Methods

Double‐blinded, randomised, placebo‐controlled, active comparator, 2‐centre study.

Parallel group.

Dummy medications given.

Single dose.

Postoperative dosing.

Participants

Males and females 16–40 years of age with at least 1 bony impacted third molar or 2 ipsilateral impacted third molars. An impaction score was assigned to the third molars to demonstrate that they were suitably impacted in bone for inclusion.

The main exclusion criteria were history of significant disease, ongoing painful conditions (other than the third molar), migraine, malabsorption states, allergy/intolerance to the study medication, gastrointestinal complaints, psychotic illness or drug abuse, concomitant medication that would have interfered with the study drugs, pregnancy/lactation and taking NSAIDs from midnight the night before surgery.

614 patients screened, 322 randomised and 318 completed study.

Interventions

Patients underwent surgical removal of 1 partially or full bone impacted mandibular third molar, or 2 ipsilateral third molars under local anaesthetic with nitrous oxide sedation. Following surgery, patients who fulfilled the inclusion criteria regarding baseline pain intensity were randomly allocated to 1 of 4 treatment groups in the ratio 1:1:1:1. These were.

• Sodium ibuprofen: 2×256 mg plus 2 matched placebo for ibuprofen/poloxamer tablets plus 2 matched placebo for 500 mg acetaminophen caplets (n = 80).

• Ibuprofen/poloxamer: 2×200 mg ibuprofen acid tablets, each tablet incorporating 60 mg of the surfactant poloxamer 407, plus 2 matched placebo for sodium ibuprofen tablets plus 2 matched placebo for 500 mg acetaminophen caplets (n = 80).

• Acetaminophen: 2×500 mg acetaminophen (Tylenol Extra Strength) caplets plus 2 matched placebo for sodium ibuprofen tablets plus 2 matched placebo for ibuprofen/poloxamer tablets (n = 81).

• Placebo: 2 matched placebo for sodium ibuprofen tablets plus 2 matched placebo for ibuprofen/poloxamer tablets plus 2 matched placebo for 500 mg acetaminophen caplets (n = 81).

Rescue medication was provided, if required within the first 4 hours following surgery, an intramuscular injection of ketorolac tromethamine (60 mg) was given. After 4 hours, acetaminophen 500 mg/hydrocodone
5 mg or ketorolac tromethamine was given.

Antibiotics were prescribed postoperatively. Caffeine‐containing foods and drinks were to be discontinued from midnight prior to surgery until the end of the 6‐hour postdose assessment period.

Patients were randomised to treatment when they rated their baseline PI as moderate or severe, and the score on the VAS was ≥ 50 mm but ≤ 85 mm.

Outcomes

Pain intensity at baseline (immediately following surgery), 5, 10, 15, 20, 25, 30, 35, 40, 45, 60, 90, 120, 180, 240, 300 and 360 minutes after dosing measured using VAS and categorical scale of 0 (none) to 3 (severe).

Pain relief measured at the same time as PI with the exception of baseline. Scale of 0‐4 used (0 = none and 4 = complete). Also asked whether starting pain has at least half gone (no = 0, yes = 1).

Stopwatches were started at the time of dosing, 1 was stopped when the patient felt any pain relief whatsoever and the second was stopped when the patient decided that the relief was meaningful to them. If the patient did not stop the watches within the first 4 hours or if rescue medication was used, the stopwatches were discontinued for that patient.

Distractibility from pain was assessed at baseline and at 60% 360 minutes after dosing. VAS was used in response to the question 'How easy is it for you to distract yourself from your pain?.'

The Rainier Scale was completed at baseline and at 60 and 360 minutes after dosing. This assessed perceived functional impairment of activities of daily living, measured on a 1‐10 scale (1 = wound not interfere at all, 10 = would completely interfere).

Time of rescue medication was recorded, patients taking rescue medication completed all pain intensity and pain relief assessments immediately before medication was taken and continued to record their pain assessments throughout the 6‐hour assessment period.

Global evaluation was scored at the end of the 6‐hour period or at the time of rescue medication. Patients were asked, 'How effective do you think the study medication is as a treatment for pain?.' Response choices were 1 = excellent, 2 = very good, 3 = good, 4 = fair or 5 = poor.

A postoperative review was conducted 5‐12 days after surgery.

