White, Paul F. PhD, MD, FANZCA; Joshi, Girish P. MB BS, MD, FFARCSI; Carpenter, Randall L. MD; Fragen, Robert J. MD
One of the most important factors in the successful management of outpatients undergoing ambulatory surgery is the ability to control post-operative pain . Adequate pain control facilitates early ambulation and improves the patients' perception of the quality of their ambulatory experience. On the other hand, inadequate pain relief delays ambulation, prolongs the discharge time, and can lead to unplanned hospital admissions. Furthermore, uncontrolled pain may increase the incidence of postoperative nausea and vomiting (PONV), as well as delay discharge after ambulatory surgery.
Oral preparations containing combinations of opioids and nonsteroidal antiinflammatory drugs (NSAIDs) are the most commonly used analgesics after ambulatory surgery. However, the use of opioid analgesics is associated with an increased incidence of side effects, including nausea, vomiting, dizziness, sedation, ileus, and respiratory depression. Ketorolac is an NSAID with analgesic and antiinflammatory properties. Early clinical trials reported that the analgesic efficacy of ketorolac was comparable to opioid analgesics in the relief of moderate to severe postoperative pain . However, the efficacy of analgesic drugs may be dependent on the type of ambulatory surgery. Furthermore, in a clinical setting, it is important to evaluate the analgesic efficacy of multiple doses of analgesic medications [3,4].
Therefore, a multicenter, double-blind, placebocontrolled, multiple-dose study was designed to compare the analgesic efficacy and side effects of oral ketorolac with those of hydrocodone plus acetaminophen after ambulatory arthroscopic or laparoscopic tubal ligation procedures.
Two hundred fifty-two healthy consenting outpatients undergoing arthroscopic or laparoscopic tubal ligation procedures were enrolled in this multicenter, double-blind, randomized, multi-dose, parallel, stratified (by surgical procedure and baseline pain) study after obtaining institutional review board approval. Exclusion criteria included allergy or contraindication to the study medications, history of alcohol or drug abuse, chronic use of analgesic or psychotropic drugs, known peptic ulcer disease, bleeding diathesis, renal dysfunction, or inability to complete the pain evaluation questionnaires.
The general anesthetic technique was standardized for all patients. Anesthesia was induced with propofol (2-2.5 mg/kg) and fentanyl (100 micro g) and maintained with isoflurane in combination with 60% nitrous oxide in oxygen. Upon emergence from anesthesia, patients experiencing moderate or severe pain were randomized according to a computer-generated randomization schedule to one of three analgesic treatment groups. Group 1 received one ketorolac 10 mg capsule and one placebo capsule for the single-dose phase of the study and then every 4-6 h as needed, Group 2 received two capsules of hydrocodone 3.75 mg and acetaminophen 375 mg for single-dose phase with the same dose repeated every 4-6 h as needed, and Group 3 received two placebo capsules for the single-dose phase followed by ketorolac 10 mg for the multiple-dose phase. All capsules were identical in appearance and were dispensed by a pharmacist who was not involved with the patient evaluations. All patients were administered rescue analgesic medications (intravenous fentanyl in the postanesthesia care unit and oral acetaminophen with codeine after discharge) if the study medication did not provide adequate pain relief.
An observer unaware of the patients' group recorded their pain using a 4-point categorical verbal rating scale (0 = none, 1 = mild, 2 = moderate, 3 = severe). Patients recorded their pain intensity values on a standardized 100-mm visual analog scale ranging from 0 = no pain to 100 = severe pain immediately prior to receiving the first dose of the study medication, at 30 min, and subsequently at every hour for 6 h. At each assessment time point, patients also rated pain relief on a 5-point categorical verbal rating scale (0 = none, 1 = a little, 2 = some, 3 = a lot, 4 = complete). The hourly evaluations were discontinued if the patient required a second dose of the study medications before 6 h had elapsed.
Patients were provided with a diary in which they were asked to record their categorical pain level and pain relief at bedtime for 3 days. In addition, they recorded the number of doses of study medication and any concomitant medication used during the day. At the end of the study period, all patients gave an overall rating of their pain relief (0 = none, 1 = a little, 2 = some, 3 = a lot, 4 = complete), drug tolerability (1 = poor, 2 = fair, 3 = good, 4 = very good, 5 = excellent), and effectiveness of the analgesic medication (1 = poor, 2 = fair, 3 = good, 4 = very good, 5 = excellent). All adverse events that occurred during the study period were recorded. The study was terminated if more than 24 h had elapsed without the patient taking any study medication, if the patient required a rescue analgesic, or at the end of 3 days. The diaries and the study medications were collected from the patients at the time of the initial follow-up visit.
