music and relaxation techniques (Table 39.5.1). The 2 meta-analyses that examined opioid intake failed to show an effect of non-pharmacologic interventions in reducing postoperative opioid consumption.

Potential for Harm

One study7 reported no adverse events in any of the trials for any of the treatment or control groups. Otherwise, adverse consequences of non-pharmacologic interventions on postoperative pain management were not addressed.

Costs and Implementation

Direct information on costs of non-pharmacologic interventions was not reported in these 7 studies. Devine and Cook8 found that the beneficial impact on length of stay and medical complications rendered non-pharmacologic interventions cost-beneficial. Another review2 hypothesized that costs could potentially be decreased by non-pharmacologic interventions, since the length of hospital stay was shortened by an average of 1.5 days (11.5%).

The most obvious direct cost of psychoeducational care is the increased staff time to provide these services. Based on average treatment duration of 42 minutes, a comprehensive version of the intervention would probably not take more than 1 hour per patient.8 Less obvious direct costs might result from staff time to plan the protocol for psychoeducational care and/or to develop patient education materials, in-service programs or staff meetings to teach or review the protocol, printing or purchasing patient education materials, transporting patients to group teaching sessions, and staff time to document the level of care provided.


Effective treatment of postoperative pain continues to be a challenge. In addition to analgesia, non-pharmacologic interventions may provide some benefit in reducing postoperative pain. Clinicians have a wide range of options to consider when developing a comprehensive version of non-pharmacologic care appropriate for their patients. As such interventions are low risk and appeal to many patients, they should be explored in practice and further research.

Table 37.4.1. Studies of non-pharmacologic interventions for postoperative pain

Study setting; Practice ExaminedStudy Design, OutcomesResults*
Adult surgical patients (# not stated), 92% in American hospitals (45% teaching, 43% general hospitals); healthcare information, skills teaching, psychosocial support2Level 1-3 A, Level 1

79-84% of studies found beneficial effects;

Average effect size values were 0.43 for recovery, 0.38 for pain, 0.36 for psychological distress;

Length of stay decreased 11.5%

Not reported; transcutaneous nerve stimulation, education/instruction, relaxation1Level 1-3 A, Level 1Simple/complex relaxation techniques, education/ instruction effective in reducing mild-moderate pain
Not reported; education and participation of patients and families in pain control4Level 1-3 A, Level 1Education improves pain control, reduces adverse outcomes
Adult pts who have undergone torso surgery (# not stated); relaxation techniques, music5Level 1-3 A, Level 1

Total pain decreased in 10 of 13 studies;

Sensory paina reduction was reported in 6 of 12 studies;

Affective painb decreased in 10 of 13 studies;

Unidimensional pain decreased in 4 of 7 studies;

Observed pain decreased in 4 of 4 studies

Not reported; sensory and procedural information (explanations of what will be happening and how patients can expect to feel)6Level 1-3 A, Level 1Combination of procedural and sensory preparation significantly better than control on all measures; effect sizes with combination larger than with either type of information alone
362 patients undergoing fractured hip repair, removal of malignant skin lesions, major elective abdominal surgery, elective cholecystectomy, abdominal hysterectomy, femoral angiography; jaw relaxation, imagery, music, breathing, relaxation tapes7Level 1-3 A, Level 13 of 7 studies showed significant decrease in pain sensation/pain distress in those who had relaxation; 1 out of 5 trials showed significant improvement in psychological outcomes; less anxiety in relaxation group
3387 adult patients; education, relaxation, music, imagery, biofeedback, multidisciplinary approach3Level 1-3 A, Level 1Effect sizes ranged from 2.25 to 1.78; strong heterogeneity

* Effect size is the standardized difference between the control and experimental groups regarding the measure of interest in each study. Positive values indicate benefit with the intervention.
a Sensory pain refers to ability to discriminate where pain is occurring and respond appropriately.
b Affective pain is the sensation of pain as something unpleasant and to be avoided.


1. AHCPR Pain Management Guideline Panel. Clinicians' quick reference guide to postoperative pain management in adults. J Pain Symptom Manage 1992;7:214-228.

2. Devine EC. Effects of psychoeducational care for adult surgical patients: a meta-analysis of 191 studies. Patient Educ Couns 1992;19:129-142.

3. Sindhu F. Are non-pharmacological nursing interventions for the management of pain effective?—a meta-analysis. J Adv Nurs 1996;24:1152-1159.

4. A Report by the American Society of Anesthesiologists Task Force on Pain Management, Acute Pain Section. Practice guidelines for acute pain management in the perioperative setting. Anesth 1995;82:1071-1081.

5. Good M. Effects of relaxation and music on postoperative pain: a review. J Adv Nurs 1996;24:905-914.

6. Suls J, Wan CK. Effects of sensory and procedural information on coping with stressful medical procedures and pain: a meta-analysis. J Consult Clin Psychol 1989;57:372-379.

7. Seers K, Carroll D. Relaxation techniques for acute pain management: a systematic review. J Adv Nurs 1998;27:466-475.

8. Devine EC, Cook TD. Clinical and cost-saving effects of psychoeducational interventions.

Return to Contents
Proceed to Next Chapter

Page last reviewed July 2001
Internet Citation: Chapter 37. Pain Management (continued). July 2001. Agency for Healthcare Research and Quality, Rockville, MD.