As patient data continues to grow, the importance of efficient and precise analysis cannot be overstated. The employment of Generative Artificial Intelligence (AI), specifically Chat GPT-4, in the realm of medical data interpretation has been on the rise. However, its effectiveness in comparison to manual data analysis has been insufficiently investigated. This quality improvement project aimed to evaluate the accuracy and time-efficiency of Generative AI (GPT-4) against manual data interpretation within extensive datasets pertaining to patients with orthopaedic injuries. A dataset, containing details of 6,562 orthopaedic trauma patients admitted to a district general hospital over a span of two years, was reviewed. Two researchers operated independently: one utilised GPT-4 for insights via prompts, while the other manually examined the identical dataset employing Microsoft Excel and IBM® SPSS® software. Both were blinded on each other's procedures and outcomes. Each researcher answered 20 questions based on the dataset including injury details, age groups, injury specifics, activity trends and the duration taken to assess the data. Upon comparison, both GPT-4 and the manual researcher achieved consistent results for 19 out of the 20 questions (95% accuracy). After a subsequent review and refined prompts (prompt engineering) to GPT-4, the answer to the final question aligned with the manual researcher's findings. GPT-4 required just 30 minutes, a stark contrast to the manual researcher's 9-hour analytical duration. This quality improvement project emphasises the transformative potential of Generative AI in the domain of medical data analysis. GPT-4 not only paralleled the accuracy of manual analysis but also achieved this in significantly less time. For optimal accurate results, data analysis by AI can be enhanced through human oversight. Adopting AI-driven approaches, particularly in orthopaedic data interpretation, can enhance efficiency and ultimately improve patient care. We recommend future investigations on large and more varied datasets to reaffirm these outcomes

Studies and meta-analyses worldwide show an increased use of one-stage revisions for treating periprosthetic hip infections, often yielding comparable or better outcomes than two-stage revisions. However, it remains unclear if these successful results can be consistently achieved nationwide besides large centers. This observational cohort study used data from the German Arthroplasty Registry (EPRD) to compare the mortality and re-revision rates between one-stage (n=8183) and two-stage (n=657) first-time revision total hip arthroplasty (RTHA). Kaplan-Meier estimates were applied to evaluate the re-revision rate and cumulative mortality for RTHA. There was a significant difference in mortality between one-stage and two-stage RTHA (p=0.02). One-year post-surgery, the mortality rate was 9.4% for one-stage revisions and 5.5% for two-stage revisions. At the five-year follow-up, the mortality rate for one-stage revisions was 25.5%, compared to 20.0% for two-stage revisions. No significant differences (p=0.30) were found in re-revision rates between one-stage and two-stage revisions after one year (one-stage 16.5% vs. two-stage 13.5%) or five years (one-stage 21.6% vs. two-stage 20.8%). For multiple revisions, the mortality differences were even larger (p<0.001), with a one-year mortality rate of 12.8% for one-stage RTHA and 5.7% for two-stage RTHA. Despite the excellent results of one-stage RTHA in the literature from individual large centers, it shows a significantly higher mortality rate with equal re-revision rate compared to two-stage revision in the nationwide care besides large centers. Significant differences can already be seen within the first year, indicating an increased perioperative mortality for one-stage revision, which might be explained by longer surgery duration, blood-loss and patient selection or maybe a lack of experience concerning proper surgical debridement for one-stage revision. This illustrates the need to establish centers for joint-revision surgery that provide interdisciplinary care and high case numbers to improve perioperative outcomes

Introduction. The American Joint Replacement Registry (AJRR) is the largest registry of total hip and knee arthroplasty (THA and TKA) procedures performed in the U.S. The National (Nationwide) Inpatient Sample (NIS) is a public database containing demographic estimates based on more than seven million hospitalizations annually. The purpose of this study was to analyze whether AJRR data is representative of the national experience with TJA as represented in NIS. Methods. Cohen's d effect sizes were computed to ascertain the magnitude of differences in demographics, hospital volume (in 50 patient increments), and geographic characteristics between the AJRR and NIS databases. Results. The study included [NIS: 2,316,345 vs. AAOS-AJRR: 557,684] primary THA [NIS: 3,417,700 vs. AAOS-AJRR: 809,494] TKA procedures. The magnitude of distribution, as determined by the Cohen's d effect size, showed the proportions of AJRR and NIS patients were similar based on overall sex [THAs (d=0.03) and TKAs (d=0.02)] and age [THAs (d=0.17) and TKAs (d=0.12)]. Similarly, only small differences (d=0.34 or less) were identified between databases considering hospital volume and geography. AJRR was underrepresented in Southern regions and hospitals with low procedure volume and overrepresented in Northern hospitals and those with larger volume. Both NIS and AJRR followed a similar overall trend with a majority of procedures performed at hospitals with <50 cases per year. Conclusion. Distributions across hospital volume, age, and geography were proportionally similar between the AJRR and NIS databases, which suggests that AJRR data is representative of national trends and may be generalized to the larger U.S. population

