After major orthopedic surgery, the adjunctive use of nonsteroidal antiinflammatory medications (NSAIDs) may reduce postoperative opioid requirements and the incidence and severity of some opioid-related side effects [1,2]. For patients who will not tolerate oral or rectal NSAIDs in the early postoperative period, a parenterally administered NSAID may be advantageous. At present, in North America, ketorolac is the only NSAID available as a parenteral preparation.
Injectable ketorolac is approved for intramuscular (IM) administration only; however, some studies report the intravenous (IV) administration of the drug as single [3-7] or repeated boluses [8-10]. Ready et al.  have demonstrated recently that IV ketorolac administered to middle-aged patients after general, gynecologic, or orthopedic surgery was equally effective when administered either as a continuous infusion or as intermittent boluses; both methods were well tolerated and were associated with a 25% reduction in morphine consumption and lower pain scores compared to placebo. The role of adjunctive NSAIDs for patients after abdominal or gynecologic surgery is controversial, with not all studies reporting significant benefits . In contrast, the analgesic efficacy of NSAIDs in major orthopedic surgery is well established; however, these patients are often elderly, and may be especially susceptible to the adverse effects of analgesic medications generally. The purpose of this study was to assess the analgesic efficacy, opioidsparing effect, and tolerability of IV ketorolac administered to older adults as a single bolus followed by a continuous infusion after total hip or knee arthroplasty.
This prospective, randomized, double-blind, multicenter study was undertaken with the approval of the Health Protection Branch of the Department of Health and Welfare Canada, and with the approval of the ethical review boards of all the hospitals involved. All subjects provided written, informed consent to participate.
ASA physical status I-III patients of either sex, 50-75 yr of age, and between 40 and 110 kg scheduled to undergo unilateral total knee or hip arthroplasty were eligible to participate in the study. Exclusion criteria included drug or alcohol dependency or abuse; use of methadone or buprenorphine within 7 days of surgery; regular use of sedative, hypnotic, or anxiolytic medications; known or suspected intolerance or allergy to any narcotic or NSAID; gastric or duodenal ulcer disease in the previous 6 mo; history of bleeding problems or anticoagulant therapy within 4 wk of surgery (pre- or postoperative anticoagulant prophylaxis was permitted); other medical or psychiatric conditions which, in the opinion of the investigator, made participation in the study inappropriate, and concurrent participation in another clinical trial. Preoperatively, all patients received instruction in the use of the patient-controlled analgesia (PCA) device and the use of the measurement scales for pain, satisfaction, and side effects. No preoperative sedatives, narcotics, or antiemetics were permitted.
All patients received a standardized general anesthetic consisting of thiopental 2-5 mg/kg, fentanyl 3 micro gram/kg, and muscle relaxant for induction, followed by nitrous oxide, oxygen, and isoflurane for maintenance. Additional fentanyl up to 2 micro gram/kg was permitted at the discretion of the anesthetist. No other narcotics or antiemetics were permitted prior to emergence from anesthesia. After surgery patients were transferred to the postanesthesia care unit (PACU). Patients were withdrawn from the study and from further analysis prior to the administration of the study drug if they received more than 4 U of blood intraoperatively, or if there were significant intraoperative protocol violations.
In the PACU, when the patient complained of pain, up to two doses of morphine 3-10 mg IV were administered at the discretion of the PACU nurse. Baseline measurements of pain and sedation were recorded immediately prior to administration of the study drug. The study drug was administered within 2 h of surgery, but only after the PACU score was at least 6 on a scale of 10 (measuring somnolence, respiration, movement, color, and blood pressure on 0-2 scales), arterial oxygen saturation measured by pulse oximetry (SaO2) was 92% or greater (supplemental oxygen was permitted), and after the patient had rated his or her pain as at least moderate in severity on a categorical scale. Patients who failed to meet these criteria did not receive the study medication and were excluded from further analysis. The study drug was administered as a 1-mL bolus (ketorolac 30 mg/mL or placebo) over 15-30 s, followed by a continuous IV infusion at a rate of 5 mL/h (ketorolac 1 mg/mL or placebo) for 24 h. Because ketorolac 30 mg/mL is yellow, the initial bolus of study drug was supplied in yellow syringes to maintain blinding. PCA morphine with a bolus of 0.02 mg/kg, a lockout interval of 6 min, and no background infusion was started at the time of study drug administration. No prophylactic antiemetics were administered, but patients could receive metoclopramide 10 mg IV every 6 h administered at the discretion of the nursing staff. If this was ineffective, dimenhydrinate 25-50 mg IV was administered. Patients who required treatment of pruritus received diphenhydramine 25 mg IV as required. Patients received supplemental oxygen therapy via nasal prongs after their return to the surgical ward if their SaO2 on room air in the PACU remained below 92%. Patients reporting inadequate analgesia were withdrawn from the study (but included in the data analysis) and provided with adequate opioid analgesia at the discretion of the investigator.