Notes

Sodium ibuprofen 256 mg is equivalent to 200 mg ibuprofen acid.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Patients "were randomly allocated to one of four treatment groups... in a 1:1:1:1 ratio according to a computer‐generated randomisation schedule that stratified patients by sex and baseline pain intensity."

Allocation concealment (selection bias)

Unclear risk

This is unclear.

Blinding (performance bias and detection bias)
All outcomes

Unclear risk

Study claims to be double blinded, but no indication of how blinding of study participants, nurses or assessors was implemented.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

All data reported on with intention‐to‐treat flow diagram presented in paper. All drop‐outs accounted for.

Selective reporting (reporting bias)

Low risk

All adverse events and outcomes reported as planned.

Other bias

Low risk

Drop‐outs, end points and adverse effects were all documented.

Methods

Double‐blinded, randomised, placebo‐controlled, single centre study (2 sites).

Parallel group.

Single dose.

Postoperative dosing.

Pennsylvania, USA and other collaborators.

Participants

Private outpatients at least 15 years of age who had 1 or more impacted third molars surgically removed.

Patients were excluded if they were pregnant or lactating; had any history of hypersensitivity or serious adverse reaction to any agent similar to the study medications; had any clinically significant condition that would affect the absorption, metabolism, or excretion of the study medications; or required concomitant medication that would make it difficult to quantify analgesia.

Long‐term users of analgesics and tranquillisers were also excluded.

269 patients were randomised, 206 patients completed the study.

Interventions

Treatments (1 or more third molar surgical extractions) were carried out under general anaesthetic with additional local anaesthetic (lignocaine). Patients were instructed to take the study medication when they had moderate or severe pain (not specified as VAS equivalent).

Interventions.

• Ketorolac 10 mg (n = 31).

• Ketorolac 20 mg (n = 35).

• Ibuprofen 400 mg (n = 32).

• Acetaminophen 600 mg (n = 36).

• Acetaminophen 600 mg + codeine 60 mg (n = 38).

• Placebo (n = 34).

The medications were issued as 2 tablets identical in appearance.

Rescue medication was provided: combinations of acetaminophen with codeine and/or oxycodone.

The patients returned to the surgeon's office 5 days postoperatively for a follow‐up visit.

Outcomes

Following the first dose of study medication, subjects responded to the following statements at hourly intervals up to 6 hours.

• My pain at this time is none (0), slight (1), moderate (2), severe (3).

• My relief from starting pain is: none (0), a little (1), some (2), a lot (3), complete (4).

• My starting pain is at least half gone: no (0), yes (1).

• At the end of the 6‐hour observation period, or when the participant took the second dose of medication, participants made a global evaluation of the study medication ranging from poor (0) to excellent (4).

Participants continued with the study drugs for 15 doses.

Adverse effect data were also collected and summarised in the paper.

Notes

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

"Using a random numbers generator, a computer assigned patient numbers, in blocks of 12, to the six treatment groups."

Allocation concealment (selection bias)

High risk

No details given.

Blinding (performance bias and detection bias)
All outcomes

Low risk

The study personnel were not aware of the identity of the study medication at the time it was administered and evaluated.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

All missing data accounted for, reasons for not completing study included in paper.

Selective reporting (reporting bias)

Low risk

All adverse events and outcomes reported as planned.

Other bias

Low risk

All outcomes well documented.

Methods

RCT, double‐blinded.

Single dose.

Postoperative dosing.

Single centre: University Dental Hospital, Pennsylvania.

Participants

210 participants.

Participants had to be at least 16 years of age, be in good general health, requiring removal of > 1 bony impacted wisdom teeth, and have no specific contraindications to the use of ibuprofen, aspirin, related NSAIDs, or acetaminophen. Women who were sexually active had to be using a medically approved method of contraception and had to have a negative urine pregnancy test on the day of surgery. Pregnant or lactating women and any patient who had received other analgesics, anti‐inflammatory drugs, sedatives (except for conscious sedation during the surgical procedure), or psychotropic agents within 12 hours of the study were excluded.

Interventions

All surgery carried out under local anaesthetic with "most patients" also receiving intravenous conscious sedation.

Treatment groups.

• Ibuprofen liquigel 200 mg (n = 61).

• Ibuprofen liquigel 400 mg (n = 59).

• Acetaminophen caplets 1000 mg (n = 63).

• Placebo (n = 27).

