The NARC (Nonsteroidal Anti-inflammatory in Renal Colic) Trial. Single-dose intravenous ketorolac versus titrated intravenous meperidine in acute renal colic: a randomized clinical trial

EM Advances

CJEM 2000;2(2):83-89

Abstract

Résumé

Introduction

Patients with renal colic often present to emergency departments with acute severe pain. Intravenous (IV) titration of opioid analgesics is a widely accepted method of providing rapid relief of severe pain;¹however, opioid titration is time consuming and requires more frequent nursing assessments than may be feasible. In addition, patients with renal colic often require higher opioid doses than many health care providers are comfortable administering.¹As a result, patients frequently receive inadequate analgesia.²Those who do receive adequate opioid doses may experience adverse effects, particularly sedation, and a common observation is that many patients are unable to resume normal activities after treatment.

Several studies have demonstrated that nonsteroidal anti-inflammatory drugs (NSAIDs) are effective in the treatment of renal colic.^3–11Ketorolac is effective for the treatment of renal colic,^10,11does not require close patient monitoring and is not associated with untoward sedation.

One case series showed that 30 mg of IV ketorolac provided good relief from renal colic pain in 10 to 20 minutes.¹⁰Another recent study suggested that single-dose IV ketorolac was superior to single-dose IV meperidine.¹¹Unfortunately, these studies do not consider the more relevant practice of opioid titration and, to date, no trials have compared single-dose IV ketorolac to titrated IV opioids in patients with acute renal colic.

This trial compares single-dose IV ketorolac to titrated IV meperidine in emergency department (ED) patients with acute renal colic. Our objectives were to compare the speed of onset and degree of analgesia offered by these agents, the incidence of adverse effects, and the patients’ functional status 60 minutes after initial study drug administration.

Our primary hypothesis was that the 2 agents provide equivalent pain relief 60 minutes after commencement of therapy. Our secondary hypotheses were that the agents would provide equivalent pain relief at 15, 30 and 45 minutes, that patients receiving ketorolac would experience fewer adverse effects, and that patients receiving ketorolac would be able to resume usual activities more frequently than those receiving meperidine.

Methods

Setting and patients

We studied a convenience sample of patients aged 18 to 65 who presented to the emergency department at 1 of 4 urban teaching hospitals. Each participating ED was staffed by full time emergency physicians (EPs) and treated from 45,000 to 65,000 patients per annum. Patients were eligible for inclusion if they were experiencing self-described mod-

The NARC Trial

erate to severe unilateral abdominal or flank pain that the treating EP suspected to be renal colic. Exclusion criteria included: pregnancy or breast-feeding; weight of less than 50 kg or more than 100 kg; allergy to meperidine, ketorolac, dimenhydrinate, acetylsalicylic acid (ASA) or another NSAID; and history of peptic ulcer disease, gastrointestinal bleeding, perforation, or inflammatory bowel disease. To be eligible for efficacy analysis, patients required objective evidence of renal colic, including an intravenous pyelogram (IVP) or ultrasound, interpreted by a radiologist as showing hydronephrosis, hydroureter or a visible calculus on the affected side; or passage of a calculus that was subsequently inspected by a pathologist.

Study design and treatments

This was a randomized, double-blind clinical trial. All patients underwent initial examination to collect baseline demographics, to assess pain severity and to assure study eligibility. Consenting patients were stratified into 2 groups — moderate and severe — based on self-described pain severity. After stratification, subjects were randomly assigned to receive meperidine or ketorolac. The randomization scheme was computer-generated and randomization was done in blocks of 6 for each stratum.

At time zero, a blinded study nurse administered 5 cc of study medication. Meperidine-treated patients received 50 mg of meperidine with 50 mg of dimenhydrinate, and ketorolac-treated patients received 30 mg of ketorolac. Study drugs were administered intravenously over 3 minutes from identical, pre-prepared, numbered syringes. At 15, 30, 45 and 60 minutes, subjects reported any adverse effects and documented their ongoing pain severity on visual analog and categorical scales.