Patients were evaluated postoperatively in the PACU (baseline) and at 2, 4, 6, and 24 h after administration of the study drug. Evaluations included level of sedation (completely awake, awake but drowsy, asleep but responds easily to verbal command, asleep but responds to touch, unresponsive), pain intensity (none, slight, moderate, severe, very severe) visual analog scale pain scores (VAS) on a 100-mm scale, and side effects. Side effects were defined as any adverse experience or unusual event, observed or volunteered by the patient, which occurred during the study period, whether or not they were considered drug-related. A record of morphine consumption was stored continuously in the memory of the PCA pump and was printed and analyzed after the 24-h study period. Patients were monitored with continuous pulse oximetry for the entire 24-h study period; these data were recorded continuously and were printed and analyzed at the end of the study to determine the frequency and severity of episodes of oxygen desaturation. Significant desaturation was defined as a SaO (2) less than 85% for at least 2 min or less than 90% for at least 10 min. After the study period (24 h) the observer rated the overall drug acceptability; the patient rated the overall pain relief and the overall acceptability of the drug. Total postoperative blood loss was determined from the nursing records (indwelling suction drainage), and a blood sample was taken to compare the pre- and postoperative hemoglobin concentrations.
Statistical analyses were undertaken by a third party who remained blinded to the study drug administered. Decisions to exclude any patient from efficacy analysis were made prior to breaking the randomization code. All patients who received the study medication were seen later for assessment of adverse events and their data are included in the results, whether or not they were withdrawn from the study prematurely. Patients withdrawn because of unsatisfactory analgesia or protocol violations were included in efficacy analyses only up to the time they were withdrawn from the study.
All treatment effects were compared using twotailed tests immediately before study drug administration (baseline) and at 2, 4, 6, and 24 h. In addition, morphine consumption was analyzed at 12 and 18 h. An alpha level of P < 0.05 was considered significant. Interval data were analyzed using analysis of variance. Ordinal data were analyzed using the Mann-Whitney U-test. The incidence of adverse events was compared between groups using the chi squared or Fisher's exact test. Interval data are reported as mean +/- SEM. Ordinal data are reported as median (range).
Of 181 patients enrolled in the study, seven were withdrawn before receiving the study medication and were excluded from further analysis. Of the remaining 174 patients, 86 received ketorolac and 88 received placebo. Twelve patients, five receiving ketorolac and seven receiving placebo, were withdrawn after study drug administration. Three patients in the placebo group were withdrawn because of inadequate analgesia. In each group, three patients were withdrawn for technical reasons (e.g., pump failure, unrecognized failure of the IV catheter). One patient in each group was withdrawn because of accidental protocol violations (administration of other NSAIDs), and one patient receiving ketorolac was found to be taking his own supply of analgesics. Two patients receiving ketorolac were excluded from all efficacy analyses. One had a baseline pain rated as mild, and therefore did not meet the inclusion criteria. For the other, the randomization code was broken prior to study completion. For the patients receiving placebo, three were excluded from all efficacy analyses; one withdrew within 1 h of study drug administration because of inadequate analgesia, one received an oral NSAID prior to surgery, and one did not receive PCA morphine as stated in the protocol.
Demographic data are summarized in Table 1. There are no significant differences between groups.