Introduction. The British Spinal Registry (BSR) is a web-based database commissioned by the British Association of Spine Surgeons (BASS) in the UK. It allows auditing of spinal surgery outcomes, patient's safety and overall experience. The clinical data include patient's demographics which is entered into the Registry by medical staff, as well as patient-reported outcome measures (PROM) that is submitted to the Registry by the patient themselves at different time periods post-operatively. It has the ability to register Device and Implants as well as co-ordinate multicentre research. This study is to identify both the staff and patient compliance regarding to data submissions to the BSR at St. George's Hospital NHS Trust. Methods and Materials. Retrospective analysis of the BSR data for all spinal surgeries that was performed at SGH by the three Complex Spinal Surgery Consultants between 1st January 2017 to 31st December 2018. This study period allowed up to 12 months PROM data analysis. Staff and Patient compliance were analysed separately depend on the data they submitted. Results. 404 (n) spinal surgeries were performed over this two-year period. Surgical demographics: 39.4% (159/404) males and 56.2% (227/404) females, 4.5% (18/404) declined to be added. Age ranging from 6 to 92, mean age was 47.9. Staff compliance was 89.4% for the registration of patient data. Trauma and tumour patients were excluded as they are not required to complete any pre-operative PROMs. For the remaining patients, 35.8% (116/324) completed their baseline pre-operative PROMs. A steady decrease in patient compliance for post-operative PROMS was observed: 29.8% (2 weeks), 24.9% (6 weeks), 21.2% (6 months), 17.1% (12 months), and 16.6% (24 months). Conclusions. We identified a poor patient compliance in both pre-operative and post-operative PROMS. Patient questionnaire fatigue can occur over time which results in falling compliance to PROMS in 12 months and 24 months. Moreover, patients who do not provide an email address are less likely to fill in their PROMS. Satisfactory compliance for staff regarding to BSR utilisation are shown. Staff should aim to achieve a higher level of compliance as this will help to keep a record of all spinal surgeries that are performed and any complications that are encountered in a particular centre. Over the past three years, the BSR had introduced various schemes to drive up utilisation including a Best Practice Tariff in 2019. Part of the tariff will be withheld when staff failure to enter patients onto the Registry. Further research is needed to determine why there is such fall off especially with patient compliance and to elucidate potential measures to improve their compliance

Aims

Knee arthroplasty surgery is a highly effective treatment for arthritis and disorders of the knee. There are a wide variety of implant brands and types of knee arthroplasty available to surgeons. As a result of a number of highly publicized failures, arthroplasty surgery is highly regulated in the UK and many other countries through national registries, introduced to monitor implant performance, surgeons, and hospitals. With time, the options available within many brand portfolios have grown, with alternative tibial or femoral components, tibial insert materials, or shapes and patella resurfacings. In this study we have investigated the effect of the expansion of implant brand portfolios and where there may be a lack of transparency around a brand name. We also aimed to establish the potential numbers of compatible implant construct combinations.

The importance of registries has been brought into focus by recent UK national reports focusing on implant (Cumberlege) and surgeon (Paterson) performance. National arthroplasty registries provide real-time, real-world information about implant, hospital, and surgeon performance and allow case identification in the event of product recall or adverse surgical outcomes. They are a valuable resource for research and service improvement given the volume of data recorded and the longitunidal nature of data collection. This review discusses the current value of registry data as it relates to both clinical practice and research. Cite this article: Bone Joint J 2023;105-B(4):356–360

Cite this article: Bone Joint Res 2023;12(4):256–258.

