Open versus arthroscopic management of acute and chronic distal radioulnar joint instability: a systematic review and meta-analysis Authors: Galal Hegazy, Ahmed I. Hammouda, Mohamed Ramadan, Abdulhamid Elzoghby, Ehab Elzahed, Ibrahem El-Sebaey, Ahmed Elgeushy, Mahmoud Seddik Published: 2026 ABSTRACT BackgroundFoveal triangular fibrocartilage complex (TFCC) repair can restore distal radioulnar joint (DRUJ) stability, but whether open or arthroscopic techniques provide superior clinical outcomes remains uncertain. This systematic review and meta-analysis compared functional outcomes, DRUJ stability, and complications between open versus arthroscopic foveal TFCC repair for acute and chronic post-traumatic DRUJ instability.MethodsPubMed, Scopus, and Cochrane CENTRAL were searched initially in January 2026 and updated on 15 April 2026 to identify comparative studies of open versus arthroscopic TFCC foveal repair published from January 2000 onward. Adult patients undergoing open or arthroscopic TFCC foveal repair for DRUJ instability were included. Non-comparative studies, abstracts, and non-English publications were excluded. Extracted outcomes included Modified Mayo Wrist Score (MMWS), Disabilities of the Arm, Shoulder and Hand (DASH), Patient-Rated Wrist Evaluation (PRWE), Visual Analogue Scale (VAS) score for pain, grip strength, range of motion, recurrent instability, and complications. Random-effects meta-analysis pooled standardized mean differences (SMDs) for continuous outcomes and odds ratios (ORs) for binary outcomes. Risk of bias was assessed using ROBINS-I, and certainty of evidence using GRADE. Exploratory subgroup analysis by injury chronicity at study level was performed.ResultsFive observational studies (Level II–III evidence; 322 wrists: 181 open, 141 arthroscopic) met inclusion criteria. ROBINS-I rated two studies at moderate and three at serious risk of bias; GRADE certainty was low to very low across outcomes. Pooled results showed no statistical differences: MMWS (SMD − 0.12, 95% CI − 0.47 to 0.23; I² = 0%, τ² = 0; p = 0.719); DASH score (SMD 0.22, 95% CI − 0.96 to 1.40; I² = 86.7%, τ² = 0.48; p < 0.0001); PRWE (SMD − 0.24, 95% CI − 2.85 to 2.38; I² = 93.3%, τ² = 1.03; p < 0.0001); VAS (SMD − 0.35, 95% CI − 1.69 to 0.99; I² = 89.9%, τ² = 0.63; p < 0.0001); recurrent instability OR (1.00, 95% CI 0.19 to 5.37; I² = 63.2%, τ² = 1.18; p = 0.028). Exploratory subgroup synthesis restricted to predominantly chronic cohorts did not materially alter pooled estimates.ConclusionsBoth open and arthroscopic foveal TFCC repair can improve function, pain, motion, grip strength, and distal radioulnar joint stability, but current comparative evidence does not demonstrate clear superiority of either technique. Because the available evidence is limited to a small number of non-randomized studies with substantial heterogeneity and low certainty, the absence of statistically significant differences should be interpreted cautiously and should not be considered evidence of clinical equivalence. In addition, although SMDs were appropriate for the present meta-analysis given heterogeneity in outcome reporting, they reduce the clinical interpretability. Surgical choice should therefore remain individualized until higher-quality comparative studies become available.Trial registrationPROSPERO registration: CRD42025638530.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12891-026-09939-y. Background Distal radioulnar joint (DRUJ) instability is a debilitating disorder that impairs forearm rotation, compromises load transfer across the wrist [1, 2]. In most cases, symptomatic DRUJ instability results from lesions of the triangular fibrocartilage complex (TFCC), particularly disruption of the deep radioulnar fibers at the ulnar fovea, which are recognized as the primary soft-tissue stabilizers of the joint [3]. Biomechanical and clinical studies have shown that failure to restore foveal TFCC continuity may lead to persistent ulnar-sided wrist pain, decreased grip strength, restricted pronation–supination, and progressive degenerative change, underscoring the importance of anatomical reconstruction of this structure [4–7]. With advances in wrist arthroscopy, arthroscopic foveal repair has emerged as an alternative to traditional open foveal reinsertion, aiming to achieve anatomic restoration while minimizing surgical morbidity and allowing concomitant intra-articular pathology to be addressed [8]. Contemporary cohort studies suggest that both open and arthroscopic foveal repair can restore DRUJ stability and improve patient-reported outcomes; however, reported failure rates, complication