Skip to main content
. 2022 Mar 3;2022(3):CD014328. doi: 10.1002/14651858.CD014328

Van de Graaf 2018.

Study characteristics
Methods Study design: non‐inferiority, parallel‐arm, multicentre, randomised controlled trial
Setting: nine hospitals in the Netherlands
Trial time period: July 2013 to November 2015
Interventions: arthroscopic partial meniscectomy (APM) versus physical therapy (PT)
Sample size calculations: initially a sample size of 402 participants was estimated to have power of 90%, an α of 0.05 and SD of 20 points; a clinically relevant difference of 8.8 points on the International Knee Documentation Committee (IKDC) ‘Subjective Knee Form’ was rounded down to a non‐inferiority threshold of 8 to increase the power. However, an interim analysis led to recalculation of SD to 18 points resulting in a sample size requirement of 320 participants (120 per group).
Analysis: intention‐to‐treat. As‐treated analysis was also conducted and results reported in 3 groups ‐ APM group, PT group and delayed APM group i.e. those who were randomised to the PT group but received APM during follow‐up.
Participants Number of participants

  • Number of participants screened: not reported

  • Number of participants at enrolment: 321

  • Number randomised: 321 participants: 159 to the arthroscopic partial meniscectomy (APM) group and 162 to the physical therapy (PT) group. 1 participant from each group withdrew immediately after randomisation and their data were not included in the analysis

  • Number included in analyses: 155 at 3 months, 151 at 6 months, 143 at 12 months and 141 at 124 months in the APM group; 158 at 3 months, 146 at 6 months, 136 at 12 months and 141 at 24 months in the PT group. 8/158 (5%) participants in the APM group refused surgery. 17/161 (10.5%) in the PT group did not complete the PT protocol. 35/161 (21.6%) participants in the PT group had APM within 6 months of randomisation.


Inclusion criteria

  • Age 45 to 70 years

  • Knee pain

  • Non‐obstructive meniscal tear confirmed by MRI


Exclusion criteria

  • Knee locking

  • Prior knee surgery

  • Instability caused by an anterior or posterior cruciate ligament rupture

  • Tumour suspected of malignancy, detectable on MRI

  • Severe osteoarthritis (Kellgren‐Lawrence score of 4)

  • BMI > 35

  • American Society of Anesthesiologists (ASA) 4‐5 participants

  • General disease that affects physical function or systemic medication/abuse of glucocorticoids

  • Any other medical condition or treatment interfering with the completion or assessment of the trial; for example, contraindications to MRI or surgery

  • Drugs or alcohol abuse

  • Participants unable to fill out the Dutch questionnaires

  • Associated injuries on the index knee consisting of symptomatic partial or total tear of the anterior cruciate ligament (ACL), posterior cruciate ligament tear, injury to the lateral or posterolateral ligament complex with significant laxity