Administered by mouth with water when postsurgical pain became moderate or severe (> 50 mm on a 100 mm VAS severity scale).

Patients who did not experience pain within 5 hours were not given medication.

Rescue medication (500 mg acetaminophen plus hydrocodone bitartrate 5 mg) was given at any time after the 1 hour assessment period.

Outcomes

Pain relief and pain intensity at 0.25, 0.5, 0.75, 1, 1.5, 2, 3, 4, 5 and 6 hours after initial dosing. Pain relief was assessed on a 5‐point scale on 0 (no relief) to 4 (complete relief). Pain intensity was assessed on a 3‐point scale 0 (none) to 3 (severe). Exact timings of onset of first perceptible relief and meaningful relief were both recorded using stopwatches.

Derived data: Hourly categorical scores for PID. SPID was scored at 2 and 6 hours (SPID2 and SPID6). Time weighted pain relief scores summed to derive 2‐ and 6‐hour total pain relief (TOTPAR2 and TOTPAR6). At conclusion patients were asked to provide a global assessment of study medication and adverse reactions if and when occurred were recorded.

Notes

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

High risk

Quote: "..double blind, double dummy, placebo controlled, randomised, parallel group clinical trial in which patients were stratified for sex and baseline pain." However, no detail given as to how patients were selected for each group.

Allocation concealment (selection bias)

High risk

The drugs appear to have been similar in number and form but this was not clearly stated and no further details were given.

Blinding (performance bias and detection bias)
All outcomes

Unclear risk

The study was said to be "double blinded" but no other details were given.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

210 patients participated and were included in statistical analysis. There were no drop‐outs.

Selective reporting (reporting bias)

Low risk

All adverse effects and outcomes reported as planned.

Other bias

High risk

Different methods of conscious sedation were used in addition to local anaesthetic. Quote: "most patients also received conscious sedation", this is not quantified.

Methods

Double‐blinded, double‐dummy, parallel‐group RCT.

Single dose.

Postoperative dosing.

Participants

Healthy male and female patients at least 15 years of age, who required surgical removal of 2 or more impacted (at least 1 partially embedded in bone) third molars. In order to be included, they had to experience moderate or severe pain associated with the surgical procedure (not specified as to how this was measured). Patients were excluded from the study if they received any analgesic within 4 hours or a long‐acting analgesic within 12 hours of the study medication; received anaesthesia other than mepiva‐Caine hydrochloride, fentanyl, or methohexital during the surgery; or were taking any concurrent medication that could confound the evaluation of analgesia or safety.

Interventions

All patients had surgery performed under general anaesthetic with supplemental local anaesthetic.

3 treatment groups.

• Ibuprofen lysine 400 mg, 2x200 mg tablets (n = 99).

• Acetaminophen 1000 mg, 2x500 mg tablets (n = 101).

• Placebo (n = 99).

Patients received a single dose of the test medication when the pain was moderate or severe (not specified as to how this was measured).

Rescue medication (backup) was provided but not stated as to what the drug was. Patients were asked not to re‐medicate during the first 1 hour period. If a patient did re‐medicate within the trial period, the time was noted and no further efficacy evaluations were taken.

Outcomes

A stopwatch was started when the study drug was administered and patients were instructed to stop the watch when they experienced meaningful pain relief. If they did not experience meaningful relief within 2 hours after dosing, use of the stopwatch was discontinued.

Response to treatment was evaluated by patient self rating of pain intensity (0 = none, 1 = slight, 2 = moderate, 3 = severe) and degree of pain relief (0 = none,1 = a little, 2 = some, 3 = a lot, 4 = complete) at 15, 30, 45, 60, and 90 minutes and 2, 3, 4, 5, and 6 hours postdose. At the last evaluation time, the patient provided a global evaluation of the study drug (0 = poor, 1 = fair, 2 = good, 3 = very good, 4 = excellent).

Adverse clinical experiences were recorded by the study co‐ordinator.

Data recorded for.

• PID up to 6 hours.

• Time to onset of analgesic effect.

• Peak analgesic effect.

• Overall analgesic effect.

• Time to PID > 1.

• Time to meaningful pain relief.

• SPID at 6 hours.

• TOTPAR at 6 hours.

• Patient global evaluation.

• Time to re‐medication.

• Number of re‐medications.

Notes

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

"Patients were assigned to one of three treatment groups... according to an allocation schedule of random numbers."