Two indices were derived from the pain intensity scores for the first-dose analyses: 1) pain intensity difference (PID) scores and 2) summated pain intensity difference (SPID) scores. PID scores were calculated for each patient by subtracting the severity pain score after the first dose from the baseline pain severity score. SPID scores at 3 and 6 h (SPID3 and SPID6) were calculated by computing the sum of PID scores during that time weighted by the length of time in hours between PID scores (area under the PID curve). Similarly, the total pain relief (TOTPAR) scores at 3 h and 6 h (TOTPAR3 and TOTPAR6) were obtained from pain relief evaluations by the computed weighted sum of the pain relief scores during those time intervals. Increasing SPID scores indicate increased pain relief, whereas negative values represent pain more intense than what was present before administration of the first dose of the study medication. With respect to TOTPAR analyses, higher TOTPAR values reflect greater pain relief, with no pain relief indicated by a score of zero.
Demographic data, treatment factors, study site, and treatment by site interaction were analyzed using analysis of variance. Data of patients who were remedicated or withdrawn from the study prior to the 1-h efficiency evaluation were considered invalid for the primary statistical efficiency analysis; however, they were included in the premature termination analysis. All patients receiving study medication were included in the safety analysis. The efficacy variables were analyzed by the Cochran-Mantel-Haenzel row mean scores test with modified ridit scoring and were adjusted for the study site. Fisher's exact and likelihood-ratio chi squared tests were used to compare the treatment groups with respect to incidence of adverse events. The Cochran-Mantel-Haenzel tests were used for categorical variables, and Kaplan Meier and log rank methods were used for all time-related variables. P values of less than 0.05 were considered statistically significant.
Sample size calculation was based on the information from the analysis of previous studies of oral analgesic medications for treatment of postoperative pain. Assuming a SD of 2.5 for SPID3 (alpha = 0.05), 35 patients in each treatment group would provide approximately 90% power to detect a difference of 2.0 between the treatment groups and 13% power to detect a difference of 0.5 between the groups. Since the surgical procedures were analyzed separately, power calculations were based on 100-125 patients for each surgical procedure, with a total of approximately 225 patients.
There were no differences in the demographic or clinical data among the three analgesic treatment groups (Table 1). All of the 252 patients enrolled in the study received at least one dose of study medication and were included in the safety analyses. Fifty-five patients (22%) remained in the study for the entire 72 h (ketorolac 18, hydrocodone-acetaminophen 15, and placebo 22). Of the 197 patients (78%) who did not complete the study, 77 required a rescue analgesic (ketorolac 24, hydrocodone-acetaminophen 26, placebo 27), 76 did not require any study medication after the initial dose (ketorolac 31, hydrocodone-acetaminophen 19, and placebo 26), 28 were lost to follow-up (ketorolac 6, hydrocodone-acetaminophen 15, placebo 7), and 16 withdrew because of adverse events (ketorolac 4, hydrocodone-acetaminophen 7, and placebo 5). The overall number of patients requiring no additional study medication after the initial dose was significantly lower in the ketorolac group compared with the hydrocodone-acetaminophen group (P < 0.01).
Of the 242 subjects whose data were considered valid for the efficacy analysis, 65 had undergone arthroscopic surgery, while 177 had had a laparoscopic tubal ligation. The data of 10 subjects were excluded from the efficacy analyses because of protocol violations. The efficacy data for the two surgical populations were analyzed separately. In the patients undergoing arthroscopic surgery, the categorical SPID, visual analog scale SPID, and TOTPAR scores for patients receiving ketorolac or hydrocodone-acetaminophen were significantly improved compared with those in the placebo group (Table 2). Although these scores were also numerically superior in the ketorolac group compared with the hydrocodone-acetaminophen group, these differences were not statistically significant. In the patients undergoing laparoscopic surgery, the three treatment groups showed similar responses to the study medications (Table 2). The daily evaluation of pain severity at the end of Postoperative Days 1, 2, and 3 showed no statistically significant differences between the two analgesics, nor were there significant differences in the pain relief scores.
In the patients undergoing arthroscopic surgery, the peak analgesic effect, as measured by the highest mean PID score achieved during the initial 6-h study period and the time to peak analgesia were significantly improved in the patients receiving ketorolac and hydrocodone-acetaminophen compared with those receiving the placebo (Figure 1 and Figure 2). However, there was no significant difference between the ketorolac and hydrocodone-acetaminophen groups. The peak analgesic effect and the time to reach peak analgesia were similar in all groups undergoing laparoscopic surgery (Figure 3 and Figure 4).