Baseline characteristics
Arthroscopic partial meniscectomy group
Mean (SD) age: 57.6 (6.5)
No. of male/female: 78/80
Mean (SD) BMI: 26.7 (3.8)
No. (%) mechanical complaints: 56 (35.4)
No. (%) medial meniscal involvement: 126 (79.7)
No. (%) osteoarthritis score (KL classification): 0 = 18 (12.0); 1 = 81 (54.0); 2 = 45 (30.0); 3 = 6 (4.0)
Mean (SD) Knee function International Knee Documentation Committee (IKDC) score (0 = most limitations to 100 = no limitations): 44.8 (16.6)
Median (IQR) Knee pain on VAS (0 = no pain to 100 = worst pain): 61.1 (44.9‐83.4)
Physical therapy group
Mean (SD) age: 57.3 (6.8)
No. of male/female: 79/81
Mean (SD) BMI: 27.2 (4.0)
No. (%) mechanical complaints: 67 (41.6)
No. (%) medial meniscal involvement: 136 (84.5)
No. (%) osteoarthritis score (KL classification): 0 = 15 (10.1); 1 = 74 (49.7); 2 = 55 (36.9); 3 = 5 (3.3)
Mean (SD) Knee function IKDC score (0 = most limitations to 100 = no limitations): 46.5 (14.6)
Median (IQR) Knee pain on VAS (0 = no pain to 100 = worst pain): 59.3 (44.9‐77.4)
Pre‐treatment group differences: there are no pre‐treatment group differences between the groups.
Interventions Arthroscopic partial meniscectomy (APM)
Arthroscopic partial meniscectomy was performed within 4 weeks of randomisation in an outpatient clinic under general or spinal anaesthesia by orthopaedic surgeons experienced in arthroscopic surgery, or orthopaedic residents skilled in arthroscopic surgery under supervision of an orthopaedic surgeon. Standard anteromedial and anterolateral portals were introduced for inspection of the knee joint. The affected meniscus was partially removed until a stable and solid meniscus remained. All participants received perioperative instructions and a home exercise program. Participants were only referred to PT after APM if they did not recover as anticipated as defined by the Dutch Orthopedic Association guidelines.
Physical Therapy (PT)
Participants were referred to PT clinics and their initial PT session was scheduled within 2 weeks after randomisation. Participating PT clinics were instructed about the exercise protocol by a knee‐specialised physical therapist or the primary investigator, prior to the first participant’s referral. The PT exercise protocol was developed by a knee‐specialised physical therapist and consisted of 16 sessions of 30 minutes each conducted over 8 weeks. The PT protocol comprised cardiovascular, coordination/balance, and closed kinetic chain strength exercises (in which the distal part of the extremity is fixed to an object that is stationary). If PT failed, the participant was allowed to attend additional PT sessions or have APM, depending on their preference.
Post‐intervention: both groups received the same home exercise instructions. The home exercise program consisted of one leg standing during 60 seconds and a step‐down exercise comprising 3, 9, 10 repetitions, twice a week.
Outcomes Outcomes were measured at 3, 6, 12 and 24 months
Primary outcome

  • Change in knee physical function from baseline to 2 years measured on the International Knee Documentation Committee (IKDC) Subjective Knee Form (scores range from 0 = most limitations to 100 = no limitations in daily and sports activities and the absence of symptoms)


Secondary outcomes

  • Knee pain on weight‐bearing and at rest measured on VAS (scores range from 0 = no pain to 100 = worst pain)

  • General health measured using SF‐36 (ranging from 0 = worst health to 100 = better health) (data only reported for Physical Component Score)

  • Quality of life measured using EQ‐5D (data not reported)

  • Radiographic progression of osteoarthritis using KL classification scores (range from 0 = no osteophytes or joint space narrowing indicating no osteoarthritis to 4 ≥ 50% joint space narrowing indicating severe osteoarthritis)

  • Activity measured on Tegner Activity Scale scores range from 0 = no activity to 10 = higher activity

  • Physical performance tests (squatting with duck‐walk, Thessaly test, McMurray test, range of motion, joint line tenderness, existence of knee joint effusion)

  • Participant‐specific complaints questionnaire

  • Participant expectation of treatment and their satisfaction (data not reported)

  • Knee replacement

  • Adverse events: minor, moderate, severe

  • Resource utilisation ‐ rehospitalisation, intervention and other healthcare costs, paid help at home, informal care, work absenteeism and presenteeism, unpaid productivity costs

  • Function measured using Patient Specific Functional Scale (PSFS)


Outcomes used in this review at 3, 6 and 24 months

  • Pain ‐ knee pain on weight‐bearing measured on VAS (0 to 100, lower score = less pain)

  • Function ‐ change in mean knee function from baseline to 2 years ‐ IKDC scores (0 to 100, higher score = better function)