Allocation concealment (selection bias)

High risk

No details given.

Blinding (performance bias and detection bias)
All outcomes

Unclear risk

The study was said to be a "double blind, double‐dummy", however, no further details were given.

Incomplete outcome data (attrition bias)
All outcomes

Unclear risk

"One of the patients in the ibuprofen lysine group had only one third molar removed and did not record efficacy evaluations; this patient was excluded from the efficacy analysis."

Otherwise all participants accounted for.

Selective reporting (reporting bias)

Low risk

All adverse effects and outcomes reported as planned.

Other bias

High risk

No detail of distribution of study medications or if they were identifiable to the participant or operator.

Methods

Randomised, double‐blinded, placebo‐controlled, parallel‐group, 2‐centre, modified factorial study.

Single dose.

Postoperative dosing.

Participants

Healthy male or female outpatients were eligible for the study if they were aged 16 to 40 years, were scheduled to undergo surgical removal of 3 to 4 impacted molars 2 of which would be mandibular molars requiring bone removal.

Impaction scoring (1‐4) was used to assess the molars.

Exclusion criteria included a history of migraine headaches within the previous year; gastrointestinal disorders such as peptic or duodenal ulcer, dyspepsia, or heartburn; hypersensitivity to the study medications; and drug or alcohol abuse. Patients with a current history of significant disease, including psychotic illness or neurosis, were also excluded, as were those who had other painful conditions, were taking medications that might confound the assessment of pain relief, or were unable to refrain from smoking. Women who were pregnant or lactating were not eligible for enrolment.

Interventions

Standard oral surgical procedures carried out under local anaesthetic and conscious sedation using nitrous oxide, diazepam, and methohexital (barbiturate drug). Following surgery, eligible patients were randomly assigned in a ratio of 2:1:2:1:1 to a single oral dose of the following.

• Ibuprofen 400 mg/paracetamol 1000 mg (n = 67).

• Ibuprofen 200 mg/paracetamol 500 mg (n = 33).

• Ibuprofen 400 mg alone (n = 69).

• Paracetamol 1000 mg alone (n = 34).

• Placebo (n = 31).

Medication was given when postoperative pain intensity was rated at least moderate on the pain intensity categorical rating scale where 0 = none; 1 = mild; 2 = moderate; and 3 = severe and pain intensity was ≥ 50 mm on a 100‐mm VAS.

Rescue medication was provided within the first 4 hours using tramadol 100 mg, and paracetamol 500 mg in combination with hydrocodone 5 mg or tramadol 100 mg after the first 4‐hour period. All assessments completed after the patient had taken rescue medication were considered missing.

Outcomes

Pain was assessed immediately after surgery (before dosing) and at specified intervals for up to 8 hours after dosing (15, 30, 45, 60, 90, 120, 180, 240, 300, 360, 420, and 480 minutes).

Primary efficacy end point was the sum of pain relief and pain intensity differences from baseline (0 hours) to 8 hours postdosing (SPRID8). This measure was defined as the area under the curve (AUC) for the sum of the 2 measures (pain relief and PID) at each time point from 0 to 8 hours.

Secondary efficacy end points were: total pain relief from 0 to 8 hours (TOTPAR8), sum of pain intensity differences from 0 to 8 hours (SPID8), SPID on the VAS from 0 to 8 hours (SPID8 VAS), TOTPAR from 0 to 4 and from 0 to 6 hours (TOTPAR4, TOTPAR6), SPID4, SPID6, SPRID4, SPRID6, SPID4 VAS, SPID6 VAS, individual pain relief from 15 minutes to 8 hours, peak pain relief over 8 hours, individual PID from 15 minutes to 8 hours, PID VAS from 15 minutes to 8 hours, peak PID and peak PID VAS over 8 hours, time to PID ≥1, PRID, time to first perceptible pain relief, time to first confirmed perceptible pain relief, time to first meaningful pain relief, time to use of rescue medication, time to pain half gone, and patient's global assessment of pain relief on a 5‐point scale (1 = poor; 2 = fair; 3 = good; 4 = very good; 5 = excellent).

The 2‐stopwatch method was also used to assess perceptible and meaningful pain relief (as in Daniels 2009).

Patients were assessed 5‐7 days postoperatively in relation to their surgery and to assess tolerability of the study medications.

