Aepala MR, Kiani S, Yeramosu T
… +1 more, Sabharwal S
Clin Orthop Relat Res
· 2026 May · PMID 42190173
·
Publisher ↗
BACKGROUND: Patient-reported outcome measures (PROMs) and clinician-reported outcome measures (CROMs) are designed to capture different dimensions of treatment response: the patient's experience and the clinician's asses...BACKGROUND: Patient-reported outcome measures (PROMs) and clinician-reported outcome measures (CROMs) are designed to capture different dimensions of treatment response: the patient's experience and the clinician's assessment, respectively. Because of that, some degree of discordance between them is expected. However, the nature and magnitude of PROM-CROM correlations across orthopaedic subspecialties have not been systematically characterized. Understanding where they converge and where they diverge may help clinicians and researchers select instrument pairings that capture complementary dimensions of treatment response for different musculoskeletal conditions. QUESTIONS/PURPOSES: (1) Do CROMs correlate differently with condition-specific PROMs compared with general health PROMs? (2) Do PROMs show distinct correlations with hard CROMs (objective, performance-based measures without patient-reported outcomes) versus soft CROMs (objective, performance-based measures including patient-reported domains)? (3) How do PROMs and CROMs compare in their ability to detect change over time as measured by responsiveness indices such as effect size, standardized response mean, and other related metrics? METHODS: We systematically searched Medline via PubMed, Embase, and Web of Science Core Collection in July 2025. Two reviewers independently performed title, abstract, and full-text screening. Studies were included when they directly compared PROMs and CROMs for orthopaedic-related topics. Extracted variables included study design, sample size, Spearman and Pearson correlation coefficients, and responsiveness. CROMs were classified as "hard" or "soft" based on the extent to which the instrument relied on patient input. Hard CROMs represent objective, performance-based or clinician-measured outcomes that do not require patient-reported information, whereas soft CROMs include clinician-administered instruments that incorporate patient-reported symptoms or functional assessments. PROMs were categorized as either generic or condition specific (joint-, region-, disease-specific) according to established definitions. All reported correlation coefficients were extracted and included regardless of statistical significance to comprehensively characterize the magnitude and variability of associations between PROMs and CROMs. Responsiveness was defined as the comparative sensitivity of PROMs and CROMs to clinical change, assessed using effect sizes, standardized response means, and related metrics. Thirty studies met inclusion criteria; 27 reported PROM-CROM correlations and nine evaluated responsiveness. The included studies encompassed a variety of musculoskeletal conditions, including musculoskeletal tumors (n = 4), distal radius fractures (n = 3), total joint arthroplasty and osteoarthritis (n = 11), pediatric limb deformities (n = 3), foot and ankle disorders (n = 7), spine (n = 1), and trauma (n = 1), with sample sizes ranging from 20 to 717 (mean = 143). Assessment of study quality, using an adapted Consensus-based Standards for the Selection of Health Measurement Instruments risk of bias framework, revealed that most studies were of "low risk" or "some concerns" because of incomplete paired PROM-CROM outcome data rather than flaws in design or analysis. RESULTS: In evaluating relationships between PROMs and CROMs across orthopaedic subspecialties, condition-specific PROMs generally showed stronger correlations with corresponding CROMs, including the Toronto Extremity Salvage Score (r = 0.75 to 0.81) and the patient-reported American Orthopaedic Foot & Ankle Society (AOFAS) score (r = 0.70). In contrast, general health measures such as the EQ-5D time trade-off (r = -0.29 to 0.13) and certain region-specific instruments, including the Oswestry Disability Index (r = 0.27) and several upper extremity PROMs (r = -0.046 to 0.41), often demonstrated weaker correlations with CROMs. When evaluating differences in correlation by CROM subtype, PROMs tended to correlate more strongly with soft CROMs, such as the AOFAS score, Musculoskeletal Tumor Society score, and Harris hip score, than with hard CROMs. In comparing responsiveness between PROMs and CROMs, condition-specific PROMs, including the Manchester-Oxford Foot Questionnaire, the WOMAC, and Knee Society Score, were more often sensitive to clinical change than CROMs, except in rehabilitation-focused settings, where hard CROM responsiveness was often greater than for PROMs and soft CROMs. CONCLUSION: Based on currently available studies related to musculoskeletal conditions, PROMs and CROMs show variable associations with one another. The stronger correlations between soft CROMs and condition-specific PROMs suggest that when the goal is to establish benchmarks, a condition-specific PROM should be paired with a soft CROM. In contrast, hard CROMs may be particularly useful in short-term rehabilitation settings, where objective performance changes are a primary focus of evaluation. We recommend that orthopaedic studies and registries include, at a minimum, one condition-specific PROM (or soft CROM) and one hard CROM to ensure a multidimensional assessment of patient outcomes and improve interpretability across studies. LEVEL OF EVIDENCE: Level III, therapeutic study.
