This investigation was a multicentre randomized, double-blind, double-dummy, controlled trial.
The study protocol was approved by the independent bioethics committee of the Institute for Maternal and Child Health IRCCS Burlo Garofolo of Trieste, Italy, and by the Italian pharmacological agency (AIFA), Eudract code: 2018-004057-25. It was recorded at ClinicalTrial.gov before the enrolment of the first participant, registration number NCT04133623.
The study was conducted at the PEDs of the Institute for Maternal and Child Health IRCCS Burlo Garofolo of Trieste, Italy (a tertiary-level university teaching children’s hospital) and the Santa Maria degli Angeli hospital of Pordenone, Italy.
Children and adolescents aged 8 to 17 who accessed the PED complaining of moderate to severe pain related to a limb trauma that occurred in the previous 48 hours were considered eligible for the study. Patients self-reported their pain through the numeric rating scale-11 (NRS-11), in which patients verbally rate their pain from 0 to 10 (0 no pain, 10 worst imaginable pain) [8].
Scores from 4 to 6 were considered moderate pain, and scores from 7 to 10 were severe.
Patients with at least one of the following features were excluded from the research: administration of any analgesic within the previous 8 hours; known allergy to ibuprofen or ketorolac; chronic use of painkiller; known hepatopathy or nephropathy, history of thrombocytopenia, coagulation disturbance; history of gastritis or esophagitis in the previous 30 days; clinical suspicion of violence or pregnancy; polytrauma; inability to self-report pain related to intellectual disability or language difficulties.
Eligible children were enrolled and randomly assigned to receive a dose of 10 mg/kg oral ibuprofen syrup (maximum of 600 mg) or a dose of 0.5 mg/kg oral drops of ketorolac (maximum of 10 mg), after obtaining written informed consent from a parent or legal guardian.
Using a computerized method by an independent epidemiologist at the Clinical Epidemiology and Public Health Research Unit at the IRCCS Burlo Garofolo Institute, we generated two separate randomization lists, one for patients with severe pain and one for patients with moderate pain, stratified per centre, with a fixed block size of 4.
Each list randomly assigned patients to receive ibuprofen or ketorolac and the placebo of the not-given drug in a double dummies fashion. Drug kits were produced, consecutively numbered and labelled to be assigned to subjects with moderate or severe pain. Each kit contained two vials closed into opaque envelopes, containing 1) a maximum extractable volume of 15ml of ibuprofen syrup (40mg/ml) and oral ketorolac placebo or 2) a maximum extractable volume of 5ml of ketorolac drops (2mg/ml) and oral ibuprofen placebo. Each vial reported the ml pro/kg dose and maximum ml dose to administer. Each drug and its corresponding placebo were indistinguishable in aspect, taste and flavour.
Consequently, each subject received one dose of active drug (ibuprofen or ketorolac) and one dose of placebo. Patients, nurses and clinicians were double-blinded on the drug and placebo administered. After the drug administration, the patients were asked to self-report their pain through the NRS-11 scale every 30 minutes until 2 hours and to report the onset of new symptoms, especially: headache, nausea, vomiting, dyspepsia, hitching, somnolence, abdominal pain or dizziness. If patients considered their pain poorly controlled after 1 hour, acetaminophen was offered as a rescue drug.
The primary study outcome was the difference in NRS-11 score reduction between the ketorolac and ibuprofen groups after 60 minutes of drug administration. The secondary results were the difference in NRS-11 score reduction between groups at 30-60-120 minutes, the number of patients achieving NRS-11 score <4 and the number of patients reaching an NRS-11 reduction >3 points at each time point between groups, and the number of patients asking for rescue analgesics in the two groups.
We hypothesized that ketorolac would be more effective than ibuprofen for severe pain and as effective as ibuprofen for moderate pain. According to the first hypothesis of ketorolac superiority over ibuprofen, based on the available literature indicating a one-point reduction in the NRS scale, a difference of minimal clinical importance for patients [9], we estimated a study population of 130 patients (65 for each group) to achieve a difference of 1 point between T0 and T60 in the NRS-11 score (with a standard deviation of 2) between the two groups with an alpha and beta error of 0.05 and 0, 20, respectively. The second hypothesis concerns the non-inferiority of ibuprofen compared to ketorolac in managing moderate to acute trauma-related pain. For the non-inferiority hypothesis, assuming a pain reduction of 1 point in the difference between T0 and T60 on the NRS-11 score as the minimum clinically significant reduction, we set the non-inferiority threshold at 50% of this value (0.5 points) (refs). The study needs 120 patients (60 for each group) according to the established non-inferiority threshold, a standard deviation of 1 point with an alpha and beta error of 0.05 and 0.20, respectively. We will consider non-inferiority confirmed if the upper limit of the 95% confidence interval of the difference between the mean of ibuprofen and ketorolac is less than the non-inferiority threshold. Both samples’ sizes were calculated for a 1:1 ratio between groups.
Statistical analysis
The statistical analysis was blinded and performed by the Clinical Epidemiology and Public Health Research Unit of the IRCCS Burlo Garofolo. Data for the two study hypotheses (superiority of ketorolac and non-inferiority of ibuprofen) were analysed separately. For the hypothesis of superiority, we planned analysis by intention-to-treat, while for the hypothesis of non-inferiority per protocol.
Continuous variables were reported as a median and interquartile range, based on data distribution, while categorical variables as numbers and percentages.
We used the Mann-Whitney test, based on data distribution, to evaluate the difference in NRS-11 score between the ibuprofen group and the ketorolac group 60 minutes after drug administration (primary study result). We calculated the 95% confidence interval of the difference between the two drugs and confirmed the non-inferiority of ibuprofen if the upper limit of the confidence interval remains below the defined non-inferiority threshold.
For secondary outcomes, differences were assessed with the Mann-Whitney's test for the continuous variables and the chi-square test (or Fisher's exact test where appropriate) for the categorical variables. Each time was evaluated separately. If the patient required another pain medication, it was evaluated as study drug failure, not as dropping out. A p-value <0.05 was considered statistically significant.