Aims. To identify variables independently associated with same-day discharge (SDD) of patients following revision total knee arthroplasty (rTKA) and to develop machine learning algorithms to predict suitable candidates for outpatient rTKA. Methods. Data were obtained from the American College of Surgeons National Quality Improvement Programme (ACS-NSQIP) database from the years 2018 to 2020. Patients with elective, unilateral rTKA procedures and a total hospital length of stay between zero and four days were included. Demographic, preoperative, and intraoperative variables were analyzed. A multivariable logistic regression (MLR) model and various machine learning techniques were compared using area under the curve (AUC), calibration, and decision curve analysis. Important and significant variables were identified from the models. Results. Of the 5,600 patients included in this study, 342 (6.1%) underwent SDD. The random forest (RF) model performed the best overall, with an internally validated AUC of 0.810. The ten crucial factors favoring SDD in the RF model include operating time, anaesthesia type, age, BMI, American Society of Anesthesiologists grade, race, history of diabetes, rTKA type, sex, and smoking status. Eight of these variables were also found to be significant in the MLR model. Conclusion. The RF model displayed excellent accuracy and identified clinically important variables for determining candidates for SDD following rTKA. Machine learning techniques such as RF will allow clinicians to accurately risk-stratify their patients preoperatively, in order to optimize resources and improve patient outcomes. Cite this article: Bone Jt Open 2023;4(6):399–407

Aims

This study aimed to explore the biological and clinical importance of dysregulated key genes in osteoarthritis (OA) patients at the cartilage level to find potential biomarkers and targets for diagnosing and treating OA.

Aims

Accurate identification of the ankle joint centre is critical for estimating tibial coronal alignment in total knee arthroplasty (TKA). The purpose of the current study was to leverage artificial intelligence (AI) to determine the accuracy and effect of using different radiological anatomical landmarks to quantify mechanical alignment in relation to a traditionally defined radiological ankle centre.

Background. Implant loosening is a common cause of a poor outcome and pain after total knee arthroplasty (TKA). Despite the increase in use of expensive techniques like arthrography, the detection of prosthetic loosening is often unclear pre-operatively, leading to diagnostic uncertainty and extensive workup. The objective of this study was to evaluate the ability of a machine learning (ML) algorithm to diagnose prosthetic loosening from pre-operative radiographs, and to observe what model inputs improve the performance of the model. Methods. 754 patients underwent a first-time revision of a total joint at our institution from 2012–2018. Pre-operative X-Rays (XR) were collected for each patient. AP and lateral X-Rays, in addition to demographic and comorbidity information, were collected for each patient. Each patient was determined to have either loose or fixed prosthetics based on a manual abstraction of the written findings in their operative report, which is considered the gold standard of diagnosing prosthetic loosening. We trained a series of deep convolution neural network (CNN) models to predict if a prosthesis was found to be loose in the operating room from the pre-operative XR. Each XR was pre-processed to segment the bone, implant, and bone-implant interface. A series of CNN models were built using existing, proven CNN architectures and weights optimized to our dataset. We then integrated our best performing model with historical patient data to create a final model and determine the incremental accuracy provided by additional layers of clinical information fed into the model. The models were evaluated by its accuracy, sensitivity and specificity. Results. The CNN we built demonstrated high performance at detecting prosthetic loosening from radiographs alone. Our first model built from scratch on just the image as an input had an accuracy of 70%. Our final model which was built by fine-tuning and optimizing a publicly available model named DenseNet, combining the AP and lateral radiographs, incorporating information from the patient history, had an accuracy, sensitivity, and specificity of 98.5%, 93.9%, and 99.5% on the patients that it was trained on, and an accuracy, sensitivity, and specificity of 88.3%, 70.2%, and 95.6% on the patients it was tested on. Conclusions. The use of machine learning (ML) can accurately detect the presence of prosthetic loosening based on plain radiographs. Its accuracy is progressively enhanced when additional clinical data is added to the loosening analysis algorithm. While this type of machine learning may not be sufficient in its present state of development as a standalone metric of loosening, it is clearly a useful augment for clinical decision making in its present state. Further study and development will be needed to determine the feasibility of applying machine learning as a more definitive test in the clinical setting. For figures, tables, or references, please contact authors directly