profiles, and recovery trajectories vary, particularly when comparing acute versus chronic presentations [8–10]. Several systematic reviews [8, 10, 11] have compared open and arthroscopic techniques for TFCC repair, but these analyses have important limitations in the context of current practice. Specifically, they identified only a small number of eligible comparative studies precluding definitive conclusions about the superiority of either approach. The objective of this systematic review and meta-analysis was to evaluate, in patients with acute or chronic DRUJ instability due to foveal TFCC lesions, the comparative effectiveness of arthroscopic versus open foveal TFCC reattachment on functional outcomes, restoration of DRUJ stability, and complications. We hypothesized that both open and arthroscopic foveal TFCC repair would improve functional outcomes and restore DRUJ stability. Materials and methods This systematic review and meta-analysis was conducted according to PRISMA 2020 guidelines (PROSPERO CRD42025638530). Eligibility criteria Eligible studies enrolled adults (≥ 18 years) with traumatic DRUJ instability due to TFCC foveal detachment undergoing either open or arthroscopic foveal TFCC reattachment. Primary extracted outcomes were Disabilities of the Arm, Shoulder and Hand (DASH), Modified Mayo Wrist Score (MMWS), and Patient-Rated Wrist Evaluation (PRWE) scores. Secondary outcomes included pain (VAS/NRS), grip strength, range of motion, recurrent instability, and complications (DSBUN neuropathy, ECU tendinitis). Eligible designs included RCTs and comparative cohort/case-control studies with ≥ 6 months follow-up. Peer-reviewed English-language studies only. Studies were excluded for: no comparator; biomechanical/cadaveric/animal data; non-traumatic DRUJ pathology (e.g., rheumatoid); insufficient extractable data; isolated TFCC debridement without foveal repair; abstract-only/case reports/editorials/reviews/non-English publications. Search strategy A systematic electronic search of PubMed, Scopus, and Cochrane CENTRAL was performed initially in January 2026 and updated on 15 April 2026 to identify eligible studies published from January 2000 onward. Database-specific syntax combined controlled vocabulary and free-text terms related to distal radioulnar joint or triangular fibrocartilage complex (TFCC), instability or foveal injury, and arthroscopic, open, or transosseous repair techniques. In PubMed, the syntax used was: ((“distal radioulnar joint“[All Fields] OR “DRUJ“[All Fields] OR “triangular fibrocartilage complex“[All Fields] OR “TFCC“[All Fields]) AND (“instabilit*“[All Fields] OR “instable“[All Fields] OR “sublux*“[All Fields] OR “dislocat*“[All Fields] OR “foveal detachment“[All Fields] OR “foveal tear“[All Fields]) AND (“arthroscop*“[All Fields] OR “open repair“[All Fields] OR “foveal repair“[All Fields] OR “foveal reattachment“[All Fields] OR “transosse*“[All Fields])) AND (2000:2026[pdat]). In Scopus, the syntax was: TITLE-ABS-KEY((“distal radioulnar joint” OR DRUJ OR “triangular fibrocartilage complex” OR TFCC) AND (instabilit* OR instable OR sublux* OR dislocat* OR “foveal detachment” OR “foveal tear”) AND (arthroscop* OR “open repair” OR “foveal repair” OR “foveal reattachment” OR transosse*)) AND PUBYEAR > 1999 AND PUBYEAR < 2027. In Cochrane CENTRAL, the syntax used was: (“distal radioulnar joint” OR DRUJ OR “triangular fibrocartilage complex” OR TFCC) AND (instabilit* OR instable OR sublux* OR dislocat* OR “foveal detachment” OR “foveal tear”) AND (arthroscop* OR arthroscopy OR “open repair” OR “foveal reattachment” OR transosseous) in Title, Abstract, Keyword. Full database-specific search strategies are provided in Supplementary Appendix (S1–S9). Study selection and data extraction Four reviewers (G.H., M.R., Ab.E., A.H) independently screened titles/abstracts; another three (G.H., I.E., E.E.) performed full-text review. Data extraction (G.H., M.S., A.H.) captured study/design characteristics, demographics, injury chronicity, follow-up, techniques, and outcomes. Means/SDs were extracted directly; Medians/IQRs, as well as means (range), were converted to means/SDs using the Wan method [12]. No additional assumptions or imputations were applied for missing dispersion data because all included studies provided either standard deviations or convertible range-based summary measures. Quantitative synthesis was therefore based exclusively on reported or mathematically derived summary statistics. Injury chronicity subgroup. Studies were classified by mean symptom duration (chronic > 6 months). Mixed presentations without subgroup outcomes were excluded