Aims. Machine learning (ML) holds significant promise in optimizing various aspects of total shoulder arthroplasty (TSA), potentially improving patient outcomes and enhancing surgical decision-making. The aim of this systematic review was to identify ML algorithms and evaluate their effectiveness, including those for predicting clinical outcomes and those used in image analysis. Methods. We searched the PubMed, EMBASE, and Cochrane Central Register of Controlled Trials databases for studies applying ML algorithms in TSA. The analysis focused on dataset characteristics, relevant subspecialties, specific ML algorithms used, and their performance outcomes. Results. Following the final screening process, 25 articles satisfied the eligibility criteria for our review. Of these, 60% focused on tabular data while the remaining 40% analyzed image data. Among them, 16 studies were dedicated to developing new models and nine used transfer learning to leverage existing pretrained models. Additionally, three of these models underwent external validation to confirm their reliability and effectiveness. Conclusion. ML algorithms used in TSA demonstrated fair to good performance, as evidenced by the reported metrics. Integrating these models into daily clinical practice could revolutionize TSA, enhancing both surgical precision and patient outcome predictions. Despite their potential, the lack of transparency and generalizability in many current models poses a significant challenge, limiting their clinical utility. Future research should prioritize addressing these limitations to truly propel the field forward and maximize the benefits of ML in enhancing patient care. Cite this article: Bone Jt Open 2025;6(2):126–134

Aims

Autologous bone graft (ABG) is considered the ‘gold standard’ among graft materials for bone regeneration. However, complications including limited availability, donor site morbidity, and deterioration of regenerative capacity over time have been reported. P-15 is a synthetic peptide that mimics the cell binding domain of Type-I collagen. This peptide stimulates new bone formation by enhancing osteogenic cell attachment, proliferation, and differentiation. The objective of this study was to conduct a systematic literature review to determine the clinical efficacy and safety of P-15 peptide in bone regeneration throughout the skeletal system.

Aims

Older adults with hip fractures are at high risk of experiencing complications after surgery, but estimates of the rate of specific complications vary by study design and follow-up period. The aim of this systematic review was to determine the prevalence of complications in older adults after hip fracture surgery.

Aims. Frozen shoulder is a common and debilitating condition characterized by pain and restricted movement at the glenohumeral joint. Various treatment methods have been explored to alleviate symptoms, with suprascapular nerve block (SSNB) emerging as a promising intervention. This meta-analysis aimed to assess the effectiveness of SSNB in treating frozen shoulder. Methods. The study protocol was registered with PROSPERO (CRD42023475851). We searched the MEDLINE, Embase, and Cochrane Library databases in November 2023. Randomized controlled trials (RCTs) comparing SSNB against other interventions were included. The primary outcome was any functional patient-reported outcome measure. Secondary outcomes were the visual analogue scale (VAS) for pain, range of motion (ROM), and complications. Risk of bias was assessed using the Cochrane risk of bias v. 2.0 tool. Results. A total of 12 RCTs were identified (702 patients; mean age 55 years (30 to 72)). Eight RCTs were deemed “low” risk-of-bias and four raised “some concerns”. Comparator interventions included intra-articular steroid injection (IAI), hydrodistension, physiotherapy, and placebo injection with 0.9% saline. Seven studies compared SSNB to IAI, with SSNB resulting in greater improvement in the Shoulder and Pain Disability Index (mean difference -4.75 (95% CI -8.11 to -1.39); p = 0.006) and external rotation (mean difference 11.64 (95% CI -0.05 to 23.33); p = 0.050). In three studies, SSNB demonstrated better VAS (mean difference -0.31 (95% CI -0.53 to 1.79); p = 0.004) compared to physiotherapy (with or without placebo injection). One study favoured hydrodistension over SSNB in improving ROM and VAS. There was no significant difference in outcomes between SSNB administered under ultrasound guidance or using a landmark technique. Conclusion. SSNB can be administered in the outpatient clinic with promising outcomes compared to IAI or physiotherapy based on level I evidence. It can therefore be considered as a first-line treatment option. Cite this article: Bone Joint J 2025;107-B(1):19–26

