Purpose: Artificial intelligence (AI) in healthcare is rapidly growing and offers novel options of data analysis. Machine learning (ML) represents a distinct application of AI, which is capable of generating predictions and has already been tested in different medical specialties with various approaches such as diagnostic applications, cost predictions or identification of risk factors. In orthopaedics, this technology has only recently been introduced and the literature on ML in knee arthroplasty is scarce. In this review, we aim to investigate which predictions are already feasible using ML models in knee arthroplasty to identify prerequisites for the effective use of this novel approach. For this reason, we conducted a systematic review of ML algorithms for outcome prediction in knee arthroplasty.Methods: A comprehensive search of PubMed, Medline database and the Cochrane Library was conducted to find ML applications for knee arthroplasty. All relevant articles were systematically retrieved and evaluated by an orthopaedic surgeon and a data scientist on the basis of the PRISMA statement. The search strategy yielded 225 articles of which 19 were finally assessed as eligible. A modified Coleman Methodology Score (mCMS) was applied to account for a methodological evaluation.Results: The studies presented in this review demonstrated fair to good results (AUC median 0.76/range 0.57–0.98), while heterogeneous prediction models were analysed: complications (6), costs (4), functional outcome (3), revision (2), postoperative satisfaction (2), surgical technique (1) and biomechanical properties (1) were investigated. The median mCMS was 65 (range 40–80) points.Conclusion: The prediction of distinct outcomes with ML models applying specific data is already feasible; however, the prediction of more complex outcomes is still inaccurate. Registry data on knee arthroplasty have not been fully analysed yet so that specific parameters have not been sufficiently evaluated. The inclusion of specific input data as well as the collaboration of orthopaedic surgeons and data scientists are essential prerequisites to fully utilize the capacity of ML in knee arthroplasty. Future studies should investigate prospective data with specific and longitudinally recorded parameters.Level of evidence: III.

Purpose: Diagnosing periprosthetic joint infections (PJI) are challenging and may be hampered by the presence of other causes of local inflammation. Conventional synovial and serum markers are not reliable under these circumstances. Synovial calprotectin has been recently shown as a promising biomarker for PJI in total hip (THA) and total knee arthroplasty (TKA). The aim of this study is to investigate if calprotectin is reliable for PJI diagnosis in cases with accompanying inflammation due to recent surgery, dislocation or implant breakage in primary and revision TKA and THA.Methods: Thirty-three patients were included in this prospective study between July 2019 and October 2021 (17 patients undergoing surgery < 9 months, 11 dislocations, five implant breakage, respectively). Synovial white blood cell count (WBC), percentage of polymorphonuclear neutrophils (PMC), serum C-reactive protein (CRP) and synovial calprotectin, using a lateral-flow-assay, were analysed. These parameters were tested against a modified European-Bone-and-Joint-Infection-Society (EBJIS) definition with adjusted thresholds to account for the local inflammation. Statistic quality criteria were calculated and compared using a binary classification test.Results: Seventeen patients were classified as confirmed infections according to the modified EBJIS definition (13 THA and 4 TKA). The calprotectin assay yielded a sensitivity of 0.88 (0.64, 0.99), a specificity of 0.81 (0.54, 0.96), a positive predictive value (PPV) of 0.83 (0.59, 0.96) and a negative predictive value (NPV) of 0.87 (0.60, 0.98).Conclusions: Even in the presence of local inflammation due to other, non-infectious causes, calprotectin is a reliable diagnostic parameter for the detection of a PJI in primary and revision THA and TKA.

Background: Machine Learning (ML) in arthroplasty is becoming more popular, as it is perfectly suited for prediction models. However, results have been heterogeneous so far. We hypothesize that an accurate ML model for outcome prediction in THA must be able to compute arthroplasty-specific data. In this study, we evaluate a ML approach applying data from two German arthroplasty-specific registries to predict adverse outcomes after THA, after careful evaluations of ML algorithms, outcome and input variables by an interdisciplinary team of data scientists and surgeons. Methods: Data of 1217 cases of primary THA from a single center were derived from two German arthroplasty-specific registries between 2016 to 2019. The XGBoost algorithm was adjusted and applied. Accuracy, sensitivity, specificity and AUC were calculated. Results: For the prediction of complications, the ML algorithm achieved an accuracy of 80.3%, a sensitivity of 31.0%, a specificity of 89.4% and an AUC of 64.1%. For the prediction of surgery duration, the ML algorithm yielded an accuracy of 81.7%, a sensitivity of 58.2%, a specificity of 91.6% and an AUC of 89.1%. The feature importance indicated non-linear outcomes for age, height, weight and surgeon. No relevant linear correlations were found. Conclusion: The attunement of input and output data as well as the modifications of the ML algorithm permitted the development of a feasible ML model for the prediction of complications and surgery duration.

