Aims The number of convolutional neural networks (CNN) available for fracture detection and classification is rapidly increasing. External validation of a CNN on a temporally separate (separated by time) or geographically separate (separated by location) dataset is crucial to assess generalizability of the CNN before application to clinical practice in other institutions. We aimed to answer the following questions: are current CNNs for fracture recognition externally valid?; which methods are applied for external validation (EV)?; and, what are reported performances of the EV sets compared to the internal validation (IV) sets of these CNNs? Methods The PubMed and Embase databases were systematically searched from January 2010 to October 2020 according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. The type of EV, characteristics of the external dataset, and diagnostic performance characteristics on the IV and EV datasets were collected and compared. Quality assessment was conducted using a seven-item checklist based on a modified Methodologic Index for NOn-Randomized Studies instrument (MINORS). Results Out of 1,349 studies, 36 reported development of a CNN for fracture detection and/or classification. Of these, only four (11%) reported a form of EV. One study used temporal EV, one conducted both temporal and geographical EV, and two used geographical EV. When comparing the CNN’s performance on the IV set versus the EV set, the following were found: AUCs of 0.967 (IV) versus 0.975 (EV), 0.976 (IV) versus 0.985 to 0.992 (EV), 0.93 to 0.96 (IV) versus 0.80 to 0.89 (EV), and F1-scores of 0.856 to 0.863 (IV) versus 0.757 to 0.840 (EV). Conclusion The number of externally validated CNNs in orthopaedic trauma for fracture recognition is still scarce. This greatly limits the potential for transfer of these CNNs from the developing institute to another hospital to achieve similar diagnostic performance. We recommend the use of geographical EV and statements such as the Consolidated Standards of Reporting Trials–Artificial Intelligence (CONSORT-AI), the Standard Protocol Items: Recommendations for Interventional Trials–Artificial Intelligence (SPIRIT-AI) and the Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis–Machine Learning (TRIPOD-ML) to critically appraise performance of CNNs and improve methodological rigor, quality of future models, and facilitate eventual implementation in clinical practice. Cite this article: Bone Jt Open 2021;2(10):879–885.

Background In cases of evolving myocardial injury not definitively attributed to coronary ischaemia precipitated by plaque rupture, referral for invasive coronary angiography (ICA) may be influenced by observed troponin profiles. We sought to explore association between early ICA and elevated high-sensitivity troponin T (hs-cTnT) concentrations with and without dynamic changes, to examine if there may be a hs-cTnT threshold associated with benefit from an initial ICA strategy. Methods Using published studies (hs-cTnT study n = 1937, RAPID-TnT study n = 3270) and the Fourth Universal Definition of Myocardial Infarction (MI), index presentations of patients with hs-cTnT concentrations 5-14ng/L were classified as ‘non-elevated’ (NE). Hs-cTnT greater than upper reference limit (14ng/L) were classified as ‘elevated hs-cTnT with dynamic change’ (encompassing acute myocardial injury, Type 1 MI, and Type 2 MI), or ‘non-dynamic hs-cTnT elevation’ (chronic myocardial injury). Patients with hs-cTnT <5ng/L and/or eGFR<15mmol/L/1.73m2 were excluded. ICA was performed within 30 days of admission. Primary outcome was defined as composite endpoint of death, MI, or unstable angina at 12 months. Results Altogether, 3620 patients comprising 837 (23.1%) with non-dynamic hs-cTnT elevations and 332 (9.2%) with dynamic hs-cTnT elevations were included. Primary outcome was significantly higher with dynamic and non-dynamic hs-cTnT elevations (Dynamic: HR: 4.13 95%CI:2.92–5.82; p<0.001 Non-dynamic: HR: 2.39 95% confidence interval [CI]:1.74–3.28, p<0.001). Hs-cTnT thresholds where benefit from initial ICA strategy appeared to emerge was observed at 110ng/L and 50ng/L in dynamic and non-dynamic elevations, respectively. Conclusion Early ICA appears to portend benefit in hs-cTnT elevations with and without dynamic changes, and at lower hs-cTnT threshold in non-dynamic hs-cTnT elevation. Differences compel further investigation.