Aims. The Exeter femoral stem has a cemented, polished taper-slip design, and an excellent track record. The current range includes short-length options for various offsets, but less is known about the performance of these stems. The aim of this study was to compare the survival of short-length stems with standard-length Exeter stems. Methods. A systematic review of all studies reporting the use of short-length Exeter stems in primary total hip arthroplasty (THA) was undertaken. Survival data, the indication for revision, and patient-reported outcomes were gathered from observational and randomized studies. Studies based on registry data were analyzed separately. Results. The review included nine studies which covered a total of 2,190 short stems. The mean follow-up was 6.4 years (2 to 12) with an all-cause survival of 95.4%. When revisions due to aseptic acetabular loosening were excluded, the survival was 97.7%. Four stems fractured (0.18%). The mean Oxford Hip Score improved from 18.29 (1.33 to 21.6) preoperatively to 41.59 (32.9 to 43.4) at final follow-up. Three studies used data from national registries. A total of 25,895 short stems (offset ≤ 35.5 mm) were used compared with 336,218 standard-length stems. In these studies, short stems had a hazard ratio (HR) for all-cause survival of 1.19 (95% CI 0.96 to 1.43) with a rate of revision per 1,000 component-years of 0.037 compared with 0.035 for standard-length stems. One study from the New Zealand Joint Registry divided short stems into standard and small offset groups. Standard offset short stems (≥ 37.5 mm) had a 0.84 HR (95% CI 0.38 to 1.88) while small offset short stems (≤ 35.5 mm) had a 1.6 HR compared with standard stems (95% CI 1.3 to 1.98). Conclusion. Short Exeter stems perform well and are a safe femoral component in primary THA, according to the current literature. There does not appear to be an increased risk of implant fracture associated with these stems. Cite this article: Bone Joint J 2025;107-B(1):27–33

Aims

This meta-analysis and systematic review aimed to comprehensively investigate the effects of vitamin K supplementation on bone mineral density (BMD) at various sites and bone metabolism in middle-aged and older adults.

Aims. Lower limb reconstruction (LLR) has a profound impact on patients, affecting multiple areas of their lives. Many patient-reported outcome measures (PROMs) are employed to assess these impacts; however, there are concerns that they do not adequately capture all outcomes important to patients, and may lack content validity in this context. This review explored whether PROMs used with adults requiring, undergoing, or after undergoing LLR exhibited content validity and adequately captured outcomes considered relevant and important to patients. Methods. A total of 37 PROMs were identified. Systematic searches were performed to retrieve content validity studies in the adult LLR population, and hand-searches used to find PROM development studies. Content validity assessments for each measure were performed following Consensus-based Standards for the selection of health measurement Instruments (COSMIN) guidelines. A mapping exercise compared all PROMs to a conceptual framework previously developed by the study team (‘the PROLLIT framework’) to explore whether each PROM covered important and relevant concepts. Results. The systematic searches found 13 studies, while hand searches found 50 PROM development studies, and copies of all 37 measures. Although several studies discussed content validity, none were found which formally assessed this measurement property in the adult LLR population. Development of many PROMs was rated as inadequate, no PROM had sufficient content validity in the study population, and none covered all areas of the PROLLIT framework. The LIMB-Q was the most promising and comprehensive measure assessed, although further validation in a wider sample of LLR patients was recommended. Conclusion. Current PROMs used in adults requiring, undergoing, or after undergoing LLR lack content validity and do not assess all important and relevant outcomes. There is an urgent need for improved outcome measurement in this population. This can be achieved through development of a new PROM, or through validation of existing measures in representative samples. Cite this article: Bone Jt Open 2024;5(12):1049–1066

Aims. Day-case knee and hip replacement, in which patients are discharged on the day of surgery, has been gaining popularity during the last two decades, and particularly since the COVID-19 pandemic. This systematic review presents the evidence comparing day-case to inpatient-stay surgery. Methods. A systematic literature search was performed of MEDLINE, Embase, and grey literature databases to include all studies which compare day-case with inpatient knee and hip replacement. Meta-analyses were performed where appropriate using a random effects model. The protocol was registered prospectively (PROSPERO CRD42023392811). Results. A total of 38 studies were included, with a total of 83,888 day-case procedures. The studies were predominantly from the USA and Canada, observational, and with a high risk of bias. Day-case patients were a mean of 2.08 years younger (95% CI 1.05 to 3.12), were more likely to be male (odds ratio (OR) 1.3 (95% CI 1.19 to 1.41)), and had a lower mean BMI and American Society of Anesthesiologists grades compared with inpatients. Overall, day-case surgery was associated with significantly lower odds of readmission (OR 0.83 (95% CI 0.73 to 0.96); p = 0.009), subsequent emergency department attendance (OR 0.62 (95% CI 0.48 to 0.79); p < 0.001), and complications (OR 0.7 (95% CI 0.55 to 0.89) p = 0.004), than inpatient surgery. There were no significant differences in the rates of reoperation or mortality. The overall rate of successful same-day discharge for day-case surgery was 85% (95% CI 81 to 88). Patient-reported outcome measures and cost-effectiveness were either equal or favoured day-case. Conclusion. Within the limitations of the literature, in particular the substantial risk of selection bias, the outcomes following day-case knee and hip replacement appear not to be inferior to those following an inpatient stay. The evidence is more robust for unicompartmental knee replacement (UKR) than for total knee replacement (TKR) or total hip replacement (THR). The rate of successful same-day discharge is highest in UKR, followed by TKR and comparatively lower in THR. Cite this article: Bone Joint J 2024;106-B(12):1385–1392