from stratified pooling. Disagreements were resolved by consensus. Extracted datasets; Supplementary Appendix (S10–S12). Data synthesis Meta-analyses used R meta package v4.4.1. Final postoperative values were pooled as SMDs (95% CI, inverse-variance) for continuous outcomes due to unavailable change-score variances/correlations across studies—confirmed robust via sensitivity analysis of baseline comparability. Binary outcomes used ORs (Mantel-Haenszel). Random-effects REML models with Hartung-Knapp adjustment assessed heterogeneity (I²: 0–25% low, 26–50% moderate, 51–75% substantial, > 75% considerable). Post-hoc sensitivity analyses included exclusion of outlier studies when I² exceeded 75% and exclusion of studies with baseline imbalance. Detailed meta-analysis; Supplementary Appendix (S12–S14). Risk of bias and certainty assessment ROBINS-I risk of bias assessments were performed independently by three reviewers (G.H., A.Elg., M.S.). Certainty of evidence was evaluated independently using the GRADE approach by the same three reviewers, with disagreements resolved by consensus. Publication bias was not formally assessed due to fewer than 10 studies per outcome. Detailed risk of bias and GRADE assessments; Supplementary Appendix (S15, S16). Results Study selection, characteristics, and risk of bias The systematic search identified 730 records. After duplicate removal (n = 270), 460 records underwent title/abstract screening (329 excluded). Excluded reports comprised TFCC- and DRUJ-related studies that did not meet the predefined comparative surgical eligibility criteria. Among non-English reports (n = 43) from Germany, France, and Spain, Japan, China, Poland, and Turkey, nine studies addressed acute injuries through early arthroscopic repair or fracture-associated treatment of fresh TFCC or DRUJ lesions. Twenty-two studies evaluated chronic instability or delayed reconstruction, including tendon grafting and ligament reconstruction. Twelve reports contained technical descriptions, diagnostic studies, or general DRUJ instability discussions. Biomechanical studies and cadaveric reports (n = 20) evaluated cadaveric fixation constructs, ligament mechanics, and stabilisation strategies without clinical outcomes. Case reports (n = 19) included eleven acute traumatic presentations involving Galeazzi fracture–associated DRUJ instability, acute flap tears of the Triangular fibrocartilage complex, irreducible distal radioulnar joint dislocation, and complex wrist fracture-dislocations. Two reports described chronic sequelae after Essex-Lopresti injury, and six described iatrogenic complications or isolated technical observations. Technical reports (n = 26) described fracture-related arthroscopic techniques, ligament reconstruction, portal modifications, cadaveric procedures, and transosseous fixation methods. Review articles (n = 106) addressed acute TFCC injury, chronic instability, degenerative pathology, wrist arthroscopy, diagnostics, and DRUJ disorders without primary comparative outcome data. Studies without primary DRUJ focus (n = 35) evaluated fracture-associated soft-tissue injury, non-DRUJ ligament instability, capsulodesis, grading systems, and non-DRUJ carpal disorders. Non-traumatic studies (n = 7) included six chronic degenerative or inflammatory conditions such as arthrosis, rheumatoid wrist involvement, ulnar impaction syndrome, and degenerative DRUJ instability, together with one salvage study for advanced DRUJ arthritis. Pediatric studies (n = 6) included one acute repair series, two chronic or delayed presentations, and three cohorts evaluated physeal-sparing reconstruction or imaging-based instability assessment. Non-surgical studies (n = 41) included three acute conservative fracture-associated reports, ten chronic imaging or rehabilitation studies, and twenty-eight diagnostic or non-operative investigations using MRI, CT arthrography, brace treatment, or consensus-based management. Editorial material (n = 8) consisted of commentaries, expert opinion papers, and errata discussing TFCC foveal repair, DRUJ instability after distal radius fracture, outcome interpretation after arthroscopic repair, and diagnostic manoeuvres such as hook or trampoline tests. Miscellaneous excluded studies (n = 18) included four acute post-fracture TFCC or DRUJ reports, five chronic salvage procedures such as distal ulnar resection or ulnar shortening osteotomy, and nine diagnostic or classification studies. Full-text exclusions (n = 126) consisted of non-comparative single-arm surgical series: arthroscopic repair reports (n = 83) included five acute studies, twenty-five chronic