de Groot TM, Gonzalez MR, van der Linden LR
… +6 more, Skornja I, Groot OQ, Schwab JH, Doornberg JN, Jutte PC, Lozano-Calderón SA
Clin Orthop Relat Res
· 2026 May · PMID 42171494
·
Publisher ↗
BACKGROUND: Prognostic support tools are increasingly used to guide treatment decisions in patients with metastatic long-bone disease. PathFX is a widely distributed survival prediction model that has been validated worl...BACKGROUND: Prognostic support tools are increasingly used to guide treatment decisions in patients with metastatic long-bone disease. PathFX is a widely distributed survival prediction model that has been validated worldwide in various settings. Despite this, to our knowledge, there has been no systematic evaluation of PathFX's algorithmic fairness across clinically relevant subgroups within external evaluation studies. QUESTIONS/PURPOSES: (1) How accurately does PathFX predict survival at 1, 3, 6, 12, 18, and 24 months in an external cohort of patients undergoing surgery for long-bone metastases? (2) Is the performance and error distribution of PathFX fair across key sociodemographic, clinical, and temporal subpopulations within an external cohort of patients undergoing surgery for long-bone metastases? METHODS: All patients 18 years or older from a tertiary orthopaedic oncology service who underwent surgery from January 2010 to December 2022 for impending or completed metastatic long-bone fracture were retrospectively studied. Of the 1018 patients, 45% (460 of 1018) were male. Race and ethnicity were self-identified through a standardized institution-wide demographic survey and recorded in the electronic health record. Among patients with available data (n = 991), 88% (874 of 991) identified as White, 5% (51 of 991) as Black, 3% (28 of 991) as Asian, and 4% (38 of 991) as Other. Race and ethnicity data were missing or not reported for 4% (36 of 1018) of patients. The primary outcome was overall survival at prespecified time points (1, 3, 6, 12, 18, and 24 months). Data on the nine predictors required by PathFX (age, sex, primary tumor group, Eastern Cooperative Oncology Group performance status, pathologic fracture status at the index site, presence of multiple skeletal metastases, presence of organ metastases, hemoglobin level, and absolute lymphocyte count) were collected for each patient. We assessed discrimination (time-specific area under the curve [AUC]/C-index with 95% confidence intervals [CIs]), calibration (slope and intercept with CIs and graphical calibration), overall accuracy (Brier score), and decision curve analysis. Discrimination (time-specific AUC/C-index) reflects how well the model distinguishes between patients who experience the event and those who do not; it ranges from 0.5 (no better than chance) to 1.0 (perfect discrimination), with values around 0.7 generally considered acceptable and ≥ 0.8 strong. Calibration assesses whether predicted probabilities agree with observed outcomes: the calibration intercept indicates systematic overestimation or underestimation (ideal = 0), while the calibration slope reflects whether risk predictions are too extreme or too moderate (ideal = 1). Overall accuracy was quantified using the Brier score, which measures the average squared difference between predicted probabilities and actual outcomes; lower values indicate better accuracy, with 0 representing perfect prediction. Finally, decision curve analysis evaluates clinical usefulness by estimating the net benefit of using the model across a range of decision thresholds compared with default strategies (treat all or treat none). We evaluated model performance and error distribution within prespecified sociodemographic, clinical, and temporal subgroups and compared subgroup estimates using Δmetrics with 95% CIs. RESULTS: In general, the accuracy and other performance parameters we observed for PathFX were inadequate for clinical use. Overall, the best-performing model was the 18-month survival model: AUC 0.63 (95% CI 0.60 to 0.67), Brier 0.22 (95% CI 0.21 to 0.23), calibration slope 0.58 (95% CI 0.33 to 0.83), and intercept 0.21 (95% CI 0.10 to 0.32). The AUC for the other models