Gauffin 2014.

Study characteristics
Methods	Study design: single centre, parallel‐group, two‐arm, randomised, controlled trial Setting: orthopaedic department at the Linkoping University hospital, Sweden Trial time period: participants were enrolled between 4 March 2010 and 5 April 2012 Interventions: arthroscopic surgery plus unsupervised exercise versus unsupervised exercise Sample size calculations: the article did not describe how the sample size was estimated in this study. Analysis: intention‐to‐treat analysis was used; for those crossing over to the other group, an 'as‐treated' analysis was performed.
Participants	Number of participants Number screened: 179 (24 ineligible, 5 declined participation) Number of participants at enrolment: 150 Number randomised: 150: 75 in the surgery group and 75 in the non‐surgery group Number included in analyses: 123 participants were included in the 3‐month analysis; 130 participants were included in the 12‐month analysis; 119 participants were included in the 3‐year analysis; 102 participants were included in the 5‐year analysis (66 in arthroscopy group, 36 in exercise group). 9 participants in the arthroscopy group did not receive surgery. By 5 years, 19 participants crossed over from the exercise group to the arthroscopy group. 4 participants were excluded from the 5‐year analysis because they underwent arthroscopic surgery in their index knee after the 12‐month follow‐up (n = 1 in arthroscopy group who hadn't had surgery as allocated and n = 3 in the exercise group). Inclusion criteria Age 45 to 64 years with suspected meniscal injury Symptom duration more than 3 months, including sudden pain onset and may include catching or locking of the joint (for < 2 seconds) Standing X‐ray with Ahlback grade of 0 (less than 50% reduction of the joint space, without consideration of possible osteophytes) Had undergone prior physiotherapy Could understand the Swedish language Exclusion criteria Had a locked knee or joint Locking for more than 2 seconds more often than once a week Rheumatic or neurological disease Fibromyalgia Replacement of hip or knee joints Contraindication for day surgery at the current unit (BMI > 35 or a serious medical illness) Baseline characteristics Knee arthroscopy (N = 75) Mean (SD) age: 54 (5) years Number of men and women: M/F = 53/22 Median (range) duration of knee pain (months): 7 (8) No. (%) expectations of treatment: 70 (97) Kellgren–Lawrence grade — no. (%): 0 ‐ 37 (49); 1 ‐ 34 (45); 2 ‐ 4 (5) No. (%) sudden onset of pain: 45 (61) No. (%) daily joint catching: 45 (61) No. (%) joint locking for > 2 seconds: 18 (24) No. (%) mod to high physical activity level (PAS 4‐6): 23 (32) Mean (95% CI) KOOS 5 subscales: Pain 55 (51‐59); Symptoms 59 (55‐62); Activities of Daily Living (ADL) 65 (61‐69); Sports 29 (25‐34); QOL 34 (30‐37) Mean (95% CI) EQ‐5D: Index 0.63 (0.57‐0.68); VAS 62 (58‐67) Exercise therapy (N=75) Mean (SD) age: 54 (6) years Number of men and women: M/F = 56/19 Median (range) duration of knee pain (months): 7 (7) No. (%) expectations of treatment: 67 (92) Kellgren–Lawrence grade — no. (%): 0 ‐ 32 (43); 1 ‐ 36 (48); 2 ‐ 7 (9) No. (%) sudden onset of pain: 34 (47) No. (%) daily joint catching: 41 (56) No. (%) joint locking for > 2 seconds: 11 (15) No. (%) mod to high physical activity level (PAS 4‐6): 23 (32) Mean (95% CI) KOOS 5 subscales: Pain 58 (54‐62); Symptoms 62 (57‐66); Activities of Daily Living (ADL) 68 (63‐73); Sports 31 (26‐37); QOL 35 (31‐39) Mean (95% CI) EQ‐5D: Index 0.62 (0.56‐0.68); VAS 64 (59‐69) Pre‐treatment group differences: there were no differences in the baseline characteristics between the two groups