Instant messaging via WhatsApp is used within hospital teams. Group messaging can lead to efficient and non-hierarchical communication. Despite being end-to-end encrypted, WhatsApp is owned by Facebook, raising concerns regarding data security. The aims of this study were: 1) to record the prevalence of WhatsApp group instant messaging amongst clinical teams; 2) to ascertain clinician attitudes towards use of instant messaging, 3) to gauge clinicians’ awareness of best practice regarding mobile data protection and 4) to create a practical guideline based off available literature that can be used to by clinicians to improve data security practice. Over a two-week period, clinical nurse specialists in the Auckland District Health Board Department of Orthopaedics retrospectively completed a blind audit of all messaging activity across the five teams WhatsApp group message threads, recording quantity of messages sent and the nature of the messages. Concurrently individuals in these WhatsApp groups completed an anonymous survey of their use of WhatsApp and their awareness of local data security policies and practice. A guideline adapted from available literature was created to compare current practice to recommended standards and subsequently adopted into local policy. 1360 messages were sent via WhatsApp in a two-week period. 384 (28%) of the messages contained patient identifiable data. Thirty-six photos were shared. Participants rated use of WhatsApp at 9.1/10 – extremely beneficial. Sixty-five per cent of clinicians reported they had not read or were unaware of the ADHB policies regarding mobile devices and information privacy and security. WhatsApp use is widespread within the Orthopaedic department and is the preferred platform of communication with many perceived benefits. Data security is a risk and implementation of an appropriate guideline to assist clinicians in achieving best practice is crucial to ensure patient data remains protected

Retrieval analysis has been valuable in the assessment of in-vivo surface damage of orthopedic devices. Historically, subjective techniques were used to grade damage on the implant's surface. Microscopy improved the ability to localize and quantify damage, but cannot measure volumetric wear due to this damage. Laser scanning provides volumetric wear, but lacks image data. Recent techniques superimpose image data on laser scan data (photorendering) and combine the strengths of both methods. Our goal is to use such methods to improve our damage assessment and potentially correlate this assessment to volumetric wear. This project focused on two areas: image-stitching and photorendering. Image-stitching registers multiple images into large-scale high-resolution composites. Six total disc replacement components were imaged with a digital microscope (Moticam 2, Motic). Three sets were taken of each component: a single template at 10x zoom (1×1), a 4-image composite at 18x zoom (2×2), and a 9-image composite at 18x zoom (3×3). The 2×2 and 3×3 sets were image-stitched to resemble their template counterpart. Measurement error was defined using common pixels identified between the composite and template images for comparison with a semi-automated feature detection algorithm (Figure 1). For photorendering, a pilot study was performed on a single retrieved tibial bearing. The component was imaged with a digital microscope (VHX-2000, Keyence) under a 3D image-stitching setting, providing a high-resolution photo embedded with height values. MATLAB was used to convert the image into a photo-rendered point cloud approximating the surfaces. The component was then laser scanned, creating a 3D point cloud with resolution 0.127 mm. The photo-rendered point cloud data was registered to the laser scan data using an iterative closest point algorithm (Geomagic Studio, Geomagic). An analysis of all composite images showed a mean error of 0.221 mm. Figure 2 compares regions of images for the template, 2×2, and 3×3 composites. Zooming in shows the effect of the increased resolution contained in the composite. The 2×2 and 3×3 composites had mean errors of 0.231 mm and 0.209 mm, respectively; these were not significantly different. Comparisons among image types showed that components with less features exhibited larger errors during image-stitching. Figure 3 shows images resulting from each step of the photorendering process. The final image of the figure shows a qualitative result of our ability to photorender the tibial bearing surface of the component. While combining microscopy and laser scan data works anecdotally, further analyses must be performed to assure the robustness of the technique. The digital microscope's embedded image-stitching software is limited in its maximum field of view; we look to extend this by taking multiple scans and using in-house software to generate a composite of a whole implant. The improved resolution provided by microscopy offer an opportunity to automate damage assessment, yielding damage mapped images which can also be overlaid on laser scan data. This may provide a means to better quantify observed damage and yield meaningful correlations with volumetric loss due to wear