did not exceed 0.68, with worse calibration metrics. Intercepts were positive for all time points, which means that the model systematically underestimated survival in this patient population. Calibration slopes were < 1 throughout, indicating overconfident (too extreme) probabilities. Brier scores ranged from 0.07 to 0.24, which is consistent with moderate probabilistic accuracy. Because the Brier score is dependent on the baseline event incidence, variation across prediction time points partly reflects changes in outcome frequency rather than pure differences in discriminative or calibration performance. The subgroup analyses suggested heterogeneity; that is, the model exhibited a better discrimination in females and poorer performance in patients who were not White with flatter calibration slopes. There were no clear differences in subgroups based on treatment period. CONCLUSION: Based on the findings of this study, PathFX in its current form is insufficient for clinical use in patients with long-bone metastases undergoing surgery, as it consistently underestimates survival. Recalibration of the model through development of an updated cohort with stepwise model updating and subgroup stability checks is warranted; however, even after recalibration, complete model redevelopment may ultimately be required before PathFX can be reliably used to guide surgical decision-making. LEVEL OF EVIDENCE: Level III, prognostic study.
Yang E, Lin HC, Shimomura S
… +6 more, Lee J, Kim HS, Yen HK, Iwata S, Lin WH, Han I
Clin Orthop Relat Res
· 2026 May · PMID 42171127
·
Publisher ↗
BACKGROUND: Patients undergoing surgery for bone metastases typically have advanced disease, and postoperative survival varies substantially. Accurate survival estimation is important for surgical decision-making and pat...BACKGROUND: Patients undergoing surgery for bone metastases typically have advanced disease, and postoperative survival varies substantially. Accurate survival estimation is important for surgical decision-making and patient counseling. Several prognostic models have been externally validated in East Asian populations, but these tools were originally developed in Western cohorts and do not incorporate region-specific epidemiology or treatment patterns. QUESTIONS/PURPOSES: (1) To develop, internally evaluate, and select a machine learning-based survival prediction model for patients undergoing surgery for nonspinal bone metastases using a multinational East Asian cohort. (2) To compare the performance of the selected model with that of an established Western prognostic tool developed by the Skeletal Oncology Research Group (SORG). (3) To identify which clinical features carried the greatest importance in the new model that we developed. METHODS: All patients who underwent surgery for nonspinal bone metastases at three tertiary referral centers in the Republic of Korea, Taiwan, and Japan between January 2009 and December 2022 were included. In total, 1045 patients met the inclusion criteria. The median (range) age at surgery was 64 years (19 to 96), 46% (478 of 1045) of patients were female, and the femur was the most common metastatic site (66% [690]). Data for 3-month, 6-month, 1-year, 3-year, and 5-year overall survival were available for 82% (854), 68% (709), 51% (529), 23% (243), and 15% (160) of patients, respectively. The corresponding survival proportions were 84%, 71%, 56%, 36%, and 31%. Data on routinely available clinical, functional, and laboratory variables were collected, and candidate predictors were predefined based on clinical relevance and data availability across institutions. Missing data were < 4% for all variables in each institution and were handled by multivariate imputation by chained equations. We trained four models using different machine-learning algorithms, and the performance of each model was evaluated using leave-one-site-out validation, in which models were trained on data from two institutions and tested on the remaining institution to ensure separation between training and testing data sets. Model performance was assessed using the Concordance Index (C-index; the ability of the model to