  • Knee replacement

  • Serious adverse events: proportion in each group with serious adverse events

  • Total adverse events: proportion in each group with any adverse events

  • Progression of knee OA (to note: data couldn't be combined in Analysis 7.4 as mean KL scores were reported for each group and not number of those with OA progression)
Notes Funding: this study was funded by the Netherlands Organization for Health Research and Development (in Dutch: ZonMw; grant 837002009), Zilverenkruis Health Insurance (grant Z436), and the Foundation of Medical Research of the OLVG, Amsterdam (grant 15u.025).
Trial registration: ClinicalTrials.gov number NCT01850719
Adverse events
Arthroscopic partial meniscectomy
Serious adverse events:
No.of events = 9
No. (%) of participants affected = 9/159 (5.7)
Nature of event: neurological events including intracranial malignancy (1), lymph node malignancy (1), rectal polyp (1), knee replacement (2), arthroscopy in affected knee (2), arthroscopy in opposite knee (1), other knee surgery (1)
Other adverse events:
No.of events = 9
No. (%) of participants affected = 9 (5.6)
Nature of event: reactive arthritis (1), knee pain resulting in extra consultation (6), Pain in back, hip or foot (2)
Total adverse events:
No. (%): 18/159 (11.32)
Physical therapy
Serious adverse events:
No.of events = 8
No. (%) of participants affected = 8/162 (4.94)
Nature of event: acute myocardial infarction (1), sudden death (1), neurological event (1), alcoholic pancreatitis (1), arthroscopy (1), knee replacement (3)
Other adverse events:
No.of events = 4
No. (%) of participants affected = 4 (2.4)
Nature of event: Knee pain resulting in extra consultation (2) other musculoskeletal (2)
Total adverse events:
No. (%): 12/162 (7.41)
Knee surgery (replacement or osteotomy)
Arthroscopic partial meniscectomy:
No. (%) of participants = 2/159 (1.2)
Other surgery: 3/159 had a re‐arthroscopy, 1/159 'other' surgery
Physical therapy:
No. (%) of participants = 3/162 (1.8)
Other surgery: 1/162 arthroscopy
Progression of knee OA: OA severity in the arthroscopy group progressed from 1.3 points at baseline to 1.6 points at 24 months (MD 0.37 points, 95% CI 0.25 to 0.49) and in the physical therapy group from 1.3 points at baseline to 1.5 points at 24 months (MD 0.18 points, 95% CI 0.04 to 0.31). Mixed‐model analysis found no significant between‐group difference (0.10 points more progression in the arthroscopy group, 95% CI ‐0.05 to 0.26, P = 0.18)
Withdrawals: two participants (1 from each group) withdrew immediately after randomisation; 18/158 in the APM group and 14/161 in the PT group due to loss to follow‐up.
Cross‐overs: 47/161 (29%) participants assigned to the exercise group had arthroscopic surgery within two years' follow‐up, and 8/159 (5%) participants assigned to the arthroscopy group did not have the procedure
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Low risk Randomisation was performed using a computerised software program (TENALEA Clinical Trial Data Management system) in a 1:1 ratio using random blocks with a maximum block size of 6.
Allocation concealment (selection bias) Low risk Randomisation to groups was concealed as it was through an online program.
Blinding of participants and personnel (performance bias)
All outcomes High risk Participants, physicians, and physical therapists were not blinded.
Blinding of outcome assessor
Self‐reported outcomes High risk As participants were unable to be blinded due to the nature of the intervention, there could be a risk of bias in the reporting of pain and function.
Blinding of outcome assessor
Assessor‐reported outcome (knee replacement) Low risk The radiologists assessing X‐rays were blinded to treatment allocation so there is low risk of bias in the assessment of osteoarthritis.
Incomplete outcome data (attrition bias)
All outcomes Low risk Two participants (1 from each group) withdrew immediately after randomisation without providing a reason. 18/158 (11%) in the APM group and 14/161 (8.6%) in the PT group were lost to follow‐up at 24 months and excluded from the final analysis. Missing outcome data reasonably balanced across groups with similar reasons for missing data across groups
Selective reporting (reporting bias) Low risk Trial registered and protocol published. Four deviations from the protocol were clearly outlined in the publication of results ‐ recalculation of SD, inclusion of all time points in the measurement of primary outcome, change to mixed‐model analysis from longitudinal analyses and correction of protocol error of 10% loss to follow‐up to 20%. Some secondary outcomes listed in the protocol such as resource utilisation, health‐related quality of life, patient‐specific complaints, participant expectations, and participant satisfaction were not analysed and authors reported they will be analysed and reported separately. The authors reported median and IQR data; however, upon request, they provided mean and SD data which has been used in the analysis
Other bias Low risk No other bias apparent