Sun 2008.

Study characteristics
Methods	Study design: parallel‐group randomized controlled trial Sample size: 76 Country: USA Setting: secondary care hospital Dates conducted: not reported Postoperative opioid used and delivery: IV fentanyl or oral hydrocodone and acetaminophen in PACU and IV morphine PCA if admitted or oral hydrocodone and acetaminophen if discharged Pain score collection: blinded researcher Concurrent postoperative analgesics: none reported postoperatively in addition to above
Participants	Inclusion criteria Age 18‐75 years old ASA 1‐3 Major plastic surgery procedures (breast augmentation or abdominoplasty with or without liposuction involving the abdomen, buttocks and lower extremities) Exclusion criteria Allergy or contraindication to NSAIDs Chronic NSAID therapy Received any analgesic medication within a 12‐hour period before the operation Pregnant or breastfeeding Alcohol or drug abuse Bleeding disorder Unstable neurological, cardiovascular, renal, hepatic or gastrointestinal diseases Unwilling to complete the follow‐up evaluations
Interventions	Group Perioperative (39 participants): 2 celecoxib 200 mg capsules 30–90 minutes before surgery and 2 placebo capsules 1 hour after surgery and celecoxib 200 mg twice daily postoperatively Group Postoperative (37 participants): 2 celecoxib 200 mg capsules 1 hour after surgery and 2 placebo capsules 30–90 minutes before surgery and celecoxib 200 mg twice daily postoperatively
Outcomes	Postoperative pain (0‐10 VRS at 0.5, 1, 2, 3, 4, 24, 48 and 72 hours) Opioid consumption (mg morphine in PACU and at 24, 48 and 72 hours) Patient satisfaction (unclear scale and time point) Return of bowel function (in days) Time to analgesia (minutes) PACU stay (minutes) Nausea and vomiting (in PACU, yes/no and requirement for anti‐emetic) Resume normal diet (days) Resume normal activity (days) Quality of recovery (0‐18 VRS at 24, 48 and 72 hours) Cardiovascular and wound complications (yes/no at 7 and 30 days)
Notes	Funding: non‐industry Declarations of interest: not reported Authors contacted: no Other: myocardial infarction not included as no events. Patient satisfaction not included as unclear scale used. Pain score data extracted from graph and 2 hours used for early pain as postoperative intervention given at one hour. Morphine consumption calculated as PACU plus 24 hours and extracted from graph with SD estimated. No outcomes reported as time‐to‐event. Time to bowel movement mean from median and IQR/1.35 to estimate SD
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated. Quote: "...computer generated random number schedule"
Allocation concealment (selection bias)	Low risk	Pharmacy‐controlled. Quote: "The study medication was prepared by a hospital pharmacist in identical‐appearing capsules according to a computer generated random number schedule".
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Double‐dummy placebo. Quote: "The patients, nurses, surgeons, and anesthesiologists directly involved in the patients’ care were blinded as to the content of the oral study medication capsules".
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Blinded. Quote: "A trained interviewer who was also blinded to the study medication contacted each patient at 24, 48, and 72 postoperatively to inquire about their maximum VRS pain score".
Incomplete outcome data (attrition bias) All outcomes	Low risk	Four participants excluded. Reasons unlikely to cause bias
Selective reporting (reporting bias)	Unclear risk	No protocol
Other bias	Low risk	Similar baseline characteristics. Non‐industry funding