Test | Median (range) | High predictive value (pv≥0.85) | Moderate predictive value (0.84≥pv≥0.70) | Low predictive value (pv≤0.69) |
Diagnosis in patients with knee pain and clinical suspicion of meniscal tear | ||||
Joint line tenderness | ||||
Medial+Lateral | PPV 0.23 (0.22–0.23) | Low | ||
NPV 0.87 (0.85 – 0.90) | Low | |||
Medial | PPV 0.42 (0.23–0.56) | Low | ||
NPV 0.93 (0.84– 0.98) | Low | |||
Lateral | PPV 0.84 (0.65–1.00) | Moderate | ||
NPV 0.93 (0.87–0.99) | Moderate | |||
McMurray | ||||
Medial + Lateral | PPV 0.37 (0.23– 0.47) | Very low | ||
NPV 0.86 (0.85–0.94) | Very low | |||
Medial | PPV 0.35 (0.20–0.65) | Very low | ||
NPV 0.88 (0.84–0.95) | Very low | |||
Lateral | PPV 0.52 (0.20–0.69) | Very low | ||
NPV 0.90 (0.84–0.94) | Very low | |||
Thessaly (20°) | ||||
Medial+Lateral | PPV 0.57 (0.24–0.90) | Very low | ||
NPV 0.92 (0.86– 0.98) | Very low | |||
Medial | PPV 0.79 (0.21–0.87) | Very low | ||
NPV 0.92 (0.84 – 0.97) | Very low | |||
Lateral | PPV 0.68 (0.20–0.84) | Very low | ||
NPV 0.95 (0.83–0.98) | Very low | |||
Thessaly (5°) | ||||
Medial | PPV 0.23 | Very low | ||
NPV 0.83 | Very low | |||
Lateral | PPV 0.24 | Very low | ||
NPV 0.82 | Very low | |||
Apley | ||||
Medial+Lateral | PPV 0.41 | Very low | ||
NPV 0.95 | Very low | |||
Medial | PPV 0.38 (0.28–0.58) | Very low | ||
NPV 0.90 (0.87–0.95) | Very low | |||
Lateral | PPV 0.47 (0.41–0.77) | Very low | ||
NPV 0.89 (0.86–0.95) | Very low | |||
McMurray and Thessaly (20°) | ||||
Medial+Lateral | PPV 0.25 | Very low | ||
NPV 0.85 | Very low | |||
McMurray and joint line tenderness | ||||
Medial | PPV 0.40 | Very low | ||
NPV 0.99 | Very low | |||
Lateral | PPV 0.75 | Very low | ||
NPV 0.98 | Very low | |||
McMurray, joint line tenderness and Apley* | ||||
Medial | PPV 0.40 | Very low | ||
NPV 0.97 | Very low | |||
Lateral | PPV 0.70 | Very low | ||
NPV 0.90 | Very low |
PPVs and NPVs are calculated based on data from the systematic search for the Danish national clinical guideline for meniscal pathology,12 and from an updated systematic search. The calculations are based on a prevalence of MRI-verified meniscal tears of 19% (women aged 50–59 years).7 Since PPV and NPV are dependent on the prevalence of meniscal tears, and since the prevalence of MRI-verified meniscal tears varies according to gender and age, this table is primarily relevant for younger patients.
*Study by Ercin et al 45 was not included as the combination of tests in that study included additional tests.
NPV, negative predictive value; PPV, positive predictive value.