Merchan 1993.

Study characteristics
Methods	Study design: single centre, parallel‐group, two‐arm, open‐label, randomised controlled trial Setting: participants recruited from outpatient clinic of an orthopaedic hospital in Spain Trial time period: January 1988 to December 1990 Interventions: arthroscopic surgery plus physiotherapy versus non‐steroidal anti‐inflammatory drugs (NSAIDs) plus physiotherapy Sample size calculations: a priori sample size calculation not described Analysis: intention‐to‐treat
Participants	Number of participants Number of participants screened: not reported Number of participants at enrolment: not reported Number randomised: 80: 40 to arthroscopic surgery and physiotherapy group and 40 to non‐steroidal anti‐inflammatory drugs and physiotherapy group Number included in the analysis: data for 35/40 (87.5%) for the arthroscopic surgery and physiotherapy group and 38/40 (95%) for the non‐steroidal antiinflammatory drugs and physiotherapy group were available at the 1‐, 2‐ and 3‐year follow‐ups Inclusion criteria Age > 50 years Painful limited degenerative osteoarthritis of the femorotibial joint Limited degenerative process (minimal joint space narrowing and formation of small osteophytes) on pre‐operative radiographs evaluated according to Ahlback 1968 Exclusion criteria Duration of pain > 6 months Body weight > 85 kg in men and > 70 kg in women History of previous surgery of the affected knee Participants with an appreciable instability or an angular deformity of > 15 degrees on standing antero‐posterior radiographs for varus and valgus femorotibial angulation Participants with femoro‐patellar joint involvement Baseline characteristics Arthroscopic surgery and physiotherapy group (N = 35) Mean (range) age: 57.1 (50 to 63) years No. of men/women: 7/28 Mean Inital Knee Score (IKS) calculated using the modified Hospital for Special Surgery Knee Rating Score: 26.85 Mean Final Knee Score (FKS) calculated using the modified Hospital for Special Surgery Knee Rating Score: 37.00 Mean Knee Score Difference (KSD = FKS‐IKS): 10.14 Non‐steroidal anti‐inflammatory drugs and physiotherapy group (N = 38) Mean (range) age: 56.9 (50 to 65) years No. of men/women: 13/25 Mean IKS: 29.86 Mean FKS: 32.76 Mean KSD: 2.89 Pre‐treatment group differences: the initial knee score (IKS), age and sex did not show statistical differences between the two groups.
Interventions	Arthroscopic surgery Arthroscopic surgery and physiotherapy. The surgical technique included debridement of synovial tissue; removal of degenerative menisci, osteophytes, and loose bodies; and limited debridement of cartilage defects. Osseous debridement to bleeding bone (abrasion arthroplasty) was not performed. The mean operative time was 50 minutes (range 35 to 70). Post‐operatively, a compression bandage was used with early exercises, motion, and weight bearing as tolerated. The operative findings of articular cartilage changes were graded according to Outerbridge (Outerbridge 1961). Superficial fibrillation was present in 8 participants, fragmentation of < 1.3cm2 in 16, fragmentation of > 1.3cm2 in 7, and eburnation to subchondral bone in 4. There were meniscal tears in 31 of the 35 knees with a ratio of medial to lateral of 4 to 1. The most common tear was a flap of the posterior horn of the medial meniscus. Osteophytes were excised from the intercondylar notch in 6 knees, loose bodies were removed in 7, and 4 were noted to have chondrocalcinosis. Physiotherapy was practiced for 4 weeks after surgery. The physiotherapy regimen included quadriceps exercises and knee flexion exercises immediately postoperatively. Non‐steroidal anti‐inflammatory drugs Non‐steroidal anti‐inflammatory drugs and physiotherapy. The non‐operative treatment consisted of non‐steroidal anti‐inflammatory drugs and a decrease in the intensity of the activities of daily living for a pain‐free knee. Physiotherapy was practiced as in the operative group (i.e. quadriceps and knee flexion exercises for 4 weeks). This non‐operative group had no further courses of treatment during the follow‐up period.
Outcomes	The mean follow‐up time was 25 months (range 12 to 36) in the arthroscopic surgery and physiotherapy group and 23 months (range 12 to 36) in the non‐steroidal anti‐inflammatory drugs and physiotherapy group. Outcomes Modified Hospital for Special Surgery Knee Rating Score (mHSSKRS) (0 to 100 points, higher score = better). mHSSKRS has two subscores: subjective score = 60 points (includes pain = 35, instability = 10, walking aids = 5, walking distance = 10); and objective score = 40 points (includes extension = 10, flexion = 20, effusion = 10). (Only group means were reported. No measures of variability reported) Treatment success defined as an increase in the post‐treatment knee rating score of at least 10 points. A failure was defined as a knee score decreasing or failing to increase by 10 points. Adverse events Outcomes used in this review at mean of 23 to 25 months Participant‐reported treatment success (at least 10 point increase in post‐treatment mHSSKRS) Serious adverse events Total adverse events Measures of variability for function outcomes using mHSSKRS were not reported so could not be used in this review.
Notes	Funding: Fundacion Caja de Madrid Trial registration: not done Adverse events: Arthroscopic surgery and physiotherapy group Serious adverse events: Death: 5/40 (12.5%) participants died after randomisation and were excluded from follow‐up. No further information provided. Deep vein thrombosis: 2/40 (5%) Other adverse events: Superficial infection: 1/40 (2.5%) Haemarthrosis: 1/40 (2.5%) Total ALL adverse events: 9/40 (22.5%) Non‐steroidal anti‐inflammatory drugs and physiotherapy group Serious adverse events: Death: 2/40 (5%) participants died after randomisation and were excluded from follow‐up. No further information provided. No other serious adverse events reported Other adverse events: none reported Total ALL adverse events: 2/40 (5%) Knee surgery (replacement or osteotomy): not reported Progression of knee OA: not reported Withdrawals: 5/40 in the arthroscopic surgery and physiotherapy group and 2/40 in the non‐steroidal anti‐inflammatory drugs and physiotherapy group due to death
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No description of sequence generation process provided. There were some baseline differences in outcome measures between the treatment groups.
Allocation concealment (selection bias)	Unclear risk	Allocation performed by "pulling consecutively numbered envelopes that had previously been randomly placed on a bulletin board". Unclear if sealed or opaque
Blinding of participants and personnel (performance bias) All outcomes	High risk	No information provided. Probably not done
Blinding of outcome assessor Self‐reported outcomes	High risk	No information provided. Probably not done. Measurement of mHSSKRS subjective score likely to be influenced by lack of blinding
Blinding of outcome assessor Assessor‐reported outcome (knee replacement)	High risk	No information provided. Probably not done. Measurement of mHSSKRS objective score and treatment success likely to be influenced by lack of blinding. Adverse events unlikely to be influenced by lack of blinding
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Imbalance in withdrawals across groups: 12.5% (5/40) due to death in the arthroscopic surgery plus physiotherapy group vs. 5% (2/40) due to death in the NSAIDs plus physiotherapy group. The cause of deaths in both groups were not reported.
Selective reporting (reporting bias)	Unclear risk	Trial registration not done and protocol not available. Insufficient information to judge high or low risk
Other bias	Low risk	No other bias apparent