Aims. Machine learning (ML), a branch of artificial intelligence that uses algorithms to learn from data and make predictions, offers a pathway towards more personalized and tailored surgical treatments. This approach is particularly relevant to prevalent joint diseases such as osteoarthritis (OA). In contrast to end-stage disease, where joint arthroplasty provides excellent results, early stages of OA currently lack effective therapies to halt or reverse progression. Accurate prediction of OA progression is crucial if timely interventions are to be developed, to enhance patient care and optimize the design of clinical trials. Methods. A systematic review was conducted in accordance with PRISMA guidelines. We searched MEDLINE and Embase on 5 May 2024 for studies utilizing ML to predict OA progression. Titles and abstracts were independently screened, followed by full-text reviews for studies that met the eligibility criteria. Key information was extracted and synthesized for analysis, including types of data (such as clinical, radiological, or biochemical), definitions of OA progression, ML algorithms, validation methods, and outcome measures. Results. Out of 1,160 studies initially identified, 39 were included. Most studies (85%) were published between 2020 and 2024, with 82% using publicly available datasets, primarily the Osteoarthritis Initiative. ML methods were predominantly supervised, with significant variability in the definitions of OA progression: most studies focused on structural changes (59%), while fewer addressed pain progression or both. Deep learning was used in 44% of studies, while automated ML was used in 5%. There was a lack of standardization in evaluation metrics and limited external validation. Interpretability was explored in 54% of studies, primarily using SHapley Additive exPlanations. Conclusion. Our systematic review demonstrates the feasibility of ML models in predicting OA progression, but also uncovers critical limitations that currently restrict their clinical applicability. Future priorities should include diversifying data sources, standardizing outcome measures, enforcing rigorous validation, and integrating more sophisticated algorithms. This paradigm shift from predictive modelling to actionable clinical tools has the potential to transform patient care and disease management in orthopaedic practice. Cite this article: Bone Joint J 2024;106-B(11):1216–1222

Aims. To determine whether obesity and malnutrition have a synergistic effect on outcomes from skeletal trauma or elective orthopaedic surgery. Methods. Electronic databases including MEDLINE, Global Health, Embase, Web of Science, ScienceDirect, and PEDRo were searched up to 14 April 2024, as well as conference proceedings and the reference lists of included studies. Studies were appraised using tools according to study design, including the Oxford Levels of Evidence, the Institute of Health Economics case series quality appraisal checklist, and the CLARITY checklist for cohort studies. Studies were eligible if they reported the effects of combined malnutrition and obesity on outcomes from skeletal trauma or elective orthopaedic surgery. Results. A total of eight studies (106,319 patients) were included. These carried moderate to high risk of bias. Combined obesity and malnutrition did not lead to worse outcomes in patients undergoing total shoulder arthroplasty or repair of proximal humeral fractures (two retrospective cohort studies). Three studies (two retrospective cohort studies, one case series) found that malnourishment and obesity had a synergistic effect and led to poor outcomes in total hip or knee arthroplasty, including longer length of stay and higher complication rates. One retrospective cohort study pertaining to posterior lumbar fusion found that malnourished obese patients had higher odds of developing surgical site infection and sepsis, as well as higher odds of requiring a revision procedure. Conclusion. Combined malnutrition and obesity have a synergistic effect and lead to poor outcomes in lower limb procedures. Appropriate preoperative optimization and postoperative care are required to improve outcomes in this group of patients. Cite this article: Bone Joint J 2024;106-B(10):1044–1049

