Rogers 1995.

Study characteristics
Methods	Study design: parallel‐group randomized controlled trial Sample size: 58 Country: UK Setting: secondary care hospital Dates conducted: not reported Postoperative opioid used and delivery: diamorphine PCA Pain score collection: not collected Concurrent postoperative analgesics: none reported
Participants	Inclusion criteria Abdominal hysterectomy by transverse lower abdominal incision Exclusion criteria ASA 3 or 4 < 45 kg or > 90 kg in weight Malignancy Endometriosis Alcohol or drug addiction Chronic analgesic use Contraindication to ketorolac
Interventions	Group Ketorolac before operation (30 participants): 10 mg IV ketorolac between induction and skin incision and saline placebo between closure of skin and recovery Group Ketorolac after operation (28 participants): saline placebo between induction and skin incision and 10 mg IV ketorolac between closure of skin and recovery
Outcomes	Diamorphine consumption (mg at 2, 4 and 12 hours) Opioid adverse events (nausea, vomiting, anti‐emetic use and pruritus at 12 hours) Intraoperative blood loss (in a standard way by weighing swabs and measuring the volume of suction loss intraoperatively)
Notes	Funding: no mention Declarations of interest: no mention Authors contacted: no Other: anti‐emetic used for nausea and vomiting. No events for pruritus so data not entered
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated. Quote: "...computer‐generated random numbers in blocks of 12"
Allocation concealment (selection bias)	Low risk	Pharmacy‐controlled. Quote: "The hospital pharmacy prepared.."
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Double‐dummy placebo used. Quote: "...containing either 0.9% normal saline 51 ml alone or 0.9% normal saline 50 ml with ketorolac 10 mg (1 ml). They were labelled with a code held by the pharmacy until the end of the study. Each patient was given one infusion between induction of anaesthesia and skin incision, and another between skin closure and arrival in the recovery ward".
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Likely blinded from above
Incomplete outcome data (attrition bias) All outcomes	High risk	Thirteen participants excluded. Quote: "...13 patients were later withdrawn; seven because a different operative procedure was performed, three because endometriosis was diagnosed at operation and three because of administrative errors in the conduct of the study".
Selective reporting (reporting bias)	Unclear risk	No protocol
Other bias	Low risk	Similar baseline characteristics