Aims. The aim of this study was to evaluate the ability of a machine-learning algorithm to diagnose prosthetic loosening from preoperative radiographs and to investigate the inputs that might improve its performance. Methods. A group of 697 patients underwent a first-time revision of a total hip (THA) or total knee arthroplasty (TKA) at our institution between 2012 and 2018. Preoperative anteroposterior (AP) and lateral radiographs, and historical and comorbidity information were collected from their electronic records. Each patient was defined as having loose or fixed components based on the operation notes. We trained a series of convolutional neural network (CNN) models to predict a diagnosis of loosening at the time of surgery from the preoperative radiographs. We then added historical data about the patients to the best performing model to create a final model and tested it on an independent dataset. Results. The convolutional neural network we built performed well when detecting loosening from radiographs alone. The first model built de novo with only the radiological image as input had an accuracy of 70%. The final model, which was built by fine-tuning a publicly available model named DenseNet, combining the AP and lateral radiographs, and incorporating information from the patient’s history, had an accuracy, sensitivity, and specificity of 88.3%, 70.2%, and 95.6% on the independent test dataset. It performed better for cases of revision THA with an accuracy of 90.1%, than for cases of revision TKA with an accuracy of 85.8%. Conclusion. This study showed that machine learning can detect prosthetic loosening from radiographs. Its accuracy is enhanced when using highly trained public algorithms, and when adding clinical data to the algorithm. While this algorithm may not be sufficient in its present state of development as a standalone metric of loosening, it is currently a useful augment for clinical decision making. Cite this article: Bone Joint J 2020;102-B(6 Supple A):101–106

Background. Advanced technologies, like robotics, provide enhanced precision for implanting total knee arthroplasty (TKA) components; however, optimal component position and limb alignment remain unknown. This study purpose was to identify the ideal target sagittal component position and coronal limb alignment that produce optimal clinical outcomes. Methods. A retrospective review of 1,091 consecutive TKAs was performed. All TKAs were PCL retaining or sacrificing with anterior lipped (49.4%) or conforming bearings (50.6%) performed with modern perioperative protocols. Posterior tibial slope, femoral flexion, and tibiofemoral limb alignment were measured with a standardized protocols. Patients were grouped by the ‘how often does your knee feel normal?’ outcome score at latest follow-up. Machine learning algorithms were used to identify optimal alignment zones which predicted improved outcomes scores. Results. Mean age and BMI were 66 years and 34 kg/m. 2. with 67% female. Demographics and relevant covariates did not affect outcomes (p≥0.145) except for BMI (p=0.077) but the difference was not clinically significant. For sagittal alignment, approximating native tibial slope within 0 to +2° with some amount of femoral flexion within 0 to +3° (possibly up to +9°) was predictive of knees always feeling normal. For knees in preoperative varus or neutral, knees were more likely to always feel normal when postoperative tibiofemoral alignment was in varus (>−1°). Knees aligned in valgus preoperatively were more likely to always feel normal in valgus (<−7°) or varus (>−4°) postoperatively. Conclusion. Superior patient-reported outcomes correlated with approximating native tibial slope and incorporating some femoral flexion while maintaining similar preoperative coronal limb alignment. Excessive deviation from native tibial slope, excessive femoral flexion or any femoral component extension, or coronal alignment overcorrection beyond the preoperative limb alignment correlated with worse outcomes

Introduction. Identifying knee osteoarthritis patient phenotypes is relevant to assessing treatment efficacy. Biomechanical variability has not been applied to phenotyping, yet features may be related to outcomes of total knee arthroplasty (TKA), an inherently mechanical surgery. This study aimed to i) identify biomechanical phenotypes among TKA candidates based on demographic and gait mechanic similarities, and ii) compare objective gait improvements between phenotypes post-TKA. Methods. TKA patients underwent 3D gait analysis one-week pre (n=134) and one-year post-TKA (n=105). Principal component analysis was applied to frontal and sagittal knee angle and moment gait waveforms, extracting major patterns of variability. Demographics (age, sex, BMI), gait speed, and frontal and sagittal pre-TKA angle and moment principal component (PC) scores previously found to differentiate sex, osteoarthritis (OA) severity, and symptoms of TKA recipients were standardized (mean=0, SD=1, [134×15]) to perform multidimensional scaling and machine learning based hierarchical clustering. Final clusters were validated by examining inter-cluster differences at baseline and gait changes (Post. PCscore. –Pre. PCscore. ) by k-way Chi-Squared, and ANOVA tests. Results. Four (k=4) TKA candidate groups yielded optimum clustering metrics, interpreted as 1) high-functioning males, 2) older stiff-kneed males, 3) slower stiff-kneed females, and 4) high-functioning females. Pre-TKA, higher-functioning clusters (1 & 4) had more dynamic loading/un-loading kinetic patterns during stance (flexion moment PC2, 3<2<4<1, P<0.001; adduction moent PC2; 3,2<4<1; P<0.001). Post-TKA, higher-functioning clusters demonstrated less gait improvement (flexion angle ΔPC2, 1,2,4<3, P<0.001; flexion moment ΔPC2, 4<2,3, P<0.001; adduction moment ΔPC2, 1<3, P=0.01). Conclusions. TKA candidates can be characterized by four clusters, interpreted as 1) high-functioning males, 2) older stiff-kneed males, 3) slower stiff-kneed females, and 4) high-functioning females, differing by demographics and biomechanical severity features. Functional gains after TKA were cluster-specific; stiff-gait clusters experienced more improvement, while higher-functioning clusters demonstrated some functional decline. Results suggest the presence of cohorts who may not benefit functionally from TKA. Cluster profiling may aid in triaging and developing osteoarthritis management and surgical strategies that meet individual or group-level function needs. For figures, tables, or references, please contact authors directly