or irreparable cases, and fifty-three traumatic cohorts without temporal stratification, whereas open repair reports (n = 28) included three acute studies, nineteen chronic reconstructive series, and six cohorts without defined chronicity, including tendon grafting, transosseous repair, capsular augmentation, and delayed foveal reattachment. Fourteen comparative studies were excluded because they compared repair with conservative treatment, debridement, ulnar shortening, immobilisation protocols, diagnostic testing, or salvage procedures rather than direct open versus arthroscopic foveal TFCC repair. One additional report lacked separately extractable treatment-group outcomes for quantitative synthesis. Five comparative studies were eligible for quantitative synthesis (Anderson et al. [13], Nakamura et al. [14], Luchetti et al. [15], Abe et al. [16], Lee et al. [17]; 322 wrists: 181 open, 141 arthroscopic) (Fig. 1). Study characteristics are summarized in Table 1. ROBINS-I rated two studies at moderate risk of bias (Luchetti et al. [15], Abe et al. [16]) and three at serious risk (Anderson et al. [13], Nakamura et al. [14], Lee et al. [17]), primarily due to confounding by non-random treatment allocation (Table 2). Baseline characteristics were generally comparable across studies, except Luchetti et al. [15] (DASH 58 ± 23 vs. 39 ± 21; PRWE 69 ± 29 vs. 54 ± 20 favoring arthroscopy). Baseline differences of 19 points for DASH and 15 points for PRWE exceeded the minimal clinically important differences (MCID) thresholds (DASH 10–15 points; PRWE 14 points), indicating clinically relevant preoperative non-equivalence. Table 1Characteristics of studies included in the systematic reviewStudyJournalCountryStudy DesignTotal Wrists (Open / Arth)Mean Age, years (Open / Arth)Male: Female (Open / Arth)Follow-up, months (Open / Arth)Injury chronicityAnderson et al. (2008) [13] J Hand Surg Am USARetrospective cohort (III)75 (39 / 36)35 ± 15 /32 ± 1217:22/18:1953 ± 20 / 32 ± 6MixedNakamura et al. (2011) [14] Hand Clinics JapanRetrospective cohort (III)90 (66 / 24)31 ± 17 / 27 ± 1538:28/13:1136 ± 11 / 36 ± 8MixedLuchetti et al. (2014) [15] J Hand Surg Eur ItalyProspective comparative (II)49 (24 / 25)32 ± 6 / 33 ± 915:9/12:1331 ± 10/ 31 ± 10ChronicAbe et al. (2018) [16] J Wrist Surg JapanRetrospective cohort (III)29 (8 / 21)22 ± 4 / 34 ± 10Total 13:16Total 34.4 ± 8ChronicLee et al. (2024) [17] J Clin Med South KoreaRetrospective cohort (III)79 (44 / 35)39 ± 14 /40 ± 1219:25/15:2024 / 24ChronicWeighted Pooled values——181 / 14133.9 / 33.789:92 / 58:8336.2 / 31.6Mixed overallPooled demographic values were calculated as sample-size-weighted descriptive summaries using the reported open and arthroscopic subgroup sample sizes from each included study. For variables incompletely reported by individual studies, pooled estimates were derived only from studies providing the relevant data; therefore, denominators varied according to the characteristic analysed. These pooled values were presented for descriptive comparison only and were not used in the meta-analytic pooling of outcome measures Table 1Characteristics of studies included in the systematic reviewStudyJournalCountryStudy DesignTotal Wrists (Open / Arth)Mean Age, years (Open / Arth)Male: Female (Open / Arth)Follow-up, months (Open / Arth)Injury chronicityAnderson et al. (2008) [13] J Hand Surg Am USARetrospective cohort (III)75 (39 / 36)35 ± 15 /32 ± 1217:22/18:1953 ± 20 / 32 ± 6MixedNakamura et al. (2011) [14] Hand Clinics JapanRetrospective cohort (III)90 (66 / 24)31 ± 17 / 27 ± 1538:28/13:1136 ± 11 / 36 ± 8MixedLuchetti et al. (2014) [15] J Hand Surg Eur ItalyProspective comparative (II)49 (24 / 25)32 ± 6 / 33 ± 915:9/12:1331 ± 10/ 31 ± 10ChronicAbe et al. (2018) [16] J Wrist Surg JapanRetrospective cohort (III)29 (8 / 21)22 ± 4 / 34 ± 10Total 13:16Total 34.4 ± 8ChronicLee et al. (2024) [17] J Clin Med South KoreaRetrospective cohort (III)79 (44 / 35)39 ± 14 /40 ± 1219:25/15:2024 / 24ChronicWeighted Pooled values——181 / 14133.9 / 33.789:92 / 58:8336.2 / 31.6Mixed overallPooled demographic values were calculated as sample-size-weighted descriptive summaries using the reported open and arthroscopic subgroup sample sizes from each included study. For variables incompletely reported by individual studies, pooled estimates were derived only from studies providing the relevant data; therefore, denominators varied according to the characteristic analysed. These pooled values were presented for descriptive comparison only and were not used in the meta-analytic pooling of outcome measures Characteristics of studies included in the systematic review