Notes

No reason given for 2:1:2:1:1 ratio used for allocating participants into different study groups.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Upon study entry, patients at each site were stratified by sex and baseline pain intensity and given a unique number in sequence according to a predefined schedule. The block design for the randomisation schedule was in groups of 7.

Allocation concealment (selection bias)

Low risk

The randomisation code was supplied to the investigator in a sealed envelope and only broken if necessary due to safety concerns.

Blinding (performance bias and detection bias)
All outcomes

Low risk

Unblinded study personnel were not involved in any part of the study other than the preparation and verification of doses. Identical medication was dispensed by blindfolded staff to blindfolded patients.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

All data reported on with intention‐to‐treat flow diagram presented in paper. All drop‐outs accounted for.

Selective reporting (reporting bias)

Low risk

All adverse effects and outcomes reported as planned. No obvious selective reporting of outcomes.

Other bias

Low risk

Drop‐outs, end points, adverse effects all documented.

Methods

Multicentre, 2‐stage, randomised, double‐blinded, parallel‐group, placebo‐controlled, factorial study. Related to Mehlisch 2010 but using an observation period of 72 hours and a 2‐stage study design.

3‐dose trial.

Postoperative dosing.

Participants

Male or female outpatients aged ≥ 16 years undergoing surgical removal of at least 3 impacted third molars (2 of which had to be mandibular impacted molars).

Impaction scoring (1‐4) was used to assess the molars.

Same exclusion criteria as Mehlisch 2010.

Interventions

Stage 1: (first 8 hours).

Patients randomly assigned to 1 of the following treatment groups.

• Ibuprofen 200 mg (n = 75).

• Ibuprofen 400 mg (n = 74).

• Paracetamol 500 mg (n = 76).

• Paracetamol 1000 mg (n = 74).

• Ibuprofen 100 mg/paracetamol 250 mg (n = 71).

• Ibuprofen 200 mg/paracetamol 500 mg (n = 143).

• Ibuprofen 400 mg/paracetamol 1000 mg (n = 149).

• Placebo (n = 73).

Stage 2: (72 hours).

Patients who had been taking the combination drugs or placebo stayed on these, but those on monotherapy received the combination drugs incorporating the same dose of active monotherapy from phase 1.

Medication was administered when postoperative pain intensity was rated at least moderate on the pain intensity categorical rating scale where 0 = none; 1 = mild; 2 = moderate; and 3 = severe and pain intensity was ≥ 50 mm on a 100‐mm VAS. Medication had to be given within 6 hours of surgery but > 3 hours after fentanyl was last administered in order for the participant to be included.

In stage 2, patients were instructed to take their assigned study medication when at least 8 hours had elapsed since their previous dose of study medication during stage 1, when their pain VAS score was ≥ 30 mm, and provided that they had not consumed > 2 doses of first‐line rescue medication in the previous 24 hours. As in stage 1, rescue medication was available as needed, but to ensure that the daily maximum dose of paracetamol was not exceeded, patients were allowed first‐line rescue medication only twice in any 24‐hour period. Patients who required > 2 doses of first‐line rescue medication in any 24‐hour period, in addition to the 3 doses of study medication, were considered treatment failures and were allowed to take tramadol 100 mg as second‐line rescue medication.

Patients were given 6 tablets to take in stage 1 and 2 tablets in stage 2 to ensure adequate concealment.

Rescue medication was provided (hydrocodone 7.5 mg and paracetamol 500 mg) at any time after dosing, but any patient in stage 1 who required rescue medication in the first 60 minutes was considered a "drop‐out" and any patient requiring > 2 doses in the first 24‐hour period of stage 2 were considered treatment failures.

735 patients initially randomised, 715 entered stage 2, 678 completed both stages.

Follow‐up assessment was carried out 7‐10 days postoperatively to assess vital signs and perform a physical examination, adverse events were also recorded.

Outcomes

Stage 1.

Same primary efficacy end points as in Mehlisch 2010 (SPRID8) along with (PRID) scores at 15, 30, 45, 60, 90, 120, 180, 240, 300, 360, 420, and 480 minutes.

The key secondary end point was the patient's global assessment of the study medication, which was evaluated in response to the question 'How do you rate the study medication?'; response choices were 1 = poor; 2 = fair; 3 = good; 4 = very good; and 5 = excellent. This was assessed at the end of 8 hours, or at the time of rescue medication if earlier than 8 hours. Other secondary end points included, among others, TOTPAR from 0 to 8 hours, SPID from 0 to 8 hours, SPID on the VAS (SPID VAS) from 0 to 8 hours, time to pain half gone, and duration of effect (time to first administration of rescue medication).

