Summary of findings: APM versus non-surgical comparators
| Outcome | Endpoint | Participants (RCTs) | Standardised MD (95% CI) | MD (95% CI) | I2 | GRADE quality* |
| (A) All patients (any type of meniscal tear with or without radiographic osteoarthritis) | ||||||
| Pain | Under 6 months | 434 (2 RCTs) | 0.18 (−0.00 to 0.37) | (KOOS pain) 4.46 better (0.22 to 8.71) | 0% | ⨁◯◯◯ Very low†‡ |
| 6–12 months | 943 (5 RCTs) | 0.22 (0.03 to 0.40) | 4.10 better (0.74 to 7.46) | 48% | ⨁⨁◯◯ Low† | |
| Over 12 months | 484 (3 RCTs) | 0.22 (0.04 to 0.40) | 4.43 better (0.72 to 8.14) | 0% | ⨁◯◯◯ Very low†‡ | |
| Function | Under 6 months | 561 (4 RCTs) | 0.08 (−0.08 to 0.25) | (Lysholm scale) 1.38 better (−1.54 to 4.31) | 0% | ⨁⨁◯◯ Low† |
| 6–12 months | 1050 (6 RCTs) | 0.18 (0.04 to 0.33) | 3.36 better (0.55 to 6.16) | 27% | ⨁⨁◯◯ Low† | |
| Over 12 months | 730 (5 RCTs) | 0.18 (0.04 to 0.33) | 2.59 better (0.29 to 4.89) | 0% | ⨁⨁◯◯ Low† | |
| Knee-specific QoL | Under 6 months | 129 (1 RCT) | 0.45 (0.10 to 0.80) | (KOOS QoL) 9.59 better (2.34 to 16.84) | NA | ⨁⨁◯◯ Low† |
| 6–12 months | 350 (3 RCTs) | 0.43 (0.10 to 0.75) | 10.36 better (3.58 to 17.14) | 56% | ⨁⨁◯◯ Low† | |
| Over 12 months | 245 (2 RCTs) | 0.30 (0.05 to 0.56) | 6.92 better (1.33 to 12.52) | 0% | ⨁◯◯◯ Very low†‡ | |
| Generic QoL | Under 6 months | None | NA | (EQ-5D VAS) NA | NA | NA |
| 6–12 months | 130 (1 RCT) | 0.01 (−0.34 to 0.35) | 0.00 better (−0.06 to 0.06) | NA | ⨁⨁◯◯ Low† | |
| Over 12 months | 113 (1 RCT) | 0.47 (0.10 to 0.85) | 0.10 better (0.02 to 0.18) | NA | ⨁◯◯◯ Very low†‡ | |
| (B) Patients with any type of meniscal tear in a non-osteoarthritic knee | ||||||
| Pain | Under 6 months | 306 (2 RCTs) | 0.16 (−0.25 to 0.57) | (KOOS pain) 3.28 better (−4.33 to 10.90) | 68% | ⨁◯◯◯ Very low†‡§ |
| 6–12 months | 402 (3 RCTs) | 0.35 (0.04 to 0.66) | 6.91 better (2.87 to 10.94) | 58% | ⨁◯◯◯ Very low†§¶ | |
| Over 12 months | 368 (3 RCTs) | 0.21 (0.00 to 0.42) | 4.15 better (0.14 to 8.16) | 0% | ⨁◯◯◯ Very low†‡ | |
| Function | Under 6 months | 411 (3 RCTs) | 0.11 (−0.13 to 0.35) | (Lysholm scale) 1.93 better (−2.40 to 6.26) | 33% | ⨁⨁◯◯ Low† |
| 6–12 months | 507 (4 RCTs) | 0.30 (0.06 to 0.53) | 5.31 better (1.12 to 9.51) | 44% | ⨁◯◯◯ Very low†‡ | |
| Over 12 months | 501 (4 RCTs) | 0.19 (0.01 to 0.36) | 2.69 better (−0.65 to 6.03) | 0% | ⨁⨁◯◯ Low† | |
| Knee-specific QoL | Under 6 months | 124 (1 RCT) | 0.52 (0.16 to 0.87) | (KOOS QoL) 10.42 better (3.36 to 17.48) | NA | ⨁⨁◯◯ Low† |
| 6–12 months | 244 (2 RCTs) | 0.59 (0.11 to 1.07) | 12.89 better (3.60 to 22.18) | 71% | ⨁⨁◯◯ Low† | |
| Over 12 months | 231 (2 RCTs) | 0.39 (0.13 to 0.65) | 8.40 better (2.86 to 13.95) | 0% | ⨁⨁◯◯ Low† | |
| Generic QoL | Under 6 months | None | NA | (EQ-5D VAS) NA | NA | NA |
| 6–12 months | 119 (1 RCT) | 0.06 (−0.30 to 0.42) | 0.01 better (−0.05 to 0.08) | NA | ⨁⨁◯◯ Low† | |
| Over 12 months | 105 (1 RCT) | 0.63 (0.23 to 1.02) | 0.13 better (0.05 to 0.21) | NA | ⨁◯◯◯ Very low†‡ | |
SMD in function was translated and re-expressed as an MD in the Lysholm knee scale, using the SD of 19.4 reported for patients with isolated meniscal tears.79
*GRADE rating explanations.41
†Risk of bias (very serious): no blinding of participants.
‡Imprecision (serious): wide CI including potential MCID (benefit) with MD less than MCID.
§Inconsistency (very serious): high level of heterogeneity (>50%) between studies.
¶Imprecision (very serious): wide CI including no benefit and potential MCID (benefit).
APM, arthroscopic partial meniscectomy; EQ-5D, EuroQol five dimensions questionnaire; GRADE, Grading of Recommendations Assessment, Development and Evaluation; I2, heterogeneity; KOOS, Knee injury and Osteoarthritis Outcome Score; MD, mean difference; QoL, quality of life; RCTs, randomised controlled trials; SMD, standardised mean difference; VAS, visual analogue pain score.