Pooled demographic values were calculated as sample-size-weighted descriptive summaries using the reported open and arthroscopic subgroup sample sizes from each included study. For variables incompletely reported by individual studies, pooled estimates were derived only from studies providing the relevant data; therefore, denominators varied according to the characteristic analysed. These pooled values were presented for descriptive comparison only and were not used in the meta-analytic pooling of outcome measures Table 2ROBINS-I risk of bias summaryStudyConfoundingParticipant SelectionIntervention ClassificationDeviationsMissing DataOutcome MeasurementSelective ReportingOverallAnderson et al. [13]SeriousModerateLowLowLowModerateLowSeriousNakamura et al. [14]SeriousSeriousLowModerateLowModerateLowSeriousLuchetti et al. [15]ModerateLowLowLowLowModerateLowModerateAbe et al. [16]ModerateModerateLowLowLowModerateLowModerateLee et al. [17]SeriousModerateLowLowLowModerateLowSeriousDetailed ROBINS-I domain assessments; Supplementary Appendix (S15) Table 2ROBINS-I risk of bias summaryStudyConfoundingParticipant SelectionIntervention ClassificationDeviationsMissing DataOutcome MeasurementSelective ReportingOverallAnderson et al. [13]SeriousModerateLowLowLowModerateLowSeriousNakamura et al. [14]SeriousSeriousLowModerateLowModerateLowSeriousLuchetti et al. [15]ModerateLowLowLowLowModerateLowModerateAbe et al. [16]ModerateModerateLowLowLowModerateLowModerateLee et al. [17]SeriousModerateLowLowLowModerateLowSeriousDetailed ROBINS-I domain assessments; Supplementary Appendix (S15) ROBINS-I risk of bias summary Detailed ROBINS-I domain assessments; Supplementary Appendix (S15) Fig. 1PRISMA flow diagram of study selection PRISMA flow diagram of study selection Meta-analysis of functional outcomes Three studies (203 wrists) reported MMWS. Random-effects meta-analysis showed SMD − 0.12 (95% CI − 0.47 to 0.23; I² = 0%, τ² = 0; p = 0.719). (Fig. 2). Study-level means, standard deviations, and pooled analyses for continuous outcomes are presented in Supplementary Appendix (S14). Fig. 2Forest plot of open vs. arthroscopic TFCC repair for MMWS at final follow-up. CI = confidence interval, N(OP/AR) = number of wrists (open repair/arthroscopic repair), SMD = standardized mean difference, I² = I-squared heterogeneity statistic, τ² = tau-squared between-study variance, p = p-value for heterogeneity test Forest plot of open vs. arthroscopic TFCC repair for MMWS at final follow-up. CI = confidence interval, N(OP/AR) = number of wrists (open repair/arthroscopic repair), SMD = standardized mean difference, I² = I-squared heterogeneity statistic, τ² = tau-squared between-study variance, p = p-value for heterogeneity test Four studies (232 wrists) reported DASH. Random-effects meta-analysis showed SMD 0.22 (95% CI − 0.96 to 1.40; I² = 86.7%, τ² = 0.48; p < 0.0001). Sensitivity analysis excluding Andersson et al. [13] (outlier) and Luchetti et al. [15] (baseline imbalance) showed SMD 0.31 (95% CI − 0.88 to 1.51; I² = 0%, τ² = 0; p = 0.639) (Fig. 3). Fig. 3Forest plot of open vs. arthroscopic TFCC repair for DASH score at final follow-up; (a) primary meta-analysis, (b) sensitivity analysis excluding Andersson et al. [13] (outlier) and Luchetti et al. [15] (baseline imbalance) Forest plot of open vs. arthroscopic TFCC repair for DASH score at final follow-up; (a) primary meta-analysis, (b) sensitivity analysis excluding Andersson et al. [13] (outlier) and Luchetti et al. [15] (baseline imbalance) Three studies (203 wrists) reported PRWE scores. Random-effects meta-analysis showed SMD − 0.24 (95% CI − 2.85 to 2.38; I² = 93.3%, τ² = 1.03; p < 0.0001). Sensitivity analysis excluding Luchetti et al. [15] (baseline imbalance) (SMD − 0.73 (95% CI − 8.29 to 6.84; I² = 91.7%, τ² = 0.65; p = 0.0005) (Fig. 4). Fig. 4Forest plot of open vs. arthroscopic TFCC repair for PRWE at final follow-up; (a) primary meta-analysis, (b) sensitivity analysis excluding Luchetti et al. [15] (baseline imbalance) Forest plot of open vs. arthroscopic TFCC repair for PRWE at final follow-up; (a) primary meta-analysis, (b) sensitivity analysis excluding Luchetti et al. [15] (baseline imbalance) Four studies (232 wrists) reported VAS scores. Random-effects meta-analysis showed SMD − 0.35 (95% CI − 1.69 to 0.99; I² = 89.9%, τ² = 0.63; p < 0.0001). Sensitivity analysis excluding Andersson et al. [13] (outlier) showing (SMD − 0.10 (95% CI − 0.60 to 0.81; I² = 0%, τ² = 0; p = 0.827) (Fig. 5). Fig. 5Forest plot of open vs. arthroscopic TFCC repair for VAS score at final follow-up; (a) primary meta-analysis, (b) sensitivity analysis excluding