Aims. The aim of this study was to perform a systematic review and bias evaluation of the current literature to create an overview of risk factors for re-revision following revision total knee arthroplasty (rTKA). Methods. A systematic search of MEDLINE and Embase was completed in accordance with the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines. The studies were required to include a population of index rTKAs. Primary or secondary outcomes had to be re-revision. The association between preoperative factors and the effect on the risk for re-revision was also required to be reported by the studies. Results. The search yielded 4,847 studies, of which 15 were included. A majority of the studies were retrospective cohorts or registry studies. In total, 26 significant risk factors for re-revision were identified. Of these, the following risk factors were consistent across multiple studies: age at the time of index revision, male sex, index revision being partial revision, and index revision due to infection. Modifiable risk factors were opioid use, BMI > 40 kg/m. 2. , and anaemia. History of one-stage revision due to infection was associated with the highest risk of re-revision. Conclusion. Overall, 26 risk factors have been associated with an increased risk of re-revision following rTKA. However, various levels of methodological bias were found in the studies. Future studies should ensure valid comparisons by including patients with identical indications and using clear definitions for accurate assessments. Cite this article: Bone Jt Open 2024;5(8):644–651

Aims. Conventional patient-reported surveys, used for patients undergoing total hip arthroplasty (THA), are limited by subjectivity and recall bias. Objective functional evaluation, such as gait analysis, to delineate a patient’s functional capacity and customize surgical interventions, may address these shortcomings. This systematic review endeavours to investigate the application of objective functional assessments in appraising individuals undergoing THA. Methods. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were applied. Eligible studies of THA patients that conducted at least one type of objective functional assessment both pre- and postoperatively were identified through Embase, Medline/PubMed, and Cochrane Central database-searching from inception to 15 September 2023. The assessments included were subgrouped for analysis: gait analysis, motion analysis, wearables, and strength tests. Results. A total of 130 studies using 15 distinct objective functional assessment methods (FAMs) were identified. The most frequently used method was instrumented gait/motion analysis, followed by the Timed-Up-and-Go test (TUG), 6 minute walk test, timed stair climbing test, and various strength tests. These assessments were characterized by their diagnostic precision and applicability to daily activities. Wearables were frequently used, offering cost-effectiveness and remote monitoring benefits. However, their accuracy and potential discomfort for patients must be considered. Conclusion. The integration of objective functional assessments in THA presents promise as a progress-tracking modality for improving patient outcomes. Gait analysis and the TUG, along with advancing wearable sensor technology, have the potential to enhance patient care, surgical planning, and rehabilitation. Cite this article: Bone Joint J 2024;106-B(8):764–774

Aims. To systematically review the predominant complication rates and changes to patient-reported outcome measures (PROMs) following osteochondral allograft (OCA) transplantation for shoulder instability. Methods. This systematic review, following PRISMA guidelines and registered in PROSPERO, involved a comprehensive literature search using PubMed, Embase, Web of Science, and Scopus. Key search terms included “allograft”, “shoulder”, “humerus”, and “glenoid”. The review encompassed 37 studies with 456 patients, focusing on primary outcomes like failure rates and secondary outcomes such as PROMs and functional test results. Results. A meta-analysis of primary outcomes across 17 studies revealed a dislocation rate of 5.1% and an increase in reoperation rates from 9.3% to 13.7% post-publication bias adjustment. There was also a noted rise in conversion to total shoulder arthroplasty and incidence of osteoarthritis/osteonecrosis over longer follow-up periods. Patient-reported outcomes and functional tests generally showed improvement, albeit with notable variability across studies. A concerning observation was the consistent presence of allograft resorption, with rates ranging from 33% to 80%. Comparative studies highlighted similar efficacy between distal tibial allografts and Latarjet procedures in most respects, with some differences in specific tests. Conclusion. OCA transplantation presents a promising treatment option for shoulder instability, effectively addressing both glenoid and humeral head defects with favourable patient-reported outcomes. These findings advocate for the inclusion of OCA transplantation in treatment protocols for shoulder instability, while also emphasizing the need for further high-quality, long-term research to better understand the procedure’s efficacy profile. Cite this article: Bone Jt Open 2024;5(7):570–580

