Ma 2012.

Methods	Design: parallel‐group RCT Recruitment period: 2009 to 2011 Administration setting: dental clinical of hospital Country: China Funding source: not stated
Participants	Number of participants randomised: 162; 162 teeth (rubber dam: 81; cotton rolls: 81) Randomisation unit: participant/tooth Age: not stated Sex: not stated Inclusion criteria (as translated): with NCCLs in mandibular premolars in dentine but without pulp exposure lesions above the gingival margins teeth with NCCLs having no occlusal trauma teeth with NCCLs having vital pulps Exclusion criteria: not stated Restorative treatments received (as translated): composite restorations of NCCLs Number of participants evaluated: 162; 162 teeth (rubber dam: 81; cotton rolls: 81) Withdrawals/loss to follow‐up: no losses to follow‐up
Interventions	Number of groups: 2 Intervention: rubber dam (as translated): "isolated with rubber dam (Optra Dam, Ivoclar Vivadent, 0.22 ˜ 0.27mm)" Control: cotton rolls (as translated): "isolated with cotton rolls placed in buccal and lingual vestibule"
Outcomes	Outcomes (as translated): failure rate Time points: 6 months after restorative treatment Diagnostic criteria: failure criteria (as translated): restorations found not to exist was regarded as failure. No further detail was provided
Notes	Adverse events: not stated Sample size calculation: no details reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote (as translated): "One hundred and sixty‐two patients with non‐carious cervical lesions were stratified randomly distributed into two groups (n = 81) from June 2009 to June 2011" Comment: method of sequence generation not stated. Insufficient information reported to make a judgement
Allocation concealment (selection bias)	Unclear risk	Not stated Comment: insufficient information reported to make a judgement
Blinding of participants and personnel (performance bias) All outcomes	High risk	Comment: operators and participants could not be blinded
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Not stated Comment: insufficient information reported to make a judgement
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: no loss to follow‐up
Selective reporting (reporting bias)	Low risk	Comment: outcomes reported as planned
Other bias	Unclear risk	Comment: no data on group comparability

ART: atraumatic restorative treatment; NCCLs: non‐carious cervical lesions; RCT: randomised controlled trial.