At each 15-minute interval, additional IV study drug was given if ongoing pain was moderate or severe. Meperidinetreated patients received 25 mg (2.5 cc) of meperidine if their pain was moderate and 50 mg (5 cc) if it was severe. Ketorolac-treated patients received 2.5 cc of saline placebo if their pain was moderate and 5 cc if their pain was severe. In total, during the 1-hour study period, meperidine-treated patients received up to 200 mg of meperidine and ketorolac-treated patients received 30 mg of ketorolac. Our dosing regimen was based on a consensus of EPs as to what would commonly be used in practice. Patients were permitted to request rescue medication or drop out of the study at any time if pain relief was inadequate. The study period ended at 60 minutes.

This study was approved by the University of British Columbia’s Investigational Review Board and by the research ethics committees of the participating hospitals.

Efficacy, safety and tolerability assessment

At 15-minute intervals after the initial injection of study medication, a research assistant collected the following data.

Categorical pain assessment: Subjects rated their ongoing pain as none, mild, moderate or severe on a verbal rating scale (VRS).

Visual analog scale (VAS) pain severity: Subjects marked their pain intensity on a non-demarcated 100-mm horizontal line that had “no pain” written to the left and “most severe pain imaginable” to the right.

Adverse effects: Patients were asked to report any side effects and were asked specifically to grade nausea, vomiting, drowsiness, headache, dizziness, injection-site pain, dry mouth, nervousness, inability to concentrate, agitation, itching and sweating on a 4-point categorical scale: none, mild, moderate or severe.

Physiologic parameters: Blood pressure, pulse and respiratory rate were recorded.

Functional status: At the final clinical assessment (60 minutes), the research nurse and patient each rated the patient’s ability to resume usual activities on a 3-point scale: not impaired (able to return to work and drive a car), mildly impaired (able to do activities of daily living at home), or severely impaired (confined to bed).

End points and statistical analysis

The primary end point was successful treatment, defined as the proportion of patients with a change in the VRS from severe or moderate to mild or none at 60 minutes after the first injection. This end point was determined a priori and was modeled after a pain study done on migraine treatment.¹²Patients who withdrew from the study or requested rescue medication before the end of the study period were considered treatment failures.

The aim of the trial was to demonstrate therapeutic equivalence. Single-dose IV ketorolac was considered equivalent to titrated IV meperidine if the relative difference in successful treatment rates between the 2 agents was less than 20%. The null hypothesis (of non-equivalence) was that the relative difference in successful treatment rates (meperidine vs. ketorolac) is greater than or equal to 20%.

Secondary end points were VRS pain relief (from severe or moderate to mild or none) at 15, 30 and 45 minutes after the first injection; VAS pain scores at 15, 30, 45 and 60 minutes after the first injection; functional status, the proportion of subjects in each group who would be able to resume their usual activity (as assessed by the research nurse and by the patient) at 60 minutes after the first injection.

Based on a consensus of experienced EPs, we assumed that 80% of subjects would be successfully treated with meperidine. Accordingly, we calculated that a sample size of 63 patients per group was necessary to reject the null hypothesis of non-equivalence with a power of 80% at a one-sided 5% significance level if the specified alternative hypothesis (of equivalence) was true.

For the primary end point, a one-sided p value (testing for equivalence) and confidence intervals (CIs) are reported. The p value for the primary end point was computed using the normal approximation to the test of equivalence given by Dunnett and Gent.¹³For successful treatment and VAS scores at different time intervals, 95% CIs are reported. For functional status and the incidence of side effects, p values for chi-square tests of homogeneity are reported. SPSS 6.1.1 for the Mac was used for chi-square testing and to compute 95% CIs.