Patients receiving ketorolac reported less pain than those receiving placebo at all evaluations after study drug administration (i.e., 2, 4, 6, and 24 h) as measured by subjective pain intensity and VAS pain scores Table 2; Figure 1. These differences were statistically significant at 2, 4, and 6 h for pain intensity, and at 2 and 24 h for VAS pain scores. After completion of the 24-h study period, both the patients and the researcher rated overall satisfaction to be superior with ketorolac compared to placebo (P = 0.002; Figure 2).
Total morphine consumption for the 24-h study period was less for patients who received ketorolac compared to placebo Figure 3, and for patients who underwent total hip arthroplasty versus total knee arthroplasty. For total hip arthroplasty, patients receiving ketorolac consumed 35% less morphine than those receiving placebo (37.3 +/- 3.9 vs 57.2 +/- 4.6 mg; P = 0.03). For total knee arthroplasty, those receiving ketorolac consumed 44% less morphine (44.9 +/- 4.3 vs 80.3 +/- 8.8 mg; P = 0.0003).
Within both groups (ketorolac and placebo), patients undergoing knee arthroplasty reported greater pain intensity and higher VAS than those undergoing hip arthroplasty throughout the study period. In the ketorolac group these differences were statistically significant for pain intensity at 2 and 6 h (P = 0.0006 and 0.034, respectively). In the placebo group differences were statistically significant for pain intensity at 6 and 24 h (P = 0.009 and 0.026, respectively). VAS pain scores were significantly higher for patients undergoing knee arthroplasty compared to hip arthroplasty at 2 and 6 h for patients receiving ketorolac, and at 6 and 24 h for those receiving placebo. For patients receiving placebo, knee arthroplasty was associated with greater morphine requirements than hip arthroplasty (80.3 +/- 8.8 vs 57.2 +/- 4.6 mg; P = 0.012).
Patients receiving ketorolac were significantly less sedated at 6 and 24 h (P = 0.007 and 0.034, respectively), but not at 2 and 4 h. The number of patients reporting nausea or vomiting did not differ between groups (51% vs 56% and 25% vs 26% for placebo versus ketorolac for nausea and vomiting, respectively); however, patients receiving placebo more frequently received antiemetic therapy than those receiving ketorolac (16% vs 6%, P = 0.03). Urinary retention was reported in 19% of patients receiving ketorolac and 25% of those receiving placebo (not significant). Postoperative blood loss from the wound drains was similar between groups (ketorolac 598 +/- 64 placebo 600 +/- 48) as was the need for postoperative blood transfusion (ketorolac 0.4 +/- 0.1, placebo 0.3 +/- 0.1 U of packed cells).
Satisfactory pulse oximetry data were obtained from 79 patients receiving ketorolac and from 75 receiving placebo. The need for supplemental oxygen therapy was not significantly different between groups (54% of the ketorolac group, 69% of the placebo group; P = 0.057). In the 24 h after study drug administration episodic desaturation occurred with similar frequency in both groups (74 episodes in 13 patients receiving ketorolac, 80 episodes in 15 patients receiving placebo). Approximately one half of these episodes occurred in the recovery room, and no patients required treatment for clinically apparent respiratory depression or hypoxia after their return to the ward.
NSAIDs are widely used as adjuncts to PCA with IV opioids and are recommended for this purpose (unless contraindicated) in the Clinical Practice Guidelines of the U.S. Department of Health and Human Services . The results of the present study support these recommendations; morphine consumption is reduced and analgesia is improved. Of equal importance, some opioid-related side effects are reduced. In particular, somnolence at 6 and 24 h after surgery was less, and fewer patients required antiemetic therapy among those patients receiving ketorolac. This combination of improved analgesia and reduced incidence and severity of side effects is reflected in the higher overall patient and observer satisfaction with ketorolac compared to placebo.
In the present study, no adverse reactions were reported with the rapid (15-30 s) IV injection of ketorolac 30 mg, and there were no side effects attributable to ketorolac during the 24-h study period. Previous studies of IV ketorolac administered as single or repeated boluses, or by continuous infusion, suggest that the incidence and severity of adverse effects is similar to that associated with IM administration [3-11].