Andersson et al. [13] (outlier) Forest plot of open vs. arthroscopic TFCC repair for VAS score at final follow-up; (a) primary meta-analysis, (b) sensitivity analysis excluding Andersson et al. [13] (outlier) Four studies (232 wrists) reported grip strength. Random-effects meta-analysis showed SMD 0.01 (95% CI − 0.41 to 0.44; I² = 0%, τ² = 0; p = 0.844) (Fig. 6). Fig. 6Forest plot of open vs. arthroscopic TFCC repair for grip strength at final follow-up Forest plot of open vs. arthroscopic TFCC repair for grip strength at final follow-up Four studies (232 wrists) reported flexion–extension arc. Random-effects meta-analysis showed (SMD − 0.32 (95% CI − 0.94 to 0.30; I² = 42.9%, τ² = 0.04; p = 0.154). Four studies (232 wrists) reported pronation–supination arc. Random-effects meta-analysis showed SMD − 0.13 (95% CI − 0.75 to 0.48; I² = 46.7%, τ² = 0.06; p = 0.130) (Fig. 7). Fig. 7Forest plot of open vs. arthroscopic TFCC repair for ROM at final follow-up; (a) flexion-extension arc, (b) pronation-supination arc Forest plot of open vs. arthroscopic TFCC repair for ROM at final follow-up; (a) flexion-extension arc, (b) pronation-supination arc Complications Five studies reported recurrent instability. Random-effects meta-analysis showed OR 1.00 (95% CI 0.19 to 5.37; I² = 63.2%, τ² = 1.18; p = 0.028), Four studies reported DSBUN neuropathy with OR 2.59 (95% CI 0.37 to 18.19; I² = 0%, τ² = 0.03; p = 0.428), and three studies reported ECU tendinitis with OR 2.47 (95% CI 0.16 to 39.08; I² = 0%, τ² = 0; p = 0.704) (Fig. 8). Fig. 8Forest plots of open vs. arthroscopic TFCC repair for complications at final follow-up; (a) recurrent instability, (b) DSBUN neuropathy, (c) ECU Tendinitis Forest plots of open vs. arthroscopic TFCC repair for complications at final follow-up; (a) recurrent instability, (b) DSBUN neuropathy, (c) ECU Tendinitis Clinical heterogeneity / injury chronicity and timing of surgery Injury chronicity varied across the included studies and contributed to clinical heterogeneity. Andersson et al. [13] and Nakamura et al. [14] included mixed acute and chronic presentations, whereas Luchetti et al. [15], Abe et al. [16], and Lee et al. [17] predominantly evaluated chronic lesions. Because postoperative outcomes were not reported separately for acute and chronic cases, formal patient-level subgroup analysis was not feasible. Exploratory synthesis restricted to predominantly chronic cohorts showed that; two studies (128 wrists) reported MMWS [SMD −0.20 (95% CI −0.72 to 0.32; I² = 0%, τ² = 0; p = 0.818)], three studies (157 wrists) reported DASH [SMD 0.55 (95% CI −0.41 to 1.51; I² = 44.3%, τ² = 0.07; p = 0.165), two studies (128 wrists) reported PRWE [SMD 0.30 (95% CI −5.51 to 6.11; I² = 83.3%, τ² = 0.34; p = 0.015), three studies (157 wrists) reported VAS scores [SMD 0.10 (95% CI −0.60 to 0.81; I² = 0.0%, τ² = 0; p = 0.827) and three studies (157 wrists) reported recurrent instability (OR 0.48 (95% CI 0.06 to 3.90; I² = 0%, τ² = 0; p = 0.493) (Fig. 9). Detailed chronic subgroup estimates are presented in Supplementary Appendix (S17, S18). Fig. 9Forest plots of open vs. arthroscopic TFCC repair for chronic subgroup estimates at final follow-up; (a) MMWS, (b) DASH, (c) PRWE, (d) VAS, (e) recurrent instability Forest plots of open vs. arthroscopic TFCC repair for chronic subgroup estimates at final follow-up; (a) MMWS, (b) DASH, (c) PRWE, (d) VAS, (e) recurrent instability GRADE certainty was low (MMWS, grip) to very low (others) due to study limitations, inconsistency, and imprecision (Table 3). Table 3Certainty of evidence for pooled outcomesOutcomeFinal certaintyExplanationMMWSLow ⊕⊕◯◯I²=0%; similar effect estimates under fixed- and random-effects modelsDASHVery low ⊕◯◯◯I²=86.7%; wide CI crossing benefit and harmPRWEVery low ⊕◯◯◯I²=93.3%; very wide CIVASVery low ⊕◯◯◯I²=89.9% in primary analysis, resolving to 0% in sensitivity analysis; confidence interval remained wideGrip strengthLow ⊕⊕◯◯I²=0%; similar effect estimates under fixed- and random-effects modelsFlexion–extension arcLow ⊕⊕◯◯I²=42.9%; confidence interval crossed no effectPronation–supination arcLow ⊕⊕◯◯I²=46.7%; confidence interval crossed no effectRecurrent instabilityVery low ⊕◯◯◯I²=63.2%; wide OR CI crossing benefit and harmDSBUN neuropathyVery low ⊕◯◯◯I²=0%; wide OR CI with sparse eventsECU tendinitisVery low ⊕◯◯◯I²=0%; very wide OR CI with very sparse eventsDetailed certainty judgments and supporting explanations are presented in Supplementary Appendix (S16) Table 3Certainty of evidence for pooled outcomesOutcomeFinal certaintyExplanationMMWSLow ⊕⊕◯◯I²=0%; similar effect estimates under fixed- and random-effects