The diagnosis of periprosthetic joint infection (PJI) requires a combination of various clinical, laboratory, microbiological and histopathological parameters. A concomitant periprosthetic fracture (PPF) further complicates the diagnosis as it causes a confounding local inflammatory response. Synovial calprotectin has been demonstrated as a promising biomarker of PJI. The purpose of the present study was to evaluate the reliability of synovial calprotectin for the pre- or intraoperative diagnosis of PJI in PFF. 30 patients with PPF and implant loosening were included in this prospective study. Synovial fluid with white blood cells and percentage of polymorphonuclear neutrophils, serum C-reactive protein, and synovial calprotectin using a lateral-flow assay were tested against the EBJIS definition with adjusted thresholds to account for the local inflammation. 14 patients were postoperatively classified as confirmed infections (ten total hip arthroplasties and fourtotal knee arthroplasties). The calprotectin assay yielded a sensitivity of 0.71 [0.48; 0.95], a specificity of 0.69 [0.46; 0.91], a positive predictive value of 0.67 [0.43; 0.91] and a negative predictive value of 0.73 [0.51; 0.96]. Calprotectin is a promising diagnostic parameter for the detection of a PJI in a PPF. The lateral flow assay offers prompt results, which may further assist the surgeon in addition to already existing parameters of PJI diagnostics to diagnose concomitant PJI in PPF during surgery.

The diagnosis and treatment of periprosthetic joint infection (PJI) currently relies on cultures, which are time-consuming and often fail. Multiplex PCR assays promise reliable and prompt results, but have been heterogeneously evaluated. In this study, we analyse multiplex PCR in pathogen identification using only tissue biopsies. 42 patients after revision arthroplasty of the hip or knee were evaluated using multiplex PCR to identify microorganisms. The patients were classified according to the diagnostic criteria published by Zimmerli et al. and the results were compared to the respective microbiological cultures. PJI was detected in 15 patients and 27 revisions were aseptic. The multiplex PCR of tissue biopsies had a sensitivity of 0.3 (95% CI 0.12–0.62), a specificity of 1.0 (0.87–1.0), a positive predictive value of 1.0 (0.48–1.0) and a negative predictive value of 0.73 (0.56–0.86). The diagnostic accuracy of multiplex PCR on tissue biopsy samples is low in comparison to routine microbiological cultures. The evaluation of tissue biopsies using multiplex PCR was prone to false negative results. However, multiplex PCR assays have the advantage of rapid pathogen identification. We therefore recommend further investigation of multiplex PCR in the setting of suspected PJI with a careful choice of specimens.

Background/Aim: Synovial calprotectin has been demonstrated as a promising biomarker for periprosthetic joint infections (PJI) in painful total hip (THA) and knee arthroplasties (TKA). However, its diagnostic utility has not been evaluated explicitly in cases with marked loosening or migration of the implant. Concerns have already been raised in cases with metallosis and severe periprosthetic osteolysis because wear-induced inflammation may yield false positive results. The purpose of this study was to evaluate calprotectin for the diagnosis of PJI in cases that preoperatively demonstrate moderate to severe periprosthetic osteolysis or implant migration as signs for implant loosening in THA and TKA. Patients and Methods: Thirty-three patients were included in this prospective study between February of 2019 and November of 2021. The extent of osteolysis was classified according to Engh et al., Paprosky et al., and the modern Knee Society Radiographic Evaluation and Scoring System. Synovial white blood cell count (WBC), percentage of polymorphonuclear neutrophils (PMC), serum C-reactive protein (CRP) and synovial calprotectin using a lateral-flow-assay were tested against the European Bone and Joint Infection Society (EBJIS) definition for PJI. Statistic quality criteria were calculated and compared using a binary classification test. Results: Ten patients were classified as confirmed infections according to the EBJIS definition (7 THA and 5 TKA). The calprotectin assay yielded a sensitivity of 0.60, a specificity of 0.61, a positive predictive value of 0.40, and a negative predictive value of 0.78. The calprotectin assay resulted in nine false positive and four false negative cases. No correlation between the extent of osteolysis and false classification by means of the calprotectin assay was observed. Conclusion: The diagnostic accuracy of synovial calprotectin is impaired if moderate to severe signs of implant loosening are present. If PJI is unlikely, the calprotectin LFT can be applied as a further exclusion tool as the negative predictive value remains relatively high.