The 2‐stopwatch method was also used as described in the previous paper.

Tolerability of the study medications was also assessed at 8 hours after dosing in stage 1, at 72 hours in stage 2 and also at the follow‐up visit.

Notes

Does not state how the molars were removed and under what type of anaesthetic. Although fentanyl is mentioned in the patients and treatment section.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Computer‐generated randomisation tables were used. The patients were stratified by sex and severity of pain, then assigned a unique number following a predefined schedule.

Allocation concealment (selection bias)

Low risk

The randomisation codes for each patient were supplied to the investigator in a sealed envelope.

Blinding (performance bias and detection bias)
All outcomes

Low risk

After preparation and checking, all tablets were deprinted and re‐packaged to match the placebos.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

All data reported on with intention‐to‐treat flow diagram presented in paper. All drop‐outs accounted for.

Selective reporting (reporting bias)

Low risk

All adverse effects and outcomes reported as planned. No obvious selective reporting of outcomes.

Other bias

Low risk

Drop‐outs, end points, adverse effects all documented.

Methods

Randomised, double‐blinded, triple‐dummy, parallel‐group study.

Single centre.

Single dose.

Postoperative dosing.

Participants

Healthy ambulatory male or female subjects, ages 16 to 65 years, who experienced moderate or severe pain after undergoing the surgical removal of 1 or more impacted third molars, 1 of which must have been at least a partial bony mandibular impaction, were included in the study.

Pregnant females, nursing mothers, and subjects with known sensitivity to acetaminophen, ketoprofen, ibuprofen, or other NSAIDs were excluded from participating in the study. Subjects with a recent history of serious medical condition or presence of bleeding disorders were excluded from the investigation. Subjects were also excluded if they had prior use of any analgesic, sedative, or psychotropic agent within 5 half‐lives for that drug before taking the study medication (except for local anaesthesia for the procedure). Prior use of any antihistamines within 48 hours of study entry was also prohibited. Subjects with a history of chronic abuse of analgesics or alcohol or substance abuse and subjects receiving other investigational drugs within 30 days of the study.

Interventions

All surgery performed under local anaesthetic (lignocaine), patients fasted from midnight the previous night.

The surgeon assessed the trauma rating of the procedure from mild to severe. Subjects remained at the study centre whilst medication was given and for 6 hours after receiving medication.

Treatment groups.

• Liquigel ibuprofen 400 mg (n = 67).

• Ketoprofen 25 mg (n = 67).

• Paracetamol 1000 mg (n = 66).

• Placebo (n = 39).

Rescue medication provided but not specified as to what drug it was.

Patients were medicated when they scored at least moderate pain on a categorical scale and a VAS of > 50.

Outcomes

Pain severity was evaluated using the categorical pain scale at 15, 30, and 45 minutes and then at 1, 1.5, 2, 3, 4, 5, and 6 hours following study drug administration.

At each assessment, patients rated their pain intensity and pain relief using the following categorical rating scale for pain intensity: none = 0, slight = 1, moderate = 2, or severe = 3. For pain relief, the following were used: none = 0, a little = 1, some = 2, a lot = 3, or complete relief = 4.

A stopwatch was used to describe meaningful pain relief time.

At the conclusion of the 6‐hour evaluation period or at the time of re‐medication (if it occurred before the 6th hour), each subject provided an overall evaluation of the study medication on a 5‐point categorical scale (from poor = 0 to excellent = 4).

If rescue medication was taken, no further measures were made. Time to rescue medication was also recorded.

Notes

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "Treatment assignments were determined by a randomisation schedule generated by the sponsor." This was independent of clinicians.

Allocation concealment (selection bias)

Low risk

Randomisation was controlled by the sponsor and all unit doses were identical in appearance.

Blinding (performance bias and detection bias)
All outcomes

Unclear risk

The study was said to be "double blinded", however, no details were given.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

239 patients included, all data appear to have been reported.

Selective reporting (reporting bias)

Low risk

All adverse effects and outcomes reported as planned. No obvious selective reporting of outcomes.

Other bias

Low risk

Drop‐outs, end points, adverse effects all documented.