modelsDASHVery low ⊕◯◯◯I²=86.7%; wide CI crossing benefit and harmPRWEVery low ⊕◯◯◯I²=93.3%; very wide CIVASVery low ⊕◯◯◯I²=89.9% in primary analysis, resolving to 0% in sensitivity analysis; confidence interval remained wideGrip strengthLow ⊕⊕◯◯I²=0%; similar effect estimates under fixed- and random-effects modelsFlexion–extension arcLow ⊕⊕◯◯I²=42.9%; confidence interval crossed no effectPronation–supination arcLow ⊕⊕◯◯I²=46.7%; confidence interval crossed no effectRecurrent instabilityVery low ⊕◯◯◯I²=63.2%; wide OR CI crossing benefit and harmDSBUN neuropathyVery low ⊕◯◯◯I²=0%; wide OR CI with sparse eventsECU tendinitisVery low ⊕◯◯◯I²=0%; very wide OR CI with very sparse eventsDetailed certainty judgments and supporting explanations are presented in Supplementary Appendix (S16) Certainty of evidence for pooled outcomes Detailed certainty judgments and supporting explanations are presented in Supplementary Appendix (S16) Discussion This systematic review and meta-analysis synthesized the currently available comparative data on open versus arthroscopic foveal TFCC repair for acute and chronic DRUJ instability. Across functional scores, pain, grip strength, range of motion, recurrent instability, and complication rates, the pooled estimates did not demonstrate clear or consistent superiority of either technique. However, these findings should not be interpreted as evidence that the two approaches are equivalent. Most analyses were based on a small number of observational cohorts, with important clinical and methodological heterogeneity and low to very low certainty of evidence, so the absence of statistically significant differences is more consistent with limited and imprecise data than with true similarity between procedures. Several outcomes, including DASH, PRWE, VAS, and chronicity-restricted subgroup analyses, were informed by only two or three studies and, in some cases, by fewer than 200 wrists overall. Although random-effects models with REML estimation and Hartung–Knapp adjustment were used because between-study heterogeneity and small study numbers were anticipated, the resulting confidence intervals remained wide and frequently included effects compatible with clinically meaningful benefit in either direction as well as no difference. These pooled estimates should therefore be regarded as exploratory and hypothesis-generating rather than definitive comparative evidence. They reduce uncertainty only partially and do not support firm conclusions regarding non-inferiority, equivalence, or interchangeability of open and arthroscopic repair. Prior systematic reviews identified only limited comparative evidence. Andersson et al. [8] reviewed two comparative studies and found no clear differences in DRUJ re-instability, range of motion, grip strength, or functional scores between approaches, with only a small DASH advantage for arthroscopy in one cohort. Robba et al. [10] similarly reported comparable postoperative DRUJ stability after open (84%) and arthroscopic (86%) repair for Palmer 1B tears. Our meta-analysis extends these findings by quantitatively pooling five comparative cohorts and incorporating structured ROBINS-I and GRADE assessments, but the overall message remains one of uncertainty: current evidence does not reliably establish that one technique yields better long-term outcomes than the other. Statistical heterogeneity varied across outcomes. Heterogeneity was low for MMWS, grip strength, and several motion measures, suggesting broadly similar study-level estimates; however, the small number of studies per analysis limits the interpretability of I² in this context. By contrast, DASH, PRWE, VAS, and recurrent instability showed substantial or considerable inconsistency. This likely reflects differences in injury chronicity, timing of surgery, surgical details, rehabilitation protocols, and study design, as well as baseline non-equivalence between treatment groups. In particular, Andersson et al. [13] contributed markedly divergent pain and PRWE estimates, whereas Luchetti et al. [15] introduced clinically relevant baseline imbalance for DASH and PRWE. Sensitivity analyses excluding these influential cohorts reduced heterogeneity for several outcomes, but confidence intervals frequently remained broad and crossed thresholds compatible with clinically important differences, indicating that residual uncertainty persists despite statistical adjustment. In the present review, most SMD estimates were small in magnitude; however, their confidence intervals were wide and crossed both positive and negative values, making it difficult to exclude clinically important