Reliable labels, uncertain meaning: Classifications for medial meniscus posterior root tears lack consensus and arthroscopy-anchored accuracy.

Abstract

Multiple, nonharmonised classification systems exist for posterior medial meniscus root tears (PMMRT), yet their diagnostic accuracy and observer reliability remain unclear. This PRISMA-2020 systematic review aimed to catalogue available PMMRT classifications and to summarise accuracy and reliability metrics. Searches (PubMed, Scopus, Cochrane CENTRAL, Ovid MEDLINE/Embase; English; humans; no date/design limits; 9 September 2025) identified 2849 records; 1920 unique records were screened and 228 full texts assessed; six studies were eligible. Risk of bias followed QUADAS-2 for diagnostic accuracy and GRRAS/COSMIN for reliability. Study-specific interobserver κ and measurement ICC (with 95% confidence interval [CIs]) were extracted. Arthroscopy-validated diagnostic accuracy was available from one study; no pooling was performed. One arthroscopy-validated MRI study reported high accuracy for radial PMMRT (sensitivity 0.90; specificity 0.94). Interobserver agreement for MRI-based classification/diagnosis was high in two studies (κ = 0.86-0.93). MRI measurement repeatability for extrusion/tear-gap was good-to-excellent (ICC: 0.80-0.92). No study prospectively linked subtype to treatment or outcomes. MRI-based PMMRT classifications show high interobserver agreement, but arthroscopy-validated accuracy evidence remains limited (single study); no system can be endorsed as most robust. A consensus, outcome-anchored anatomical taxonomy and multicenter, arthroscopy-validated studies linked to clinical outcomes are required before standardisation. Level III. PROSPERO registration number: CRD420251103590.