Background. The advent of value-based conscientiousness and rapid-recovery discharge pathways presents surgeons, hospitals, and payers with the challenge of providing the same total hip arthroplasty episode of care in the safest and most economic fashion for the same fee, despite patient differences. Various predictive analytic techniques have been applied to medical risk models, such as sepsis risk scores, but none have been applied or validated to the elective primary total hip arthroplasty (THA) setting for key payment-based metrics. The objective of this study was to develop and validate a predictive machine learning model using preoperative patient demographics for length of stay (LOS) after primary THA as the first step in identifying a patient-specific payment model (PSPM). Methods. Using 229,945 patients undergoing primary THA for osteoarthritis from an administrative database between 2009– 16, we created a naïve Bayesian model to forecast LOS after primary THA using a 3:2 split in which 60% of the available patient data “built” the algorithm and the remaining 40% of patients were used for “testing.” This process was iterated five times for algorithm refinement, and model performance was determined using the area under the receiver operating characteristic curve (AUC), percent accuracy, and positive predictive value. LOS was either grouped as 1–5 days or greater than 5 days. Results. The machine learning model algorithm required age, race, gender, and two comorbidity scores (“risk of illness” and “risk of morbidity”) to demonstrate excellent validity, reliability, and responsiveness with an AUC of 0.87 after five iterations. Hospital stays of greater than 5 days for THA were most associated with increased risk of illness and risk of comorbidity scores during admission compared to 1–5 days of stay. Conclusions. Our machine learning model derived from administrative big data demonstrated excellent validity, reliability, and responsiveness after primary THA while accurately predicting LOS and identifying two comorbidity scores as key value-based metrics. Predictive data has the potential to engender a risk-based PSPM prior to primary THA and other elective orthopaedic procedures

Digital Ventilated Cages (DVC) offer an innovative technology to obtain accurate movement data from a single mouse over time [1]. Thus, they could be used to determine the occurrence of a tendon damage event as well as inform on tissue regeneration [2,3]. Therefore, using the mouse model of tendon experimental damage, in this study it has been tested whether the recovery of tissue microarchitecture and of extracellular matrix (ECM) correlates with the motion data collected through this technology. Mice models were used to induce acute injury in Achilles tendons (ATs), while healthy ones were used as control. During the healing process, the mice were housed in DVC cages (Tecniplast) to monitor animal welfare and to study biomechanics assessing movement activity, an indicator of the recovery of tendon tissue functionality. After 28 days, the AT were harvested and assessed for their histological and immunohistochemical properties to obtain a total histological score (TSH) that was then correlated to the movement data. DVC cages showed the capacity to distinguish activity patterns in groups from the two different conditions. The data collected showed that the mice with access to the mouse wheel had a higher activity as compared to the blocked wheel group, which suggests that the extra movement during tendon healing improved motion ability. The histological results showed a clear difference between different analyzed groups. The bilateral free wheel group showed the best histological recovery, offering the highest TSH score, thus confirming the results of the DVC cages and the correlation between movement activity and structural recovery. Data obtained showed a correlation between TSH and the DVC cages, displaying structural and movement differences between the tested groups. This successful correlation allows the usage of DVC type cages as a non-invasive method to predict tissue regeneration and recovery. Acknowledgements: This research is part of the P4FIT project ESR13, funded by the H2020-ITN-EJD MSCA grant agreement No.955685

The aim of this study was to improve data collection, audit and research data by integrating a comprehensive patient data collection database into the day to day running of an orthopaedic department. The day to day processes of the house surgeon and registrar junior staff were analysed and tasks identified that would allow accurate recording and recall of orthopaedic-specific patient data by automating and performing tasks that would improve junior staff efficiency. A database was then designed and implemented with a “front end” that performed such tasks as generating operating lists, tracking ward locations of patients, producing discharge summaries and auditing complications. This database was then introduced on the hospital intranet and the “back end” constructed to gather accurate patient and injury data to allow improved data collection and research. Information and data collected was a significant improvement on previous methods available to the department. Junior doctor compliance for data entry was high but the system needed monitoring and “cleaning” on a weekly basis to maintain its accuracy. This method of data collection was more accurate than anything available in the hospital and has been surprisingly useful in producing data to support concerns within the department regarding managerial changes in hospital systems. Use of a database collection system that gathers information by performing day to day tasks for junior staff has been an effective and reasonably accurate method of obtaining useful patient data. It requires regular monitoring to be most effective but has been easily integrated and accepted within the hospital system