Aims. Proximal humeral fractures are the third most common fracture among the elderly. Complications associated with fixation include screw perforation, varus collapse, and avascular necrosis of the humeral head. To address these challenges, various augmentation techniques to increase medial column support have been developed. There are currently no recent studies that definitively establish the superiority of augmented fixation over non-augmented implants in the surgical treatment of proximal humeral fractures. The aim of this systematic review and meta-analysis was to compare the outcomes of patients who underwent locking-plate fixation with cement augmentation or bone-graft augmentation versus those who underwent locking-plate fixation without augmentation for proximal humeral fractures. Methods. The search was carried out according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines. Articles involving patients with complex proximal humeral fractures treated using open reduction with locking-plate fixation, with or without augmentation, were considered. A meta-analysis of comparative studies comparing locking-plate fixation with cement augmentation or with bone-graft augmentation versus locking-plate fixation without augmentation was performed. Results. A total of 19 studies were included in the qualitative synthesis, and six comparative studies were included in the meta-analysis. Overall, 120 patients received locking-plate fixation with bone-graft augmentation, 179 patients received locking-plate fixation with cement augmentation, and 336 patients received locking-plate fixation without augmentation. No statistically relevant differences between the augmented and non-augmented cohorts were found in terms of the Disabilities of the Arm, Shoulder and Hand questionnaire score and Constant-Murley Score. The cement-augmented group had a significantly lower rate of complications compared to the non-augmented group. Conclusion. While locking-plate fixation with cement augmentation appears to produce a lower complication rate compared to locking-plate fixation alone, functional outcomes seem comparable between augmented and non-augmented techniques. Cite this article: Bone Joint J 2024;106-B(7):646–655

Aims

Intra-articular (IA) injection may be used when treating hip osteoarthritis (OA). Common injections include steroids, hyaluronic acid (HA), local anaesthetic, and platelet-rich plasma (PRP). Network meta-analysis allows for comparisons between two or more treatment groups and uses direct and indirect comparisons between interventions. This network meta-analysis aims to compare the efficacy of various IA injections used in the management of hip OA with a follow-up of up to six months.

Aims

The aims of this study were to identify and evaluate the current literature examining the prognostic factors which are associated with failure of total elbow arthroplasty (TEA).

Aims. Robotic-assisted unicompartmental knee arthroplasty (R-UKA) has been proposed as an approach to improve the results of the conventional manual UKA (C-UKA). The aim of this meta-analysis was to analyze the studies comparing R-UKA and C-UKA in terms of clinical outcomes, radiological results, operating time, complications, and revisions. Methods. The literature search was conducted on three databases (PubMed, Cochrane, and Web of Science) on 20 February 2024 according to the guidelines for Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA). Inclusion criteria were comparative studies, written in the English language, with no time limitations, on the comparison of R-UKA and C-UKA. The quality of each article was assessed using the Downs and Black Checklist for Measuring Quality. Results. Among the 3,669 articles retrieved, 21 studies on 19 series of patients were included. A total of 3,074 patients (59.5% female and 40.5% male; mean age 65.2 years (SD 3.9); mean BMI 27.4 kg/m. 2. (SD 2.2)) were analyzed. R-UKA obtained a superior Knee Society Score improvement compared to C-UKA (mean difference (MD) 4.9; p < 0.001) and better Forgotten Joint Score postoperative values (MD 5.5; p = 0.032). The analysis of radiological outcomes did not find a statistically significant difference between the two approaches. R-UKA showed longer operating time (MD 15.6; p < 0.001), but reduced complication and revision rates compared to C-UKA (5.2% vs 10.1% and 4.1% vs 7.2%, respectively). Conclusion. This meta-analysis showed that the robotic approach for UKA provided a significant improvement in functional outcomes compared to the conventional manual technique. R-UKA showed similar radiological results and longer operating time, but reduced complication and revision rates compared to C-UKA. Overall, R-UKA seems to provide relevant benefits over C-UKA in the management of patients undergoing UKA. Cite this article: Bone Jt Open 2024;5(5):374–384

