Burney 2004.

Methods	Single‐blinded (outcome assessor), clinical RCT Sequence generation by random number tables Follow‐up: 6 months
Participants	Participants: 34 adults in a university setting in Ann Arbor, Michigan, USA Operation: unilateral inguinal hernia repair 2 groups, size: 15/18 Age: not reported Men/women: not reported Remarks: recurrent hernias or bilateral hernias were excluded
Interventions	Group 1 (spinal): spinal with lidocaine (5% with 7.5% dextrose, volume not reported), postincision: illio‐inguinal block with bupivacaine (0.5%, 8 mL to 10 mL), post‐op regimen not reported Group 2 (control): GA (fentanyl), postincision: illio‐inguinal block with bupivacaine (0.5%, 8‐10 mL), post‐op regimen not reported. Adjuvants: none Immediate post‐op pain control: significantly improved
Outcomes	Dichotomous: none reported Continuous: health status measured by SF‐36 at 6 months, but without randomization list
Notes	We contacted the study author for missing information on SF‐36 outcome. He provided original data and comments, but regretted that the randomization list was no longer available. Therefore the data could not be included. Funding sources: this study was supported by a grant from the Aetna Foundation, Hartford, Conn, USA. Conflicts of interest: no conflict of interest statement was provided.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "randomization was carried out using a blocked and balanced random number table."
Allocation concealment (selection bias)	Low risk	Quote: "a sealed opaque envelope with the randomization assignment was opened only after the patient had given informed consent for the study." The well‐described method makes bias unlikely.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Participants and caregivers were not blinded, but this is acceptable.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Outcome assessor blinding was not reported, but participants filled out the questionnaire alone. Study author responded: "research assistants collecting the data were blinded as to experimental groups during initial data collection. All data collection was by questionnaire. Research assistants were present for early data collection, but at 6 months I think it was only by mail."
Incomplete outcome data (attrition bias) All outcomes	High risk	Loss to follow‐up reported, but not assigned to groups or outcomes. Initially 34 participants were recruited, but only 23 questionnaires were collected at 6 months. Participants erroneously assigned to the wrong group were analysed with ITT. Bias is likely due to the unclear group allocation of participants lost to follow‐up.
Selective reporting (reporting bias)	Low risk	Primary outcomes fully reported on
Null bias	Unclear risk	Quote: "twelve (80%) of 15 patients in group 1 and 17 (94%) of 18 in group 2 received pain medication in the PACU (P = .3). In group 1, 10 (67%) of 15 patients received narcotic medication, and 6 (40%) of 15 patients received non‐ narcotic medication. In the group 2, 17 (94%) of 18 received narcotic medication, and 7 (39%) of 18 received nonnarcotic medication (P = .07 for narcotic medication; P > 0.99 for nonnarcotic medication)." No significantly decreased analgesic consumption in the PACU, however pain scores not reported.