There were no significant differences in drug tolerability among the treatment groups at the end of the first-dose phase (Table 3). At the end of the multidose phase, the ketorolac group, which included patients who received ketorolac throughout the study as well as those who received placebo for the first dose and ketorolac for the subsequent doses, scored higher in tolerability than the hydrocodone-acetaminophen group (Table 3). The incidence of postoperative dizziness was significantly higher in the patients receiving hydrocodone-acetaminophen compared with ketorolac (Table 4). However, there were no significant differences in the incidence of other side effects between the two analgesic treatment groups.
Ketorolac provides effective postoperative analgesia when administered either orally or parenterally . Preliminary studies suggested that oral ketorolac possesses potent analgesic activity in the postoperative period . When ketorolac was administered in an intravenous-oral sequence after ambulatory surgery, it produced fewer side effects than the combination of acetaminophen and codeine . However, the interpretation of these results are limited because patients undergoing a variety of surgical procedures (e.g., laparoscopy, arthroscopy, breast biopsy, and hernia repair) were studied. After oral surgery procedures, Forbes et al.  reported better analgesia and a decreased incidence of side effects with ketorolac, 10 mg per os (PO) compared with a codeine-acetaminophen combination. In the present study, the two oral analgesic treatments were distinguished from placebo in patients undergoing arthroscopic surgery, which indicates that the model used was sensitive to the effects of these drugs.
In the laparoscopic tubal ligation patients, the analgesic treatments were similar to the placebo group. This negative finding may be due to the high incidence of PONV in this patient population. In addition, laparoscopic (versus arthroscopic) surgery may be associated with greater impairment of gastrointestinal function, which thereby reduces tolerance of oral medications in the early postoperative period. Importantly, pain after laparoscopic surgery may be related to multiple factors. Postoperative abdominal pain may arise from the subumbilical and suprapubic entry sites, and shoulder pain may result from the effects of residual gas in the peritoneal cavity. Tubal ligation procedures have been reported to cause greater postoperative pain than diagnostic laparoscopy  as a result of the release of prostaglandins . Although the type of laparoscopic surgery (diagnostic versus tubal ligation) may influence the severity of the postoperative pain , it does not necessarily predict the usefulness of NSAID administration . Finally, NSAIDs appear to exhibit a "ceiling effect" with respect to their postoperative analgesic efficacy . Due to their limited analgesic efficacy after more painful operative procedures, NSAIDs alone may provide inadequate pain relief in the early recovery period. Analogous to the findings in this study, other investigators have also failed to demonstrate a significant opioid-sparing effect with the use of ketorolac after laparoscopic sterilization .
The efficacy of these two oral analgesics were evaluated using single- and multiple-dose models. Multidosing studies may provide more clinically relevant information regarding the optimal dosing schedule  because some analgesics display increased efficacy with repeated administration . Furthermore, multidose studies provide a better evaluation of the side effect profile of analgesic medications compared with single-dose studies .
The results of this study corroborate with previous reports, which suggested that oral ketorolac is well tolerated after single and multiple doses [15-20]. Wong et al.  reported a lower incidence of nausea and somnolence in the patients receiving ketorolac, 10 mg PO, compared with codeine 60 mg plus acetaminophen 600 mg PO. Ketorolac does not depress central respiratory drive  or produce adverse effects on gastrointestinal function . However, anecdotal reports of renal failure associated with ketorolac have become a cause of concern . The renal dysfunction has been temporary and may have been related to other factors (e.g., dehydration, preexisting renal disease, concomitant nephrotoxic drug therapy) [2,23].
One of the deficiencies of this study design is the residual analgesic effects of intraoperatively administered opioids, which could have affected evaluation in the initial dosing phase of the study. However, the study medication was administered only when patients complained of moderate or severe pain, and, therefore, the residual anesthetic or analgesic effects should have been minimized. Another limitation of the study design was related to the rigid dosing of analgesic medication, which decreased the sensitivity of the pain model. It is well recognized that there is significant pharmacokinetic variability not only among patients but also within patients at different time points during the postoperative period. Greater flexibility in the dosing of analgesic medications may have helped to compensate for this pharmacokinetic and dynamic variability. With acute postoperative pain, more flexible dosage regimens (e.g., patient-controlled) provide superior pain relief .
In conclusion, ketorolac, 10 mg, and hydrocodone 7.5 mg in combination with acetaminophen, 750 mg PO, were equally effective in the management of pain after arthroscopic surgery. However, neither oral analgesic was very effective after laparoscopic tubal ligation. In these outpatient surgical populations, oral ketorolac was reported by patients to be better tolerated than the hydrocodone-acetaminophen combination in the treatment of acute postoperative pain, although side effect profiles, per se, were similar.
The authors thank Yifeng Ding, MD, and Cathy Martin, RN, for their assistance with the study. In addition, the support of Thomas J. Maneatis, MD, Arnz P. Shefrin, MS, and Marjorie Cadden, MS, at Syntex (Palo Alto, CA) is appreciated.