Aims. Posterior malleolar (PM) fractures are commonly associated with ankle fractures, pilon fractures, and to a lesser extent tibial shaft fractures. The tibialis posterior (TP) tendon entrapment is a rare complication associated with PM fractures. If undiagnosed, TP entrapment is associated with complications, ranging from reduced range of ankle movement to instability and pes planus deformities, which require further surgeries including radical treatments such as arthrodesis. Methods. The inclusion criteria applied in PubMed, Scopus, and Medline database searches were: all adult studies published between 2012 and 2022; and studies written in English. Outcome of TP entrapment in patients with ankle injuries was assessed by two reviewers independently. Results. Four retrospective studies and eight case reports were accepted in this systematic review. Collectively there were 489 Pilon fractures, 77 of which presented with TP entrapment (15.75%). There were 28 trimalleolar fractures, 12 of which presented with TP entrapment (42.86%). All the case report studies reported inability to reduce the fractures at initial presentation. The diagnosis of TP entrapment was made in the early period in two (25%) cases, and delayed diagnosis in six (75%) cases reported. Using modified Clavien-Dindo complication classification, 60 (67%) of the injuries reported grade IIIa complications and 29 (33%) grade IIIb complications. Conclusion. TP tendon was the commonest tendon injury associated with pilon fracture and, to a lesser extent, trimalleolar ankle fracture. Early identification using a clinical suspicion and CT imaging could lead to early management of TP entrapment in these injuries, which could lead to better patient outcomes and reduced morbidity. Cite this article: Bone Jt Open 2024;5(3):252–259

Aims. Perthes' disease (PD) is a relatively rare syndrome of idiopathic osteonecrosis of the proximal femoral epiphysis. Treatment for Perthes' disease is controversial due to the many options available, with no clear superiority of one treatment over another. Despite having few evidence-based approaches, many patients with Perthes' disease are managed surgically. Positive outcome reporting, defined as reporting a study variable producing statistically significant positive (beneficial) results, is a phenomenon that can be considered a proxy for the strength of science. This study aims to conduct a systematic literature review with the hypothesis that positive outcome reporting is frequent in studies on the treatment of Perthes' disease. Methods. We conducted a systematic review of all available abstracts associated with manuscripts in English or with English translation between January 2000 and December 2021, dealing with the treatment of Perthes' disease. Data collection included various study characteristics, surgical versus non-surgical management, treatment modality, mean follow-up time, analysis methods, and clinical recommendations. Results. Our study included 130 manuscripts. Overall, 110 (85%) reported positive (beneficial) results, three (2%) reported negative results, and 17 (13%) reported no significant difference. Despite only 10/130 studies (8%) having a testable hypothesis, 71 (55%) recommended the use of their studied treatment methods for the patients, five (4%) made recommendations against the use of the studied treatment modality, and 54 (42%) did not make any recommendations. Conclusion. The overall rate for positive outcomes among included manuscripts regarding different treatment methods for Perthes' disease (85%) is higher than the 74% positive outcome rate found among studies for other surgically treated disorders and significantly higher than most scientific literature. Despite the lack of testable hypotheses, most manuscripts recommended their studied treatment method as a successful option for managing patients solely based on the reporting of retrospective data. Cite this article: Bone Joint J 2024;106-B(2):121–127

Aims. Machine-learning (ML) prediction models in orthopaedic trauma hold great promise in assisting clinicians in various tasks, such as personalized risk stratification. However, an overview of current applications and critical appraisal to peer-reviewed guidelines is lacking. The objectives of this study are to 1) provide an overview of current ML prediction models in orthopaedic trauma; 2) evaluate the completeness of reporting following the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) statement; and 3) assess the risk of bias following the Prediction model Risk Of Bias Assessment Tool (PROBAST) tool. Methods. A systematic search screening 3,252 studies identified 45 ML-based prediction models in orthopaedic trauma up to January 2023. The TRIPOD statement assessed transparent reporting and the PROBAST tool the risk of bias. Results. A total of 40 studies reported on training and internal validation; four studies performed both development and external validation, and one study performed only external validation. The most commonly reported outcomes were mortality (33%, 15/45) and length of hospital stay (9%, 4/45), and the majority of prediction models were developed in the hip fracture population (60%, 27/45). The overall median completeness for the TRIPOD statement was 62% (interquartile range 30 to 81%). The overall risk of bias in the PROBAST tool was low in 24% (11/45), high in 69% (31/45), and unclear in 7% (3/45) of the studies. High risk of bias was mainly due to analysis domain concerns including small datasets with low number of outcomes, complete-case analysis in case of missing data, and no reporting of performance measures. Conclusion. The results of this study showed that despite a myriad of potential clinically useful applications, a substantial part of ML studies in orthopaedic trauma lack transparent reporting, and are at high risk of bias. These problems must be resolved by following established guidelines to instil confidence in ML models among patients and clinicians. Otherwise, there will remain a sizeable gap between the development of ML prediction models and their clinical application in our day-to-day orthopaedic trauma practice. Cite this article: Bone Jt Open 2024;5(1):9–19