Saeed 2015.

Study characteristics
Methods	Study design: single centre, two‐arm, randomised controlled trial Setting: Department of Orthopedics, Ch. Rehmat Ali Memorial Trust Hospital attached to Continental Medical College Lahore, Pakistan Trial time period: January 2012 to December 2014 Interventions: arthroscopic surgery versus five intra‐articular hyaluronic acid injections given at weekly intervals Sample size calculations: not reported Analysis: intention‐to‐treat
Participants	Number of participants Number of participants screened: not reported Number of participants at enrolment: 120 Number randomised: 120: 60 in the arthroscopic debridement group and 60 in the intra‐articular hyaluronic acid injection group Number included in analyses: 120: 60 in the arthroscopic debridement group and 60 in the intra‐articular hyaluronic acid injection group Inclusion criteria Age > 40 years History of pain in the knee joint Grade 2 & 3 Kellgren‐Lawrence grading system Exclusion criteria Age < 40 years History of injury or accident Prior intervention with intra‐articular glucocorticoid injections within the past three months Baseline characteristics Arthroscopic surgery No. (%) women: 48 (80) No. (%) men: 12 (20) No. (%) of Kellgren‐Lawrence Grade 2 ‐ 38 (63.33), Grade 3 ‐ 22 (36.66) No. (%) Pain score measured on the Knee Society Score System (0 = severe, 50 = none): 10 = 4 (6.7), 20 = 42 (70), 30 = 14 (23.3) Intra‐articular hyaluronic acid injections No. (%) women: 50 (83.3) No. (%) men: 10 (16.6) No. (%) of Kellgren‐Lawrence Grade 2 ‐ 36 (60), Grade 3 ‐ 24 (40) No. (%) Pain score measured on the Knee Society Score System (0 = severe, 50 = none): 10 = 8 (13.4), 20 = 26 (43.3), 30 = 26 (43.3) Pre‐treatment group differences: there were some differences between groups in pain at baseline: in the intra‐articular hyaluronic acid injection group, 13% of participants had pain score 10, 43% had 20 and 43% had 30, while in the arthroscopic group, 7% of participants had pain score 10, 70% had 20 and 23% had 30.
Interventions	Arthroscopic surgery Arthroscopic debridement. Participants were admitted to the hospital, arthroscopic debridement was performed in the operation theatre by using two portals in all cases under spinal anaesthesia. Participants were discharged on the next day. Monitoring of electrocardiogram (ECG) and blood pressure was standard in all cases during the entire duration of the procedure. All the debridements were done by a single surgeon to minimise the bias for the study. Intra‐articular hyaluronic acid injections Participants were injected with intra‐articular hyaluronic acid after being given intradermal anaesthesia. The injections were given weekly for five weeks with a 24‐gauge needle under strict aseptic conditions in the operation theatre as an outpatient. In case of joint effusion, aspiration was done before the injection to prevent dilution of the injection.
Outcomes	Outcome were assessed at baseline and at 1, 3 and 6 months' follow‐up Outcomes Knee Society Score System (KSSS) assessing knee pain scores; range from 0 ‐ severe to 50 ‐ none. This study categorised the scores into KSSS 20, KSSS 30, KSSS 40, KSSS 45 Adverse events Outcomes used in this review at 3 and 6 months Pain measured on KSSS, score 30 or higher Total adverse events
Notes	Funding: not reported Trial registration: not reported Adverse events: Arthroscopic debridement group Serious adverse events: none reported Other adverse events: No.(%): 13 (26) Nature of event: pain and mild effusion Total adverse events: No.(%): 13 (26) Intra‐articular hyaluronic acid injection group Serious adverse events: none reported Other adverse events: No.(%): 8 (13.4) Nature of event: pain at injection site Total adverse events: No.(%): 8 (13.4) The frequency of adverse events in the arthroscopy group was higher than the injection group Knee surgery (replacement or osteotomy): not reported Progression of knee OA: not reported Withdrawals: none
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Described as 'randomized experimental study' but no information about sequence generation process provided
Allocation concealment (selection bias)	Unclear risk	No information on whether allocation concealment was done or not
Blinding of participants and personnel (performance bias) All outcomes	High risk	Blinding of participants and study personnel was not reported. Probably not done due to the nature of the intervention
Blinding of outcome assessor Self‐reported outcomes	High risk	Risk of detection bias in assessment of participant‐reported knee pain and adverse events as blinding of participants was not reported but probably not done
Blinding of outcome assessor Assessor‐reported outcome (knee replacement)	Low risk	Blinding of study personnel not reported but adverse effects unlikely to be influenced by lack of blinding
Incomplete outcome data (attrition bias) All outcomes	Low risk	No withdrawals or loss to follow‐up
Selective reporting (reporting bias)	High risk	Trial registration not done and protocol not available. The authors measured overall knee function (using the Knee Society Score System), but it is unclear if they reported overall scores or a pain subscore.
Other bias	Low risk	No other bias apparent