Results

Patient population

One hundred and ninety-five patients were enrolled and received study drug. One patient was excluded because of weight over 100 kg, leaving 194 eligible for analysis. Of these, 137 had a diagnostic IVP or ultrasound, 5 passed a calculus and 52 lacked objective evidence of renal colic, leaving 142 eligible for efficacy analysis. At 15 minutes, 1 meperidine-treated patient and 1 ketorolac-treated patient dropped out because of inadequate pain control. At 30 minutes, 4 meperidine-treated patients and 3 ketorolac-treated patients dropped out. At 45 minutes, 2 patients in each group dropped out, leaving 129 patients who completed the full study protocol.

Baseline characteristics

Baseline demographic and clinical data are summarized in Table 1. The characteristics of the patients in the 2 groups were similar, except for patient weight, which was greater in the meperidine-treated group.

Actual meperidine doses

Because meperidine was titrated to pain, the actual dose administered varied between patients. At 0 minutes, all patients received 50 mg. At 15 minutes, 19 patients received 50 mg and 22 received 25 mg. At 30 minutes, 14 patients received 50 mg and 19 received 25 mg. At 45 minutes, 10 patients received 50 mg and 16 received 25 mg. The mean total meperidine dose administered during the 1-hour study period was 99 mg.

Primary end point (Fig. 1)

Successful treatment (VRS pain relief from severe or moderate to mild or none) at 60 minutes occurred in 49 of 77 (64%; 95% CI, 53%–75%) meperidine-treated patients and in 47 of 65 (72%; 95% CI, 61%–83%) ketorolactreated patients (p value for equivalence = 0.002). The difference in success rates (percentage of meperidine-treated patients compared to percentage of ketorolac-treated patients) was –8% (95% CI, –23% to 7%), favouring ketorolac.

Other end points

Figure 1 also shows treatment success rates by time. Overlapping confidence intervals at all time intervals indicate that differences between groups were not statistically significant. The p values for equivalence were 0.05 at 15 minutes and less than 0.001 at 30 and 45 minutes, demonstrating statistical equivalence as defined above.

Figure 2 shows the proportion of patients, stratified by pain severity at study entry, who obtained adequate pain relief. There was a trend favouring ketorolac at all time intervals, even in patients with severe pain. Figure 3 shows VAS pain scores by treatment group at each time interval, demonstrating a similar trend.

Fifty-three patients were enrolled and treated, but were excluded from the efficacy analysis because they lacked objective evidence of urolithiasis (n = 52) or weighed more than 100 kg (n = 1). A secondary analysis, including these patients, demonstrated the same relative analgesic efficacy as the primary analysis.

Functional ability

Figure 4 shows that ketorolac-treated patients were more likely to be able to resume usual activity at study conclusion. Overall, 44% of ketorolac recipients vs. 10% of meperidine recipients rated themselves “unimpaired” at 60 minutes (p < 0.001). The research nurses’ assessment of functional ability was similar, also favouring ketorolac.

Safety and tolerability

The most frequently occurring adverse events are shown in Table 2. There were no statistically significant differences between treatment groups at any time interval for agitation, injection-site pain, itchiness, nausea, nervousness, sweating or vomiting.

Discussion

The aim of this multicentre trial was to determine whether single-dose IV ketorolac provides equivalent pain relief to titrated IV meperidine. We chose to perform an equivalence trial because we felt the 2 agents would have similar effectiveness and we thought that demonstrating equivalence would be more convincing than failing to show a statistically significant difference using a standard superiority trial. Many physicians will be unfamiliar with the statistical methodology of equivalence trials, and readers are referred to 2 recent thrombolytic trials^14,15that used designs similar to ours.