Since 1977 we did implant ceramic on ceramic THR in younger and active population. In 1991 we published (JBJS B N°4) data's in a consecutive group of patients under 50 years of age. What about the same patients for more than 30 years? Eighty-six hips in 75 patients, 41 males 34 females, mean age 43 (18 to 50), mean weight 68 kg (36 to 100), Charnley class: 38 A, 28 (38 hips) B, 9 patients (10 hips) C. Sixty-six primary procedures, 20 revisions (18 failed arthroplasties: 6 THR, 5 resurfacing, four single cup, two hemiarthroplasty, one bipolar), one after acetabular fractures. Four hips previously infected. Eight Patients deceased (8 hips) prosthesis still in place, ten lost to follow-up before 2 years, eight hips in 8 patients were revised before the review, partially followed: from 2 to 20 years: 35, completely followed: 25 hips in 23 patients resuming in: No pain in 20, slight pain in 2, severe disability in 3 not related to the hip, no radiolucent lines in 22, radiolucent lines in 3, no osteolysis in 25. Revision for: early sepsis in one, socket loosening in 8 (3 revisions cases), femoral head fractures in 2: one extra small head (22mm) for Crowe 3 DDH, one fractured at 24 years. Inertness, stability related to fibrous tissue generation, no noise

Primary malignant bone tumours are a scarce entity with limited population-based data from developing countries. The aim of the study is to investigate the frequency and anatomical distribution of primary malignant bone tumours in a local South African population. This will be an epidemiological retrospective study. Data will be used of patients that were diagnosed with primary malignant bone tumours over a period of nine years spanning from 1 January 2014 to 31 December 2022. This data will be received from private and government laboratories. Data to be considered are type of primary malignant bone tumours diagnosed, incidence of primary malignant bone tumours over a period of nine years and the most common anatomical sites of primary malignant bone tumours. The rationale behind our study is to assess the frequency of different primary malignant bone tumours in another geographic area of South Africa and to compare these findings to local and international literature. With a projected increase in diagnosis of primary malignant bone tumours in developing countries it is important to have more available data about primary malignant bone tumours from these areas to have a better understanding of these conditions and to understand the impact of the burden they impose on healthcare systems so that management of these conditions can also be improved. Preliminary results show that 23.83% of primary malignant bone tumours occurred in the age group 0–24 years of age, 49.22% in the 25–59 age group and 26.95% in the 60+ age group. The most common tumour that occurred was chondrosarcoma (49.21%) followed by osteosarcoma (41.80%) then Ewing's sarcoma (4,69%) and lastly chordoma (4.30%). From the 256 samples that met the inclusion criteria the five most common anatomical sites were distal femur (63), proximal tibia (41), proximal humerus (38), pelvis (34) and proximal femur (20)

Data of high quality are critical for the meaningful interpretation of registry information. The National Joint Registry (NJR) was established in 2002 as the result of an unexpectedly high failure rate of a cemented total hip arthroplasty. The NJR began data collection in 2003. In this study we report on the outcomes following the establishment of a formal data quality (DQ) audit process within the NJR, within which each patient episode entry is validated against the hospital unit’s Patient Administration System and vice-versa. This process enables bidirectional validation of every NJR entry and retrospective correction of any errors in the dataset. In 2014/15 baseline average compliance was 92.6% and this increased year-on-year with repeated audit cycles to 96.0% in 2018/19, with 76.4% of units achieving > 95% compliance. Following the closure of the audit cycle, an overall compliance rate of 97.9% was achieved for the 2018/19 period. An automated system was initiated in 2018 to reduce administrative burden and to integrate the DQ process into standard workflows. Our processes and quality improvement results demonstrate that DQ may be implemented successfully at national level, while minimizing the burden on hospitals. Cite this article: Bone Jt Open 2022;3(9):716–725

Introduction. Despite the increasing numbers of ankle replacements that are being performed there are still limited studies on the survival of ankle replacements and comparisons between different implants. The primary aim of this study is to link NJR data with NHS digital data to determine the true failure rates of ankle replacements. Secondary outcomes include analysis risk factors for failure, patient demographics and outcomes of individual prosthesis. Methods. A data linkage study combined National Joint Registry Data and NHS Digital data. The primary outcome of failure is defined as the removal or exchange of any components of the implanted device inserted during ankle replacement surgery. Life tables and Kaplan Meier survival charts demonstrated survivorship. Cox proportional hazards regression models with the Breslow method used for ties were fitted to compare failure rates. Results. 5,562 primary ankle replacement were recorded on the NJR. The 1-year survivorship was 98.8% (95% CI 98.4%–99.0%), 5-year survival in 2725 patients was 90.2% (95% CI 89.2%–91.1%), and 10-year survival in 199 patients was 86.2% (95% CI 84.6%–87.6%). When using a Cox regression model for all implants with over 100 implantations using the Infinity as the reference, only the Star (Hazard ratio 1.60 95% CI 0.87–2.96) and Inbone (HR 0.38 95% CI 0.05–2.84) did not produce significantly worse survivorship. Conclusion. Ankle replacements have increased in numbers over the past decade, and the currently used implants have lower failure rates than older prosthesis. It is expected that in the future the outcomes of ankle replacements will continue to improve