Interventions	Arthroscopic surgery Arthroscopic surgery plus unsupervised exercise program. All operations were performed with full or local anaesthetics by an experienced arthroscopist at an independent daycare clinic. After the arthroscope was inserted in the joint and the joint was visually inspected, the surgeon judged, according to their experience, whether a meniscal resection or any other surgical treatment was indicated. After surgery, all participants were allowed immediate, full weight‐bearing activity. The participants were advised to resume the exercise programme according to phase 1 for 1 week, and then switch to phase 2. Phase 1 program was for 1 week and consisted of a brisk walk 20‐30 min, 10 x 2 sets of squats, pelvic lifts, pelvic lifts with ball between knees and extension of one knee, heel raise, wall squat, and standing on a pillow on one leg (30 sec x 2). Phase 2 was done twice a week for 3 months and consisted of 3 sets of all of the above‐mentioned exercises. Exercise Unsupervised exercise program. At an independent clinic, five physiotherapists experienced in knee rehabilitation gave individual instructions for the exercise programme. The exercise programme aimed to increase muscle function and postural control and was done twice a week for 3 months, unsupervised. Participants could exercise either at the gym or at home. Compliance was monitored with self‐reported exercise diaries.
Outcomes	Outcomes were measured at baseline, at 3 and 12 months, 3 years and 5 years of follow‐up. Primary outcome KOOS (Knee injury and Osteoarthritis Outcome Score) Pain subscale at 12 months, where scores range from 0 to 100 (100 indicates good knee function and less pain) Secondary outcomes KOOS subscales (pain, symptoms, ADL, sports/recreation, QoL); scores range from 0 to 100 where higher scores indicate better knee function Health‐related quality of life using EQ‐5D Index, a single index summarising responses in 5 dimensions (mobility, self‐care, usual activities, pain/discomfort, anxiety/depression). EQ‐5D VAS scores range from 0 (worst imaginable health state) to 100 (best imaginable health state) Physical activity scale (PAS), a 6‐point Likert scale ranging from "1: no physical activity" to "6: heavy physical activity several times a week" Functional tests used at 3 months: squatting (possible without pain, possible with pain, not possible). SOLEC (standing on one leg with eyes closed test), 30‐second chair stand test on one leg (maximum repetitions in 30 sec). Symptom satisfaction scale, a 6‐point Likert scale that ranged from “delighted” to “terrible" Participant expectations about recovery, a 4‐point Likert scale that ranged from "no recovery" to "full recovery" Clinically important change in pain ‐ improved (> 10‐point higher KOOS Pain subscore at 5 years compared with baseline); stable (KOOS Pain subscore at 5 years within 10 points from baseline); deteriorated (> 10‐point lower KOOS Pain subscore at 5 years compared with baseline) Adverse events ‐ electronic medical charts checked at 12 months and 3 years; and via questionnaire at 5 years Radiographic changes at 5 years ‐ radiographic OA defined as grade equal or greater than Kellgren‐Lawrence grade 2 (definite osteophytes and possible joint space narrowing). Kellgren‐Lawrence classification (grade 1 (doubtful): doubtful joint space narrowing and possible osteophytic lipping; grade 2 (minimal): definite osteophytes and possible joint space narrowing; grade 3 (moderate): moderate multiple osteophytes, definite narrowing of joint space and some sclerosis and possible deformity of bone ends; grade 4 (severe): large osteophytes, marked narrowing of joint space, severe sclerosis and definite deformity of bone ends). Deterioration in radiographic findings also assessed. Outcomes included in this review at 3 and 12 months and at 5 years: Pain measured on KOOS Pain subscale (0 to 100; higher score = less pain) (we multiplied the mean values by –1 so that lower scores = less pain, as per Cochrane Handbook guidance) Function measured on KOOS ADL subscale (0 to 100; higher score = better function) Knee‐specific health‐related quality of life measured on KOOS QoL subscale (0 to 100; higher score = better QoL) Generic health‐related quality of life measured on EQ‐5D Index (0 to 1, higher score = better QoL) Participant‐reported treatment success measured as improvement in the KOOS‐Pain subscore of > 10 points at 5 years compared with baseline Serious adverse events Total adverse events Progression of knee OA