The COVID-19 pandemic has led to unprecedented times worldwide. From lockdowns to masks now being part of our everyday routine, to the halting of elective surgeries, the virus has touched everyone and every part of our personal and professional lives. Perhaps, now more than ever, our ability to adapt, change and persevere is critical to our survival. This year's closed meeting of The Knee Society demonstrated exactly those characteristics. When it became evident that an in-person meeting would not be feasible, The Knee Society leadership, under the direction of President John Callaghan, MD and Program Chair Craig Della Valle, MD created a unique and engaging meeting held on September 10–12, 2020. Special recognition should be given to Olga Foley and Cynthia Garcia at The Knee Society for their flexibility and creativeness in putting together a world-class flawless virtual program. The Bone & Joint Journal is very pleased to partner with The Knee Society to once again publish the proceedings of the closed meeting of the Knee Society. The Knee Society is a United States based society of highly selected members who have shown leadership in education and research in knee surgery. It invites up to 15% international members; this includes some of the key opinion leaders in knee surgery from outside the USA. Each year, the top research papers from The Knee Society meeting will be published and made available to the wider orthopaedic community in The Bone & Joint Journal. The first such proceedings were published in BJJ in 2019. International dissemination should help to fulfil the mission and vision of the Knee Society of advancing the care of patients with knee disorders through leadership, education and research. The quality of dissemination that The Bone & Joint Journal provides should enhance the profile of this work and allow a larger body of surgeons, associated healthcare professionals and patients to benefit from the expertise of the members of The Knee Society. The meeting is one of the highlights of the annual academic calendar for knee surgeons. With nearly every member in attendance virtually throughout the 3 days, the top research papers from the membership were presented and discussed in a virtual format that allowed for lively interaction and discussion. There are 75 abstracts presented. More selective proceedings with full papers will be available after a robust peer review process in 2021, both online and in The Bone & Joint Journal. The meeting commenced with the first group of scientific papers focused on Periprosthetic Joint Infection. Dr Berry and colleagues from the Mayo Clinic further help to clarify the issue of serology and aspirate results to diagnose TKA PJI in the acute postoperative setting. 177 TKA's had an aspiration within 12 weeks and 22 were proven to have PJI. Their results demonstrated that acute PJI after TKA should be suspected within 6 weeks if CRP is ≥81 mg/L, synovial WBCs are ≥8500 cells/μL, and/or synovial neutrophils≥86%. Between 6– 12 weeks, concerning thresholds include a CRP ≥ 32 mg/L, synovial WBC ≥7450, and synovial neutrophils ≥ 84%. While historically the results of a DAIR procedure for PJI have been variable, Tom Fehring's study showed promise with the local delivery of vancomycin through the Intraosseous route improved early results. New member Simon Young contrasted the efficacy of the DAIR procedure when comparing early infections to late acute hematogenous PJI. DAIR failed in 63% of late hematogenous PJIs (implant age>1 year) compared to 36% of early (<1year) PJIs. Dr Masri demonstrated in a small group of patients that those with well-functioning articulating spacers can retain their spacers for over 12 months with no difference in infection from those that had a formal two stage exchange. The mental toll of PJI was demonstrated in a longitudinal study by Doug Dennis, where patient being treated with 2 stage exchange had 4x higher rates of depression compared to patient undergoing aseptic revision. The second session focused on both postoperative issues with regards to anticoagulation and manipulation. Steven Haas demonstrated high complication rates with utilization of anticoagulation for treatment of postoperative pulmonary embolism with modern therapeutic anticoagulation (warfarin, enoxaparin, Xa inhibitors) with the Xa inhibitors demonstrating lower complication rates. Two papers focused on the topic of manipulation. Mark Pagnano presented data on timing of manipulation under anesthesia up to even past 12 months. While gains were modest, a subset of patients did achieve substantial gains in ROM > 20degrees even after 3 months post op. Dr Westrich's study demonstrated no difference in MUA outcomes with either IV sedation or neuraxial anesthesia although the length of stay was shorter in the IV sedation group. Several studies in Session II focused on kinematics and femoral component position. Dr Li's in vivo kinematic study during weightbearing flexion and gait demonstrated that several knees rotated with a lateral pivot motion and not all knees can be described with a single motion character. Dr Mayman and his group utilized a computational knee model to demonstrate that additional distal femoral resection results in increasing levels of mid -flexion instability and cautioned against the use of additional bony resection