Worldwide, most spine imaging is either “inappropriate” or “probably inappropriate”. The Choosing Wisely recommendation is “Do not perform imaging for lower back pain unless red flags are present.” There is currently no detailed breakdown of lower back pain diagnostic imaging performed in New Brunswick (NB) to inform future directions. A registry of spine imaging performed in NB from 2011-2019 inclusive (n=410,000) was transferred to the secure platform of the NB Institute for Data, Training and Research (NB-IRDT). The pseudonymized data included linkable institute identifiers derived from an obfuscated Medicare number, as well as information on type of imaging, location of imaging, and date of imaging. The transferred data did not include the radiology report or the test requisition. We included all lumbar, thoracic, and complete spine images. We excluded imaging related to the cervical spine, surgical or other procedures, out-of-province patients and imaging of patients under 19 years. We verified categories of X-ray, Computed Tomography (CT), and Magnetic Resonance Imaging (MRI). Red flags were identified by ICD-10 code-related criteria set out by the Canadian Institute for Health Information. We derived annual age- and sex-standardized rates of spine imaging per 100,000 population and examined regional variations in these rates in NB's two Regional Health Authorities (RHA-A and RHA-B). Age- and sex-standardized rates were derived for individuals with/without red flag conditions and by type of imaging. Healthcare utilization trends were reflected in hospital admissions and physician visits 2 years pre- and post-imaging. Rurality and socioeconomic status were derived using patients’ residences and income quintiles, respectively. Overall spine imaging rates in NB decreased between 2012 and 2019 by about 20% to 7,885 images per 100,000 people per year. This value may be higher than the Canadian average. Females had 23% higher average imaging rate than males. RHA-A had a 45% higher imaging rate than RHA-B. Imaging for red flag conditions accounted for about 20% of all imaging. X-rays imaging accounted for 67% and 75% of all imaging for RHA-A and RHA-B respectively. The proportions were 20% and 8% for CT and 13% and 17% for MRI. Two-year hospitalization rates and rates of physician visits were higher post-imaging. Females had higher age-standardized hospitalization and physician-visit rates, but the magnitude of increase was higher for males. Individuals with red flag conditions were associated with increased physician visits, regardless of the actual reason for the visit. Imaging rates were higher for rural than urban patients by about 26%. Individuals in the lowest income quintiles had higher imaging rates than those in the highest income quintiles. Physicians in RHA-A consistently ordered more images than their counterparts at RHA-B. We linked spine imaging data with population demographic data to look for variations in lumbar spine imaging patterns. In NB, as in other jurisdictions, imaging tests of the spine are occurring in large numbers. We determined that patterns of imaging far exceed the numbers expected for ‘red flag’ situations. Our findings will inform a focused approach in groups of interest. Implementing high value care recommendations pre-imaging ought to replace low-value routine imaging

Knee osteoarthritis (KOA) diagnosis is based on symptoms, assessed through questionnaires such as the WOMAC. However, the inconsistency of pain recording and the discrepancy between joint phenotype and symptoms highlight the need for objective biomarkers in KOA diagnosis. To this end, we study relationships among clinical and molecular data in a cohort of women (n=51) with Kellgren-Lawrence grade 2–3 KOA through Support Vector Machine (SVM) and a regulation network model (RNM). Clinical descriptors (i.e., pain catastrophism (CA); depression (DE); functionality (FU); joint pain (JP); rigidity (RI); sensitization (SE); synovitis (SY)) are used to classify patients. A Youden's test is performed for each classifier to determine optimal binarization thresholds for the descriptors. Thresholds are tested against patient stratification according to baseline WOMAC data from the Osteoarthritis Initiative, and the mean accuracy is 0.97. For our cohort, the data used as SVM inputs are KOA descriptors, synovial fluid (SL) proteomic measurements (n=25), and transcription factors (TF) activation obtained from RNM [2] stimulated with the SL measurements. The relative weights after classification reflect input importance. The performance of each classifier is evaluated through AUC-ROC analysis. The best classifier with clinical data is CA (AUC = 0.9), highly influenced by FU and SE, suggesting that kinesophobia is involved in pain perception. With SL input, leptin strongly influences every classifier, suggesting the importance of low-grade inflammation. When TF are used, the mean AUC is limited to 0.608, which can be related to the pleomorphic behaviour of osteoarthritic chondrocytes. Nevertheless, FU has an AUC of 0.7 with strong importance of FOXO downregulation. Though larger and longitudinal cohorts are needed, this unique combination of SVM and RNM shall help to map objectively KOA descriptors. Acknowledgements: Catalan & Spanish governments 2020FI_b00680; STRATO-PID2021126469ob-C21-2, European Commission (MSCA-TN-ETN-2020-Disc4All-955735, ERC-2021-CoG-O-Health-101044828). ICREA Academia