correctly rank patients according to their expected survival), Brier score (overall prediction error), time-dependent area under the curve (tdAUC; how well the model distinguishes patients with different survival outcomes at specific time points), calibration slope and intercept (agreement between predicted and observed survival), and decision curve analysis (the potential clinical benefit of using the model to guide treatment decisions). The best-performing model was designated as the East Asian Survival Tool for Bone Metastasis Surgery (EAST-BMS) and was compared with the SORG model. To allow a fair comparison, the performance of the SORG model was evaluated on the same held-out test data sets in each iteration of the leave-one-site-out validation, applying the same performance metrics used to select the final model. RESULTS: Gradient boosting survival analysis demonstrated the most favorable overall performance and was selected as the EAST-BMS. The number of outcome events used for model evaluation was 170 at 3 months and 447 at 12 months. The EAST-BMS achieved tdAUC values of 0.81 (95% confidence interval [CI] 0.78 to 0.85) at 3 months and 0.78 (95% CI 0.70 to 0.84) at 12 months, compared with 0.81 (95% CI 0.74 to 0.86) and 0.76 (95% CI 0.67 to 0.83), respectively, for the SORG model, indicating comparable ability to distinguish patients with different survival outcomes. Brier scores were 0.12 (95% CI 0.09 to 0.15) and 0.23 (95% CI 0.17 to 0.28) for EAST-BMS versus 0.14 (95% CI 0.12 to 0.16) and 0.25 (95% CI 0.15 to 0.34) for SORG, indicating lower prediction error in EAST-BMS. Calibration intercepts were -0.08 (95% CI -0.25 to 0.09) versus -1.06 (95% CI -1.26 to -0.86) at 3 months and -0.35 (95% CI -0.49 to -0.22) versus -1.23 (95% CI -1.37 to -1.08) at 12 months, indicating better agreement between predicted and observed survival in EAST-BMS. Decision curve analysis showed wider threshold probability ranges with positive net clinical benefit for EAST-BMS (0.04 to 0.96 versus 0.05 to 0.66 at 3 months; 0.17 to 0.77 versus 0.08 to 0.67 at 12 months), which means that using the EAST-BMS to guide treatment decisions may provide greater clinical benefit than the SORG model. Albumin, Karnofsky performance status, percentage of lymphocytes, and C-reactive protein level were among the most influential predictors. CONCLUSION: The EAST-BMS, the first multinational machine-learning survival model for patients from East Asia undergoing surgery for nonspinal bone metastases of which we are aware, demonstrated favorable predictive accuracy and clinical utility. This web-based tool may support personalized prognostic assessment and surgical decision-making. It is freely available as a web-based tool at https://bms.east-mskonco.org. LEVEL OF EVIDENCE: Level III, therapeutic study.
van Loon DFR, van Es EM, Siemensma MF
… +4 more, Eygendaal D, Stockmans F, Veeger DHEJ, Colaris JW
Clin Orthop Relat Res
· 2026 May · PMID 42102859
·
Publisher ↗
BACKGROUND: Despite the clinical relevance of forearm fractures and malunions and the impact of a functional limitation, the link between forearm malalignment and limited pronation and supination remains poorly understoo...BACKGROUND: Despite the clinical relevance of forearm fractures and malunions and the impact of a functional limitation, the link between forearm malalignment and limited pronation and supination remains poorly understood and still relies on anatomical alignment expressed as angulation. Using recently developed technologies, mechanisms that limit function can be automatically detected by modeling individual forearm kinematics using three-dimensional (3D) bone models of the radius and ulna. QUESTIONS/PURPOSES: We evaluated the accuracy of a personalized 3D kinematic model to identify limitations in forearm rotation in pronation and supination and to answer the following questions: (1) How accurately does the model-predicted ROM agree with the corresponding clinical measurements? (2) How accurately does the model classify malunited forearms according to the presence of clinically relevant functional limitations, defined as a range of pronation or supination less