It is unlikely that any 2 agents are exactly equivalent, and to prove exact equivalence would require an infinite sample size. Therefore, when designing an equivalence trial, it is important to define a range of difference (or zone of indifference) that is accepted as “equivalent.” A relative difference of 20% has been proposed¹⁶as the criterion for declaring that an innovative therapy provides sufficient efficacy when tested against an active control; therefore, we defined equivalence as less than a 20% relative difference in successful treatment at 60 minutes. Our null hypothesis (of non-equivalence) was that meperidine is 20% better than ketorolac. Our research hypothesis (equivalence) was that meperidine is less than 20% superior. Consequently, if the upper limit of the 95% confidence range for the observed difference between drugs reached 20%, with meperidine superior (one-sided test), we could not conclude equivalence.

In this study, 72% of ketorolac-treated patients and 64% of meperidine-treated patients experienced adequate analgesia. The absolute difference is 8%, and the 95% CI around this difference extends from –7% (meperidine superior) to +23% (ketorolac superior). The corresponding range for relative success is –11% (meperidine superior) to +36% (ketorolac superior). Our calculated p value for equivalence (p = 0.002) means there is a 0.2% chance of observing the difference we did, favouring ketorolac, if meperidine is actually 20% better. Therefore, based on our pre-specified criteria, we rejected the null hypothesis and accepted our research hypothesis, that the 2 treatments are statistically equivalent.

One of our secondary hypotheses was that the 2 treatments would also provide equivalent pain relief at 15, 30 and 45 minutes. Figure 2 demonstrates that there was a trend favouring ketorolac at these time intervals, independent of whether the patient initially had moderate or severe pain.

We also hypothesised that ketorolac-treated patients would be better able to resume usual activity than meperidine-treated patients. Figure 4 shows that 44% of ketorolac recipients vs. 10% of meperidine recipients were judged able to resume usual activity at 60 minutes. At the same time, approximately 15% of ketorolac recipients and 50% of meperidine recipients were judged severely impaired. This impairment is primarily due to a statistically and clinically significant difference in adverse drug effects (Table 2), since VAS pain scores were not significantly different between groups (Fig. 3). These data suggest that meperidine’s relatively poor performance was not related to underdosing, which is a criticism of many analgesic trials.

Patients in the meperidine-treated group were heavier than patients in the ketorolac group; however, a linear regression analysis showed no effect of weight on pain intensity at 60 minutes.

Our trial is the first to compare titrated intravenous meperidine to single-dose ketorolac. In a previous study,¹¹Cordell and colleagues demonstrated that single-dose IV ketorolac (60 mg) provided better early analgesia than single-dose IV meperidine (50 mg). In the Cordell study, VAS pain scores in meperidine-treated patients dropped from approximately 80 mm at baseline to 55 mm at 15 minutes, but did not improve further at 30 minutes. In the current study, meperidine-treated patients improved from a similar baseline to 50 mm at 15 minutes, then to 42 mm at 30 minutes (after patients with moderate or severe pain received additional meperidine). In the Cordell study, ketorolac recipients improved to 35 mm at 15 minutes and 25 mm at 30 minutes, whereas in the current study, ketorolac recipients improved to 44 mm at 15 minutes and 28 mm at 30 minutes. These differences are not clinically important, but it is possible that 60 mg of ketorolac provides faster pain relief than the 30-mg dose we used.

A secondary analysis, including the patients who lacked objective evidence of urolithiasis, showed similar outcomes for efficacy, functional status and adverse effects, suggesting that our conclusions are robust and that IV ketorolac is effective for treating patients with “clinical” renal colic, whether or not confirmatory evidence is present.

We believe that the results of this study apply to everyday practice, since the clinical and demographic characteristics of the study population were similar to those of most patients with renal colic.

Limitations

First, it is possible that some patients with extremely severe pain may not have entered the study and that these patients might have had a different response to analgesics; however, our data suggest that patients with severe pain responded at least as well to ketorolac as to meperidine. Second, like many ED studies, we used a convenience sample, which may have introduced a selection bias. Third, we do not know whether intravenous NSAID administration will provide more rapid pain relief than rectal administration. Finally, although the combination of effective analgesia and good functional status suggests that there are economic and productivity benefits to using IV ketorolac, we did not perform a formal economic analysis to demonstrate these benefits.