as the first line for flexion contractures. Using computer navigation, Dr Huddleston's study nicely outlined the variability in femoral component rotation to achieve a rectangular flexion gap utilizing a gap balanced method. The third session opened the meeting on Friday morning. The focus was on unicompartmental knee arthroplasty and the increasing utilization of robotic assisted total knee arthroplasty. David Murray showed using registry data that for patient with higher comorbidities (ASA >3), UKA was safer and more cost effective than TKA while Dr Della Valle's group demonstrated overall lower average healthcare costs in UKA patients compared to TKA in the first 10 years after surgery. Dr Geller assessed UKA survivorship among 3 international registries. While survivorship varied by nation and designs, certain designs consistently had better overall performance. Dr Nunley and his group showed robotic navigation UKA significantly reduced outliers in alignment and overhang compared to manual UKA. Dr Catani's data demonstrated that full thickness cartilage loss should still be considered a requirement for UKA success even with robotic assistance. Despite a high dislocation rate of 4%, Mr Dodd demonstrated high survivorship for lateral UKA despite historical contraindications. The growing evidence for robotics TKA was demonstrated in two studies. Professor Haddad showed less soft tissue injury, reduced bone trauma and improved accuracy or rTKA compared to manual TKA while Dr Gustke single surgeon study showed his rTKA had improved forgotten joint scores and less ligament releasing required for balancing. Despite these finding, Dr Lee's study demonstrated that a robotic TKA could not guarantee excellent pain relief and other factors such a patient expectations and psychological factors play a role. Our fourth session was devoted to machine learning and smart tools and modeling. Dr Meneghini used machine learning algorithms to identify optimal alignment outcomes that correlated with patient outcomes. Several parameters such as native tibial slope, femoral sagittal position and coronal limb alignment correlated with outcomes. Along the same lines, Bozic and coauthors demonstrated that using AI algorithms incorporated with PROM's improved levels of shared decision making and patient satisfaction. Dr Lombardi demonstrated that a mobile patient engagement platform that provided smart phone-based exercise and education was comparable to traditional methods. Dr Mahfouz demonstrated the accuracy of using ultrasound to produce 3D models of the bone compared to conventional CT based strategies and Dr Mahoney showed the valued of a preop 3D model in reproducing more normal knee kinematics. The last two talks of the session focused on some of the positives of the COVID-19 pandemic, namely the embracing of telemedicine by patients and surgeons as demonstrated by Dr Slover and the increasing and far reaching educational opportunities made available to residents and fellows during the pandemic. Session five focused on risk stratification and optimization prior to TKA. Dr O'Connor demonstrated that that the implementation of an optimization program preoperatively reduced length of stay and ED visits, and Charles Nelson's study showed that risk stratification tool can lower complication rates in obese patients undergoing TKA comparable to those that are nonobese. Dr Markel's study demonstrated that those who have preoperative depression and anxiety are at higher risk of complications and readmissions after surgery and these issues should be addressed preoperatively. Interestingly, a study by Dr Callaghan demonstrated that care improvement pathways have not lowered the gap in complications for morbidly obese patients undergoing TKA, Dr Barsoum argued that the overall complication rates were low and this patient cohort had significant gains in PROMS after TKA that would not be experienced if arbitrary cutoff for limited surgery were established. The final session on Friday, Session six, had several well done and interesting studies. There continues to be mounting evidence that liposomal bupivacaine has little effect on managing post-operative pain to warrant its increased use. Bill Macaulay and colleagues showed no change in pain scores, opioid consumption and functional scores when liposomal bupivacaine was discontinued at a large academic medical center. Dr Bugbee importantly demonstrated that a supervised ambulation program reduced falls in the early postoperative period. Several paper on healthcare economics were presented. Rich Iorio showed that stratifying complexity of total joint cases between hospitals with a system can be efficient and cost savings while Dr Jiranek demonstrated in his study that complex TKAs can be identified preoperatively and are associated with prolonged operative time and cost of care and consideration should be given in future reimbursement models to a complexity modifier. Dr Springer, in their evaluation of Medicare bundled payment models, demonstrated that providers and hospitals in historical bundled models that became efficient were penalized in the new model, forcing many groups to drop out and return to a fee for service model. Ron Delanois important work showed that social determinants can have a major negative impact on outcomes following TKA. Our