Osteoarthritis of the hip is common and the mainstay of surgical treatment for end-stage disease is total hip replacement. There are few RCTs comparing long-term outcomes between prostheses; therefore, surgeons and patients are reliant on single-centre case-series and recently, analysis of joint registries, when making evidence-based implant choices. We conducted a systematic review, conforming to PRISMA, of Medline and Embase in September 2017. Single-centre case-series and papers analysing registries were included. Series looking at disease-specific cohorts (other than OA), under 15 years follow-up or lacking survival analyses were excluded. Resurfacings, revisions and complex-primaries were also excluded. 2750 abstracts were screened, resulting in 299 full-text articles. Following full review 124 articles were excluded and 21 series added from references, resulting in 150 analyses of individual prostheses/constructs and 12 papers from registries. We also analysed annual reports of registries. Registry data indicated cemented prostheses tended to better outcomes at late follow-ups, whereas case-series showed cementless prostheses tended to have better survival past 15 years with revision for any reason (of stem, cup or either component) as the end-point. The discrepancy between results from registry data and single-centre case series is stark, and whilst the reasons for these differences may be multifactorial, single-centre case-series included in this review often lacked sufficient power to provide precise estimates of survival. This is contrasted to data from registries, which tended to have far greater numbers from multiple centres, allowing results to be generalised to the population. The difference between these two modes of analysis suggests bias exists in selection and outcomes from single-centre series. The varied quality of reporting in case-series make it difficult for a reader to adequately assess bias, and accurately inform contemporary decision making. Surgeons and patients should be cautious when interpreting single-centre case series and systems relying on data generated from them

This study investigates the relationships between Intervertebral Disc (IVD) morphology and biomechanics using patient-specific (PS) finite element (FE) models and poromechanical simulations. 169 3D lumbar IVD shapes from the European project MySpine (FP7-269909), spanning healthy to Pfirrmann grade 4 degeneration, were obtained from MRIs. A Bayesian Coherent Point Drift algorithm aligned meshes to a previously validated structural FE mesh of the IVD. After mesh quality analyses and Hausdorff distance measurements, mechanical simulations were performed: 8 and 16 hours of sleep and daytime, respectively, applying 0.11 and 0.54 MPa of pressure on the upper cartilage endplate (CEP). Simulation results were extracted from the anterior (ATZ) and posterior regions (PTZ) and the center of the nucleus pulposus (CNP). Data mining was performed using Linear Regression, Support Vector Machine, and eXtreme Gradient Boosting techniques. Mechanical variables of interest in DD, such as pore fluid velocity (FLVEL), water content, and swelling pressure, were examined. The morphological variables of the simulated discs were used as input features. Local morphological variables significantly impacted the local mechanical response. The local disc heights, respectively in the mid (mh), anterior (ah), and posterior (ph) regions, were key factors in general. Additionally, fluid transport, reflected by FLVEL, was greatly influenced (r2 0.69) by the shape of the upper and lower cartilage endplates (CEPs). This study suggests that disc morphology affects Mechanical variables of interest in DD. Attention should be paid to the antero-posterior distribution and local effects of disc heights. Surprisingly, the CEP morphology remotely affected the fluid transport in NP volumes around mid-height, and mechanobiological implications shall be explored. In conclusion, patient-specific IVD modeling has strong potential to unravel important correlations between IVD phenotypes and local tissue regulation. Acknowledgments: European Commission: Disc4All-MSCA-2020-ITN-ETN GA: 955735; O-Health-ERC-CoG-2021-101044828