Gulsen 2017.

Study characteristics
Methods	Study design: split‐mouth RCT Conducted in: Turkey
Participants	Inclusion criteria: fit study requirements including follow‐ups and informed consent; healthy without significant medical diseases or a history of bleeding problems; symmetrical impacted third molars with same level of surgical difficulty, requiring the same surgical technique to be performed; third molars in Class I, Level B position (according to Pell and Gregory) and in vertical positions according to Winter. Exclusion criteria: pregnant and lactating women; signs of pericoronitis; chronic use of medications such as antihistamines, non‐steroidal anti‐inflammatory drugs, steroids, and antidepressants, which would complicate the evaluation of their postoperative response. Age: between 17 to 27 years Number randomised: 30 participants/60 teeth
Interventions	PRF versus none Intervention group (n = 30): on the intervention side, the socket was filled with 3 pieces of PRF membrane, and the flap was primarily closed with 3–0 silk sutures Control group (n = 30): none
Outcomes	Facial swelling was determined by measuring distances from gonion ‐ commissura labiorum, tragus – commissura labiorum and tragus – lateral canthus. Pain evaluated with a VAS, with endpoints of 0 (no pain) to 100 (worst pain).
Notes	Infection was noted in 3 participants who did not have PRF. Sample size calculation: not reported Infection was not stated as an intended outcome, but was reported (infection not defined in the paper).
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "study sides and control sides randomly selected" Comment: method of randomisation not described
Allocation concealment (selection bias)	Unclear risk	Comment: concealment methods not described. It is not mentioned how the side of the intervention was selected and whether it was concealed or not.
Blinding (performance bias and detection bias) patient	Unclear risk	Comment: participants were informed of the nature of the surgical and experimental procedures. The side in which the PRF was inserted into may have been known to the participant, but this is unclear.
Blinding (performance bias and detection bias) assessor	Unclear risk	Quote: "For standardization all measurements were performed by the same surgeon (UG)." Comment: measurements were all done by the same surgeon (UG), but it is not described if "UG" was also the operating surgeon
Incomplete outcome data (attrition bias) All outcomes	Low risk	No dropouts reported.
Selective reporting (reporting bias)	Low risk	All planned outcomes were reported, and infection incidence was also reported, although not as a primary outcome variable.
Other bias	High risk	Quote: "Bilateral removal of the third molar was performed in a single appointment. For the study side, the sockets were filled with PRF, whereas for the control side, the sockets were left empty." Comment: in this split‐mouth study, both procedures were carried out at the same time. This may have affected the participant perception of pain.