differences in either direction. In addition, SMDs may be influenced by between-study variation in measurement methods, dispersion, and population heterogeneity, so numerically small effect sizes should not automatically be interpreted as negligible clinical impact. Consequently, these findings are better understood as showing no clear evidence of large or consistent differences between techniques rather than demonstrating clinical equivalence. The methodological quality of the included evidence also tempers the conclusions. All five comparative cohorts were non-randomized retrospective studies, with three judged at serious risk of bias and two at moderate risk according to ROBINS-I. Major concerns included non-randomized treatment allocation, time-sequential group assignment, and potential confounding related to surgeon preference, injury chronicity, and baseline imbalance. GRADE therefore rated certainty as low for relatively stable outcomes such as MMWS and grip strength, and very low for several patient-reported and complication outcomes because of heterogeneity, sparse events, and wide confidence intervals. Under GRADE, such ratings imply that the true effect may be substantially different from the pooled estimate, reinforcing that our results should be viewed as uncertain and non-definitive rather than as proof of comparable performance of the two techniques. Because several studies reported only final postoperative means and standard deviations, pooled continuous analyses were based on final rather than change scores. Although baseline imbalance was evident in Luchetti et al. [15] for DASH and PRWE, sensitivity analyses excluding this study did not materially alter the overall comparative estimates, but they also did not eliminate imprecision. In addition, injury chronicity was inconsistently reported, preventing robust patient-level subgroup analysis of acute versus chronic lesions. Exploratory synthesis restricted to predominantly chronic cohorts did not materially change pooled estimates but should be regarded as hypothesis-generating only. From a clinical perspective, the available evidence suggests that both open and arthroscopic foveal TFCC repair can improve function and DRUJ stability in selected patients but does not establish that they are interchangeable or that one is categorically superior. Open foveal repair offers direct visualization and broad access to the foveal insertion, whereas arthroscopic repair provides a less invasive approach and may facilitate management of associated intra-articular pathology [18–28]. Given the limitations of the existing data, surgical decision-making should remain individualized, considering lesion characteristics, chronicity, associated pathology, and surgeon expertise rather than an expectation of equivalence based on the present meta-analytic findings. Several limitations should be acknowledged. Sample sizes were modest, studies were predominantly single-center, and surgical techniques and rehabilitation protocols varied across cohorts. Formal assessment of publication bias was not performed because each meta-analysis included fewer than ten studies, making funnel-plot-based methods and statistical tests unreliable in this setting. Nevertheless, the possibility of small-study effects, selective outcome reporting, and unpublished negative studies cannot be excluded, and these sources of bias may have influenced the pooled estimates in either direction. Conclusion Based on the currently available comparative evidence, open and arthroscopic foveal TFCC repair both appear capable of improving functional outcomes, pain, forearm motion, grip strength, and DRUJ stability, but the data do not demonstrate clear superiority of either technique. At the same time, the small number of non-randomized studies, substantial heterogeneity for several outcomes, and low to very low certainty of evidence for most endpoints mean that these findings should be interpreted with caution and should not be taken as evidence of clinical equivalence. Surgical choice should therefore remain individualized according to lesion characteristics, surgeon expertise, and associated pathology until higher-quality prospective comparative studies become available. Future research should incorporate prospectively collected timing data, use randomized or rigorously matched comparative designs, and report standardized outcome measures with established MCIDs, ideally stratified by injury chronicity, to clarify whether either technique offers advantages in specific clinical subgroups. Supplementary Information Supplementary Material 1. Supplementary Material 1.