than 50°? (3) What is the frequency at which the model detects bone impingement and central band block during pronation and supination? METHODS: This retrospective study evaluated a diagnostic model using the preoperative CT scans of 45 patients with unilateral diaphyseal forearm malunions, all of whom underwent corrective osteotomy due to a clinically relevant limitation in pronation or supination function. In all, 53% (24) of patients were male; the mean ± SD age at the time of the CT scan was 16 ± 6 years, and the mean time since the original trauma was 6 ± 5 years. Twenty patients had a clinically relevant loss of pronation, 15 patients had a loss of supination, and 10 patients had a loss of both. We generated 3D bone models with landmarks to simulate forearm rotation in 5° steps from 100° of pronation to 100° of supination. Two mechanisms that limit function after diaphyseal malunions-bone impingement and central band blockage-were identified in the simulation, resulting in a predicted ROM. For the first study question, differences between clinical and predicted function were expressed as mean absolute error, root mean square error, and mean error to illustrate typical error size, penalize outliers, and quantify the direction of error deviation, respectively. Acceptable errors were around 15°, comparable to the range seen in clinical measurements. For question two, clinical measurements and predictions were dichotomized based on a threshold of 50°. Accuracy, sensitivity, specificity, positive and negative predictive values, and the area under the receiver operating characteristic (ROC) curve for detecting clinically relevant limitations were calculated separately for pronation and supination. Acceptable diagnostic values should be above 60%, which is normal for angulation measurements. For question three, the blocking mechanisms detected during the simulation were counted. RESULTS: Mean absolute errors between prediction and clinical measurement for pronation, supination, and ROM were 19°, 23°, and 22°, respectively. Root mean square errors were 22° for pronation, 28° for supination, and 28° for ROM. Mean errors were 3° for pronation, 1° for supination, and 5° for ROM. Errors were substantially higher than the clinical measurement uncertainty, with some outliers. Accuracy for finding a relevant pronation or supination limitation was 91% and 82%, respectively. Diagnostic values for detecting pronation limitations were 91% for accuracy, 87% for sensitivity, 100% for specificity, 79% for negative predictive value, and 100% for positive predictive value. For supination, the values were 82% for accuracy, 84% for sensitivity, 80% for specificity, 80% for negative predictive value, and 84% for positive predictive value. Area under the curve values were 0.97 (95% confidence interval [CI] 0.93 to 1) for detecting pronation limitations and 0.93 (95% CI 0.87 to 1) for supination limitations. These values are higher than those reported by studies using angulation thresholds. Bone impingement was mainly seen during pronation, and a central band block was the most common reason for a supination limitation. CONCLUSION: Individualized kinematic modeling of forearm malunions reliably detects clinically relevant limitations of forearm rotation without requiring dynamic imaging. Because of simplifications on the exact location and status of the central band and the neutral position of the forearm, exact ROM prediction is not possible. CLINICAL RELEVANCE: This study represents an important step toward functional rather than anatomical evaluation of forearm anatomy and correction of malunited forearm fractures. The next step would be to use the model in preoperative planning optimization, focusing on functional outcomes rather than purely anatomical correction. Given the model's high diagnostic accuracy, personalized 3D kinematic modeling has potential as a decision tool for determining whether a forearm fracture should undergo operative treatment or whether it can be managed nonoperatively. However, challenges regarding fracture remodeling and stability in a cast, along with low-dose 3D imaging, must be addressed.