Future research

This study was underpowered (ß = 0.73) to show a difference favouring ketorolac. A future study would require 417 patients per treatment arm to detect an 8% difference favouring ketorolac at the alpha = 0.05 level with 80% power. Such a trial would be difficult to perform and of little clinical significance, since many clinicians prescribe analgesics in combination for patients with severe pain. A more important study would be an equivalence trial comparing IV ketorolac to “balanced” analgesia using an opioid/NSAID combination.

Conclusions

In the doses studied, single-dose ketorolac is as effective as titrated IV meperidine for the ED relief of acute renal colic. In addition, IV ketorolac causes fewer adverse effects and it better preserves the patient’s functional ability 60 minutes after commencement of treatment.

References

1. Rosen P, Barkin R, Danzl D, editors. Emergency medicine, concepts and clinical practice, 4th ed. St. Louis: Mosby Year-Book; 1998. p. 282.
2. Rosen P, Barkin R, Danzl D, editors. Emergency medicine, concepts and clinical practice, 4th ed. St. Louis: Mosby Year-Book 1998. p. 279.
3. Lundstam S, Wahjander L, Kral JG. Treatment of ureteral colic by prostaglandin synthetase inhibition with diclofenac sodium. Cur Ther Res 1980;28:355-8.
4. Lundstam S, Wahjander L, Kral JG, Leissner K. Prostaglandin synthetase inhibition with diclofenac sodium in treatment of renal colic: comparison with use of a narcotic analgesic. Lancet 1982;1:1096-7.
5. Vignoni A, Fierro A, MoreschiniG, Cau M, Agostino A, Daniele E, et al. Diclofenac sodium in ureteral colic: a double blind comparison trial with placebo. J Int Med Res 1983;11:303-7.
6. Grenabo L, Aurell M, Delin K, Holmlund D, Sjodin JG. Antidiuretic hormone levels and the effect of indomethacin on ureteric colic. J Urology 1983;129:941-3.
7. Sommer P, Kromann-Andersen, Lendorf A, Lyngdorf P, Moller P. Analgesic effect and tolerance of Voltaren and Ketogan in acute renal or ureteric colic. Br J Urol 1989;63:4-6.
8. Wolfson AB, Yealy DM. Oral indomethacin for acute renal colic. Am J Emerg Med 1991;9:16-9.
9. Thompson JF, Pike JM, Chumas PD, Rundle JSH. Rectal diclofenac compared with pethidine injection in acute renal colic. Br Med J 1989;299:1140-1.
10. Larsen LS, Miller A, Allegra JR. The use of intravenous ketorolac for the treatment of renal colic in the emergency department. Am J Emerg Med 1993;11:197-9.
11. Cordell WH, Wright SW, Wolfson AB, Timerding BL, Maneatis TJ, Lewis RH, et al. Comparison of intravenous ketorolac, meperidine, and both (balanced analgesia) for renal colic. Ann Emerg Med 1996;28:151-8.
12. The Subcutaneous Sumatriptan International Study Group. Treatment of migraine attacks with sumatriptan. N Engl J Med 1991;325:316-21.
13. Dunnett CW, Gent M. Significance testing to establish equivalence between treatments, with special reference to data in the form of 2X2 tables. Biometrics 1977;33:593-602.
14. The COBALT Investigators. A comparison of continuous infusion of alteplase with double-bolus administration for acute myocardial infarction. N Engl J Med 1997;337:1124-30.
15. Tebbe U, Michels R, Adgey J, Boland J, Caspi A, Charbonnier B, et al. Randomized, double-blind study comparing saruplase with streptokinase therapy in acute myocardial infarction: The COMPASS equivalence trial. J Am Col Card 1998;31:487-93.
16. Ware JH, Antman EM. Equivalence trials. N Engl J Med 1997; 337:1159-61.