de Araujo Nobre 2007

Methods	Randomised, parallel group study. Patients and outcome assessors blind. No withdrawals.
Participants	Adults. No smokers. Four immediately loaded mandibular implants. 30 enrolled and results given for 30.
Interventions	2 groups. 0.2% hyaluronic acid gel (Gengigel) versus 0.2% chlorhexidine gel (Elugel) applied on a toothbrush as only oral hygiene measure. Recall every 2 months. Study duration: 6 months.
Outcomes	Implant failure. Probing pocket depth (mm), modified plaque index (mPI) by Mombelli and modified sulcus bleeding index (mBI) by Mombelli at 10 days, 2, 4 and 6 months. 6‐month data used.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	After correspondence with the author: "The randomisation procedure was performed by allocating the patients by order of admission with the use of a pre‐determined random numbers table".
Allocation concealment (selection bias)	Low risk	After correspondence with the author: "Group allocation was concealed to the investigators during the study’s data collection and analysis. The investigators only knew in which group the patients were included after the conclusion of the statistical analysis".
Blinding (performance bias and detection bias) All outcomes	Low risk	"Investigator blinding: "The patients data was separated in 2 databases with an encryption file (database 1: patients personal id and group allocation; database 2: clinical data). The investigators only knew in which group the patients were included after the conclusion of the statistical analysis".
Incomplete outcome data (attrition bias) All outcomes	Low risk	No loss to follow‐up.
Selective reporting (reporting bias)	Low risk
Other bias	Low risk
Blinding of the patient?	Low risk	After correspondence with the author: "Patient blinding was performed by giving to the patients blank cartridges (with no information on them discriminating the product or products' characteristics").