Peters ST, Wouters RM, Selles RW
… +2 more, Slijper HP, and the Hand-Wrist Study Group
Clin Orthop Relat Res
· 2026 May · PMID 42102783
·
Publisher ↗
BACKGROUND: A frequently used strategy to improve patient-reported outcome measure (PROM) response rates and reduce patients' burden with PROM completion is item reduction. In this context, a shortened decision tree vers...BACKGROUND: A frequently used strategy to improve patient-reported outcome measure (PROM) response rates and reduce patients' burden with PROM completion is item reduction. In this context, a shortened decision tree version of the patient-rated wrist evaluation (DT-PRWE) was developed, reducing the number of items from 15 to 5. The DT-PRWE demonstrated excellent psychometric properties in simulated data; however, its psychometric properties have not yet been evaluated in a real-world clinical setting. QUESTIONS/PURPOSES: (1) What is the interversion reliability and agreement between the DT-PRWE and the patient-rated wrist evaluation (PRWE) in a clinical setting? (2) What is the difference in completion time between the DT-PRWE and the PRWE? METHODS: We conducted a prospective study at Xpert Clinics in the Netherlands, a multicenter, referral-based outpatient practice, with both urban and regional locations, that specializes in hand and wrist surgery and hand therapy to assess the interversion reliability and agreement between the PRWE and the DT-PRWE. Between January and April 2025, a total of 427 adult patients were treated for wrist-related conditions and completed the PRWE at baseline as part of routine outcome measurement at our clinic. Subsequently, we asked patients to complete the DT-PRWE again 5 to 10 days after the initial assessment. Of those, we considered adult patients with wrist conditions who completed both versions of the PRWE to be potentially eligible. Based on this, 55% (235 of 427) were potentially eligible; a further 27% (116) were excluded because of intervening treatment before completion of the PRWE and the DT-PRWE, including corticosteroid injection (11% [45]), surgery before completion of both versions (6% [25]), other treatment before completion of both versions (5% [23]), and concomitant treatment (5% [23]), leaving 28% (119) for analysis here. Primarily, we evaluated interversion reliability using intraclass correlation coefficients (ICC) as the main outcome; an ICC > 0.75 was considered acceptable for clinical use. We also calculated Pearson correlation coefficients. We assessed the agreement by evaluating paired between-version mean differences, standard error of measurement (SEM), and Bland-Altman plots. Additionally, we compared the SEM values with the minimum important change (MIC) thresholds of the PRWE to assess the level of agreement. Finally, we calculated the median (IQR) completion time and compared completion efficiency between versions using the paired Wilcoxon signed-rank test. RESULTS: The DT-PRWE demonstrated good interversion reliability compared with the PRWE for total score (ICC 0.88 [95% confidence interval (CI) 0.83 to 0.91]), pain subscore (ICC 0.78 [95% CI 0.69 to 0.84]), and hand function subscore (ICC 0.83 [95% CI 0.77 to 0.88]). Additionally, the scores of the PRWE and DT-PRWE were highly correlated (total score r = 0.88 [95% CI 0.83 to 0.92], pain subscore r = 0.81 [95% CI 0.74 to 0.87], hand function subscore r = 0.83 [95% CI 0.77 to 0.88]). The agreement between versions was high, with between-version mean differences of -5.3 (95% CI -7.2 to -3.5) on the total score (score range 0 to 100), -3.4 (95% CI -4.6 to -2.2) on the pain subscore (score range 0 to 50), and -2.0 (95% CI -3.2 to -0.7) on the hand function subscore (score range 0 to 50). The SEM values were 7.1 for the total score, 4.6 for the pain subscore, and 4.9 for the hand function subscore, all falling below the MIC thresholds. The Bland-Altman plots indicated high agreement. The median (IQR) time to complete the PRWE was 3 minutes 33 seconds (2 minutes 20 seconds to 7 minutes 18 seconds), whereas for the DT-PRWE it was 1 minute 5 seconds (50 seconds to 1 minute 29 seconds), representing a 70% reduction with a median difference of 2 minutes 28 seconds (p < 0.001). CONCLUSION: The DT-PRWE is a reliable alternative to the full-length version, requiring substantially less time for patients to complete. While preserving both pain and function subscores, its simple digital implementation and comparability with existing PRWE data make the DT-PRWE well suited to replace the PRWE in routine clinical practice and for research applications. Future research should focus on cross-cultural validation of the DT-PRWE and test-retest reliability. Also, future research may investigate whether implementing shortened PROMs, such as the DT-PRWE, improves compliance with PROMs. LEVEL OF EVIDENCE: Level I, diagnostic study.