Renvert 2009

Methods	Randomised, parallel group study. One implant considered per patient. Patients cannot be blind, outcome assessors blind. 2 withdrawals (1 after therapy, 1 at 3 months) from the hand instrument group; 4 withdrawals (2 after therapy; 1 after 1 months, 1 after 3 months) from the ultrasonic device group.
Participants	Adults. Mean age 61.5; SD 12.4. 37 enrolled, 31 analysed.
Interventions	Mechanical debridement using titanium curettes (Deppeler SA, Rolle, Switzerland) versus debridment with an ultrasonic device (the Vector systems, Dürr Dental AG, Bietigheim‐Bissingen, Germany) with a specially designed tip (LM Instruments Oy, Parainen, Finland). All implants were polished with rubber cups and polishing paste. Oral hygiene instructions on an individual basis and at all study time points (1, 3 and 6 months).
Outcomes	Presence/absence of hyperplasia, full‐mouth plaque score (presence/absence in percentage within each subject ‐ four sites per tooth and implant), local plaque score at four sites of the treated implant, PPD at the worst site of implant, mean PPD based on scores from four sites per implant, presence /absence of BOP at the implant, bleeding appearing after PPD measurements.
Notes	Described as double blind, but only assessors were blind. Sample size calculation assuming a 0.6 mm of difference between treatments. 18 subjects to be included in each group, 17 and 14 analysed.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"The allocation was carried out using a computer software program (SPSS Inc.) for the randomisation".
Allocation concealment (selection bias)	Low risk	After correspondence with the author: "Closed envelopes were opened when treatment had to be performed".
Blinding (performance bias and detection bias) All outcomes	Low risk	"Study subjects were instructed not to discuss the therapy with the study examiner. The study examiner was unaware of study treatment allocation and performed all clinical measurements".
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	6 losses to follow‐up.
Selective reporting (reporting bias)	Unclear risk	Hyperplasia reported in the outcome measures, but not in the results.
Other bias	Low risk
Blinding of the patient?	Low risk	Not possible. This was judged to be irrelevant for the risk assessment.