final day on Saturday opened with Session seven, and several interesting paper on metal ions/debris in TKA. Dr Whitesides simulator study showed the absence of scratches and material loss in a ceramic TKA compared with Co-Cr TKA and suggested an advantage to this material in patients with metal sensitivity. Conversely, in a histological study of failed TKA, perivascular lymphocytic infiltration was not associated with worse clinical outcomes or differences in revision in a series of 617 aseptic revisions, 19% of which had PVLI found on histology. The Mayo group and Dr Trousdale however, noted that serum metal ion levels can be helpful in identifying implant failure in a group of revision TKAs, especially those with metallic junctions. Dr Dalury demonstrated nicely that use of maximally conforming inserts did not have a negative effect on implant loosening in a series of 76 revision TKA's at an average follow up of 7 years, while Kevin Garvin and his group showed no difference in end of stem pain between cemented and cementless stems in revision TKA. The final two studies in the session by Bolognesi and Peters respectively showed that metaphyseal cones continue to demonstrate excelled survivorship in rTKA setting despite extensive bone loss. Session eight was highlighted by a large series of revision reported by new member Dr Schwarzkopf, who showed that revision TKA done by high volume surgeons demonstrated better outcomes and lower revision rates compared to surgeon who did less than 18 rTKA's per year. Dr Maniar importantly showed that preoperatively, patients with high activity level and low pain and indicated by a high preop forgotten joint score did poorly following TKA while David Ayers nicely demonstrated that KOOS scores that assess specific postoperative outcomes can predict patient dissatisfaction after TKA. The final paper in this session by Max Courtney showed that the majority of surgical cancellations are due to medical issues, yet a minority of these undergo any intervention specifically for that condition, but they resulted in a delay of 5 months. The first two studies of Session nine focused on polyethylene thickness. Dr Backstein demonstrated no difference in KSS scores, change in ROM and aseptic revision rates based on polyethylene thickness in a series of 195 TKA's. An interesting lab study by Dr Tim Wright showed a surprising consistency in liner thickness choice among varying levels of surgeon experience that did not correlate with applied forces or gap stability estimates. Two studies looked specifically at the issue of tibial loosening and implant design. Nam and colleagues were not able to demonstrate concerning findings for increasing tibial loosening in a tibial baseplate with a shortened tibial keel at short term follow up, while Lachiewicz demonstrated a 19% revision or revision pending rate in 223 cemented fixed bearing ATTUNE TKA at a mean of 30 months. Our final session of the meeting, began with encouraging news, that despite only currently capturing about 40% of TJA's done in the US, the American Joint Replacement Registry data is representative of data in other representative US databases. An interesting study presented by Robert Barrack looked at bone remodeling in the proximal tibia after cemented and cementless TKA of two different designs. No significant difference was noted among the groups with the exception of the cemented thicker cobalt chrome tray which demonstrated significantly more bone mineral density loss. Along the same lines, a study out of Dr Bostrom's lab demonstrated treatment of a murine tibial model with iPTH prevents fibrous tissue formation and enhances bone formation in cementless implants. New Member Jamie Howard showed no difference in implant migration and kinematics of a single radius cementless design using either a measured resection or gap balancing technique and Dr Cushner show no difference in blood loss with cemented or cementless TKA with the use of TKA. The final two studies looked at staging and bilateral TKA's. Peter Sharkey showed that simultaneous TKA's were associated with higher complication compared to staged TKA and that staged TKA with less than a 90-day interval was not associated with higher risk. However, Mark Figgie showed that patients undergoing simultaneous TKA compared to staged TKA, missed 17 fewer days of work. In spite of the virtual nature of the meeting, there were some outstanding scientific interactions and the material presented will continue to generate debate and to guide the direction of knee arthroplasty as we move forwards

Aims

To map literature on prognostic factors related to outcomes of revision total knee arthroplasty (rTKA), to identify extensively studied factors and to guide future research into what domains need further exploration.

Aims

This study used an artificial neural network (ANN) model to determine the most important pre- and perioperative variables to predict same-day discharge in patients undergoing total knee arthroplasty (TKA).

Aims

The purpose of this study was to develop a personalized outcome prediction tool, to be used with knee arthroplasty patients, that predicts outcomes (lengths of stay (LOS), 90 day readmission, and one-year patient-reported outcome measures (PROMs) on an individual basis and allows for dynamic modifiable risk factors.