Notes	Funding: no funding source reported Trial registration: NCT01288768 Adverse events: Arthroscopic surgery Serious adverse events: No.(%): 3/66 (4.5%) Nature of event: two repeat arthroscopies in the surgery group ‐ one at 10 months and the other at 21 months after intervention. One participant died three years after surgery; it was not reported whether it was related to the intervention. Other adverse events: none reported Total adverse events: No.(%): 3/66 (4.5%) Exercise therapy Serious adverse events: none reported Other adverse events: none reported Total adverse events: none reported Knee surgery (replacement or osteotomy): not reported Progression of knee OA: radiographic deterioration from baseline to the 5‐year follow‐up, assessed according to the Kellgren‐Lawrence grade, occurred in 33/55 (60%) of the arthroscopy group and 10/27 (37%) of the exercise group. Withdrawals: no withdrawals were reported in this study; however, 9/75 in the arthroscopy group and 18/75 in the exercise group did not complete the 3‐month analysis; 5/75 in the arthroscopy group and 15/75 in the exercise group did not complete the 12‐month analysis; 9/75 in the arthroscopy group and 39/75 in the exercise group did not complete the 5‐year questionnaire and 20/75 in the arthroscopy group and 48/75 in the exercise group did not complete the 5‐year weight‐bearing radiographs. Reasons for lack of outcome data were not reported by group. Data analysis: ITT data were used for 3‐ and 12‐month follow‐up but only as‐treated data were reported at 5‐year follow‐up for outcomes: pain (Analysis 2.1), function (Analysis 2.2) and knee‐specific and generic quality of life (Analysis 2.3; Analysis 2.4) and participant reported success (Analysis 2.5).
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No description of sequence generation process provided
Allocation concealment (selection bias)	Low risk	"The orthopaedic surgeon who enrolled and assessed participants was blinded to the allocation sequence. The allocations were placed in sequentially numbered, opaque, sealed envelopes in 15 blocks, block size 10"
Blinding of participants and personnel (performance bias) All outcomes	High risk	Participants were not blinded to group assignments. Blinding of physiotherapists was attempted but some participants revealed their group. Surgeons were not blinded
Blinding of outcome assessor Self‐reported outcomes	High risk	Participants were not blinded to group assignments; thus, there was a risk of bias in the measurement of pain, function, knee‐specific quality of life, generic quality of life
Blinding of outcome assessor Assessor‐reported outcome (knee replacement)	Low risk	It is unclear whether outcome assessors checking electronic medical files for adverse events at 12 months were blind to group allocation, but we judged outcome measurement unlikely to be influenced by lack of blinding. Blinding of radiologist assessing X‐rays not reported. Blinding of surgeon assessing X‐rays not done. Blinding of physiotherapists assessing functional tests (not used in our analysis) attempted but broken
Incomplete outcome data (attrition bias) All outcomes	High risk	9/75 in the arthroscopy group and 18/75 in the exercise group did not complete the 3‐month analysis; 5/75 in the arthroscopy group and 15/75 in the exercise group did not complete the 12‐month analysis; 8/74 in the arthroscopy group and 36/72 in the exercise group did not complete the 5‐year analysis (4 participants underwent surgery in the affected knee and were excluded from the study after 1 year). In the 3‐month assessment, 14 participants (per group information was not given) completed the questionnaire at 5 months and these data were excluded
Selective reporting (reporting bias)	High risk	Authors have not collected or reported Tegner activity scale and have collected but not reported KOOS total score (secondary outcomes in the trial registration form)
Other bias	Low risk	No other bias apparent