1. Joshi GP. Postoperative pain management. In: White PF, ed. Ambulatory anesthesia and surgery. London, WB Saunders. 1997:477-86.
2. Souter AJ, Fredman B, White PF. Controversies in the perioperative use of nonsteroidal antiinflammatory drugs. Anesth Analg 1994;79:1178-90.
3. Quiding H, Oikarinen V, Sane J, et al. Analgesic efficacy after single and repeated doses of codeine and acetaminophen. J Clin Pharmacol 1984;24:27-34.
4. Lasagna L. Analgesic methodology: a brief history and commentary. J Clin Pharmacol 1980;20:373-6.
5. Brown CR, Moodie JE, Dickie G, et al. Analgesic efficacy and safety of single-dose oral and intramuscular ketorolac tromethamine for postoperative pain. Pharmacotherapy 1990;10:59S-70S.
6. Wong HY, Carpenter RL, Kopacz DJ, et al. A randomized, double-blind evaluation of ketorolac tromethamine for postoperative analgesia in ambulatory surgery patients. Anesthesiology 1993;78:6-14.
7. Forbes JA, Butterworth GA, Burchfield WH, et al. Evaluation of ketorolac, aspirin, and an acetaminophen-codeine combination in postoperative oral surgery pain. Pharmacotherapy 1990;10:77S-93S.
8. Edwards ND, Barcklay K, Catling SJ, et al. Day case laparoscopy: a survey of postoperative pain and an assessment of the value of diclofenac. Anaesthesia 1991;46:1077-80.
9. Brodie BL, Casper RF. Prostaglandins mediate postoperative pain in falope ring sterilization. Am J Obstet Gynecol 1985;151:175-7.
10. Green CR, Pandit SK, Levy L, et al. Intraoperative ketorolac has an opioid-sparing effect in women after diagnostic laparoscopy but not after laparoscopic tubal ligation. Anesth Analg 1996;82:732-7.
11. O'Hara DA, Fragen RJ, Kinzer M, et al. Ketorolac tromethamine as compared with morphine sulphate for the treatment of post-operative pain. Clin Pharmacol Ther 1987;41:556-61.
12. Beaver WT. Measurement of analgesic efficacy in man. In: Bonica JJ, Lindbolm U, and Iggo A, Eds. Advances in pain research and therapy. New York: Raven Press; 1983:411-34.
13. McQuay HJ, Carroll D, Watts PG, et al. Codeine 20 mg increases pain relief from ibuprofen 400 mg after third molar surgery. A repeat-dosing comparison of ibuprofen and an ibuprofencodeine combination. Pain 1989;37:7-13.
14. McQuay HJ, Bullingham RES, Moore RA, et al. Zomepriac, dihydrocodeine and placebo compared in postoperative pain after day-case surgery: the relationship between the effects of single and multiple doses. Br J Anaesth 1985;57:412-9.
15. Honig WJ, Van OJ. A multiple-dose comparison of ketorolac tromethamine with diflunisal and placebo in postmeniscectomy pain. J Clin Pharmacol 1986;26:700-5.
16. McQuay HJ, Poppleton P, Carroll D, et al. Ketorolac and acetaminophen for orthopedic postoperative pain. Clin Pharmacol Ther 1986;39:89-93.
17. Kagi P. A multiple-dose comparison of oral ketorolac and pentazocine with treatment of postoperative pain. Curr Ther Res 1989;45:1049-59.
18. Forbes JA, Kehm CJ, Grodin CD, et al. Evaluation of ketorolac, ibuprofen, acetaminophen, and an acetaminophen-codeine combination in postoperative oral surgery pain. Pharmacotherapy 1990;10:945-1055.
19. Johansson S, Josefesson G, Malstam J, et al. Analgesic efficacy and safety comparison of ketorolac tromethamine and doleron for the alleviation of orthopedic post-operative pain. J Intern Med Res 1989;17:324-32.
20. Fernandez-Sabate A, Portabella F. Comparative multiple-dose study of ketorolac tromethamine and diflunisal for pain following orthopaedic surgery. J Intern Med Res 1991;19:210-8.
21. Bravo LJ, Mattie H, Spierdijik J, et al. The effects on ventilation of ketorolac in comparison with morphine. Eur J Clin Pharmacol 1988;35:491-4.
22. Yee MK, Evans WD, Facey PE, et al. Gastric emptying and small bowel transit in male volunteers after IM ketorolac and morphine. Br J Anaesth 1991;67:426-31.
23. Kenny GNC. Potential renal, haematological and allergic adverse effects associated with nonsteroidal anti-inflammatory drugs. Drugs 1992;44 (suppl. 5):31-7.