Inhibition of prostaglandin synthesis which occurs with ketorolac and other NSAIDs  is associated with an increased bleeding time, gastrointestinal ulceration, and an increased risk of acute renal failure in some patient populations [13-15]. Although bleeding time is increased with these drugs it generally remains within the normal range, and there is little evidence that surgical bleeding is increased . This and other controlled studies of bleeding after major orthopedic surgery [16,17], laparotomy , or prostatic resection  found no significant increase in bleeding associated with NSAID administration. It is common practice after major orthopedic surgery to administer warfarin or low-dose heparin to decrease the risk of deep vein thrombosis; to date there is no evidence of an interaction with NSAIDs which could increase postoperative blood loss .
In the well hydrated patient with normal renal function, NSAIDs have little adverse renal effect [13-15]. Postoperative hypovolemia may occur after major orthopedic surgery, yet reports of acute renal failure attributed to postoperative NSAID administration are rare. Renal function was not assessed in the present study but the risk must be considered, particularly in patients at high risk for severe postoperative hypovolemia.
Gastrointestinal ulceration and bleeding are the most frequently reported serious adverse reactions associated with NSAID administration; however, these complications are reported primarily in association with prolonged use [21,22]. The short-term use (less than 1 wk) of these drugs is not associated with an increased risk of severe gastrointestinal complications . In the present study, no patient was withdrawn because of gastrointestinal complaints, and only two patients, both in the placebo group, complained of abdominal pain. Whether IV ketorolac would be as well tolerated over several days is not known.
If one chooses to use ketorolac, is a continuous IV infusion necessary? The morphine-sparing effect and improvement in analgesia described in this study are consistent with most studies that examine the effects of NSAIDs after major orthopedic surgery, and it is not clear that continuous IV infusion is superior to enteral, IM, or intermittent IV bolus administration [1,17]. The bioavailability of oral ketorolac is almost 100%, and the metabolism and elimination half-life of the parent compound is similar regardless of the route of administration (i.e., oral, IM, or IV) . With an elimination half-life of 5 h, ketorolac 30 mg IM or IV every 6 h as recommended by the manufacturer may be as effective as a continuous infusion. Ready et al.  found no differences in analgesia, morphine consumption, or side effects with intermittent or continuous IV administration. In contrast, Burns et al.  compared ketorolac administered as repeated IM boluses with an IM loading dose followed by an IM infusion and demonstrated similar pain scores but greater morphine-sparing in the patients receiving an infusion; however, the study protocol dictated a larger ketorolac dose in the first hour of the study for the patients receiving the infusion, thus confounding the results. At present, the evidence favoring a continuous infusion is inconclusive.
This study also demonstrates that patients undergoing total knee arthroplasty have greater morphine requirements and higher pain scores than those undergoing hip arthroplasty. However, these are also the patients who gain the most from the addition of adjunctive therapy with ketorolac; the percentage reduction in morphine consumption is greater in these patients.
In conclusion, the administration of ketorolac 30 mg IV followed by an IV infusion at a rate of 5 mg/h for 24 h after total hip or knee arthroplasty was associated with improved analgesia, lower morphine consumption, less sedation and reduced requirements for antiemetic therapy compared to placebo. IV ketorolac was well tolerated and was not associated with an increased incidence of side effects. In particular, there was no difference in postoperative blood loss or the incidence of gastrointestinal symptoms. IV ketorolac is an effective alternative to IM ketorolac or enteral NSAID administration when used as an adjunct to PCA morphine after major orthopedic surgery.
1. Moote C. Efficacy of nonsteroidal anti-inflammatory drugs in the management of postoperative pain. Drugs 1992;44:14-30.
2. Dahl JB, Kehlet H. Non-steroidal anti-inflammatory drugs: rationale for use in severe postoperative pain. Br J Anaesth 1991;66:703-12.
3. Camu F, Van Overberge L, Bullingham R, Lloyd J. Hemodynamic effects of two intravenous doses of ketorolac tromethamine compared with morphine. Pharmacotherapy 1990;10:122S-6S.
4. Smith I, Shively RA, White PF. Effects of ketorolac and bupivacaine on recovery after outpatient arthroscopy. Anesth Analg 1992;75:208-12.
5. Ding Y, White PF. Comparative effects of ketorolac, dezocine, and fentanyl as adjuvants during outpatient anesthesia. Anesth Analg 1992;75:566-71.
6. Brown CR, Moodie JE, Wild VM, Bynum LJ. Comparison of intravenous ketorolac tromethamine and morphine sulfate in the treatment of postoperative pain. Pharmacotherapy 1990;10:116S-21S.
7. Kenny GN, McArdle CS, Aitken HH. Parenteral ketorolac: opiate-sparing effect and lack of cardiorespiratory depression in the perioperative patient. Pharmacotherapy 1990;10:127S-31S.
8. Liu J, Ding Y, White PF, et al. Effects of ketorolac on postoperative analgesia and ventilatory function after laparoscopic cholecystectomy. Anesth Analg 1993;76:1061-6.
9. Parker RK, Holtmann B, Smith I, White PF. Use of ketorolac after lower abdominal surgery. Anesthesiology 1994;80:6-12.
10. 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.
11. Ready LB, Brown CR, Stahlgren LH, et al. Evaluation of intravenous ketorolac administered by bolus or infusion for treatment of postoperative pain. Anesthesiology 1994;80:1277-86.
12. Acute Pain Management Guideline Panel. Acute pain management: operative or medical procedures or trauma. Clinical practice guideline. Washington, DC: US Department of Health and Human Services, Agency for Health Care Policy and Research Pub. No. 92-0032, 1992:76.
13. Insel PA. Analgesic-antipyretics and antiinflammatory agents; drugs employed in the treatment of rheumatoid arthritis and gout. In: Goodman LS, Gilman A, Rall TW, Nies AS, Taylor P, eds. Goodman and Gilman's the pharmacologic basis of therapeutics. Elmsford, NY: Pergammon Press, 1990:642.
14. Kenny GN. Potential renal, haematological and allergic adverse effects associated with nonsteroidal anti-inflammatory drugs. Drugs 1992;44Suppl 5:31-6.
15. Kehlet H, Dahl JB. Are perioperative nonsteroidal anti-inflammatory drugs ulcerogenic in the short term? Drugs 1992;44Suppl 5:38-41.
16. Taivainen T, Hiller A, Rosenberg PH, Neuvonen O. The effect of continuous intravenous indomethacin infusion on bleeding time and postoperative pain in patients undergoing emergency surgery of the lower extremities. Acta Anaesthesiol Scand 1989;33:58-60.
17. Lindgren U, Djupso H. Diclofenac for pain after hip surgery. Acta Anaesthesiol Scand 1985;56:28-31.
18. Rorarius MGF, Baer GA, Metsa-ketela T, et al. Effects of perioperatively administered diclofenac and indomethacin on blood loss, bleeding time and plasma prostanoids in man. Eur J Anaesthesthiol 1989;6:335-42.
19. Bricker SRW, Savage ME, Hanning CD. Peri-operative blood loss and nonsteroidal antiinflammatory drugs: an investigation using diclofenac in patients undergoing transurethral prostatic resection. Eur J Anaesthesthiol 1987;4:429-34.
20. Toon S, Holt BL, Mullins FG, Bullingham R, et al. Investigations into the potential effects of multiple dose ketorolac on the pharmacokinetics and pharmacodynamics of racemic warfarin. Br J Clin Pharmacol 1990;30:743-50.
21. Nuki G. Pain control and the use of non-steroidal analgesic antiinflammatory drugs. Br Med Bull 1990;46:262-78.
22. Fries J, Miller S, Spitz P, et al. Toward an epidemiology of gastropathy associated with nonsteroidal antiinflammatory drug use. Gastroenterology 1989;96:647-55.
23. Mroszczak EJ, Jung D, Yee J, et al. Ketorolac tromethamine pharmacokinetics and metabolism after intravenous, intramuscular, and oral administration in humans and animals. Pharmacotherapy 1990;10:33S-9S.
24. Burns JW, Aitken HA, Bullingham RE, et al. Double-blind comparison of the morphine sparing effect of continuous and intermittent i.m. administration of ketorolac. Br J Anaesth 1991;67:235-8.