Patient Factors That Matter in Predicting Hip Arthroplasty Outcomes: A Machine-Learning Approach

Associations of cognitive appraisal and patient activation on disability and mental health outcomes: a prospective cohort study of patients undergoing spine surgery

BACKGROUND

With the increased use of patient-reported outcomes measures (PROMs) to assess spine surgery outcomes, it is important to understand how patients interpret their health changes over time. The measurement of cognitive-appraisal processes enables the quantification of how individuals think about quality of life (QOL). This study examined how appraisal processes were associated with patients' views of their role in managing their health-patient activation.

METHODS

This longitudinal cohort study from August 2019 to January 2022 included 222 adults undergoing spine surgery for cervical (n = 107) and/or lumbar (n = 148) pathology at an academic medical center. PROMs assessed disability (Neck Disability Index for cervical or Oswestry Disability Index for lumbar) and mental health (PROMIS-29 v2.0), cognitive-appraisal processes (QOLAPv2-SF), and patient activation (Patient Activation Measure). ANOVA models were used to examine the relationships between QOL and cognitive appraisal processes before and after surgery, overall and stratified by patient-activation stage. Effect sizes facilitated interpretation.

RESULTS

There were significant improvements in pain-related disability and mental health following surgery. Cognitive appraisal processes explained substantial amounts of variance, particularly with changes in mental health (45% before surgery, 75% at three months, and 63%, at 12-months after surgery). With respect to physical disability, less disability was associated with a lesser focus on negative aspects of QOL. Appraisal explained the most variance before surgery for high-activation patients. At 12-months post-surgery, however, appraisal explained the most variance for the low-activation patients. Appraisal explained similar amounts of variance in mental health at baseline and three-months post-surgery for all activation groups, but substantially more variance in the low-activation group at 12-months post-surgery. There were differences in the direction of appraisal-outcome associations by activation group in selected appraisal items/domains.

CONCLUSIONS

Cognitive-appraisal processes demonstrate a significant relationship with QOL among spine surgery patients. These processes explain substantial variance in pain-related disability and mental health, especially among those high in activation before surgery and those low in activation at 12-months post-surgery. Our findings suggest that patients' ways of thinking about their health may be effective targets of motivational coaching, to help them become more engaged over the recovery trajectory.

Mental health after lumbar spine surgery: cognitive appraisal processes and outcome in a longitudinal cohort study

BACKGROUND CONTEXT

A not uncommon finding following spine surgery is that many patients do not achieve mental health improvement up to population norms for their age cohort, despite improvement in pain and functioning.

PURPOSE

This study examined how patients who were categorized as depressed versus not depressed think about health-related quality of life as assessed by cognitive-appraisal processes. It examined cross-sectional and longitudinal differences over 12 months postsurgery.

DESIGN

Prospective longitudinal cohort study with data collected at presurgery and at ∼3- and ∼12-months postsurgery from August 2013 to August 2023.

PATIENT SAMPLE

We included 173 adults undergoing lumbar spine surgery for degenerative spinal conditions at an academic medical center. The study sample was 47% female, with a mean age of 61 (SD=15.0), and a median level of education of college graduate.

OUTCOME MEASURES

Depression was defined as a Mental Component Score (MCS)≤38 on the Rand-36, building on studies that equated MCS scores with significant depression as assessed by clinically validated depression scales. The Quality-of-Life Appraisal Profile assessed the cognitive-appraisal domains of Experience Sampling and Standards of Comparison.

METHODS

The analysis focused on two comparisons: cross-sectionally comparing those who were not depressed (n=82) to those who were depressed (n=77) at baseline; and comparing longitudinal trajectories among those depressed before surgery and improved (n=54) versus did not improve (n=23). T-tests characterized group differences in appraisal endorsement; analysis of variance evaluated appraisal items in terms of explained variance; and Pearson correlation coefficients assessed direction of association in predicting mental health.

RESULTS

There were presurgical and longitudinal differences in both cognitive appraisal domains. Before surgery, depressed patients were less likely than nondepressed patients to endorse emphasizing the positive; more likely to focus on worst moments, recent flare-ups, their spinal condition, and the future; and more likely to compare themselves to high aspirations (eg, perfect health). Over time, among those who were depressed before surgery, those who improved focused decreasingly on worst moments and on the time before their spinal condition, and increasingly on emphasizing the positive and balancing the positives/negatives. Appraisal explained more variance in mental health among those who did not improve as compared to those who did, at all timepoints. All appraisal items were more highly correlated with mental health among those who remained depressed as compared to those who improved, particularly over time.

CONCLUSIONS

Endorsement of cognitive appraisal processes was different for depressed versus nondepressed spine-surgery patients before surgery and distinguished those who were depressed before surgery and improved versus those who did not improve. These findings suggest that targeted interventions could be beneficial for addressing mental health concerns during the spine surgery recovery trajectory. These interventions might use appraisal measures to identify patients likely to remain depressed after surgery, and then focus on helping these patients shift their focus and standards of comparison.

Computed tomography radiomics in predicting patient satisfaction after robotic-assisted total knee arthroplasty

PURPOSE

After robotic-assisted total knee arthroplasty (RA-TKA) surgery, some patients still experience joint discomfort. We aimed to establish an effective machine learning model that integrates radiomic features extracted from computed tomography (CT) scans and relevant clinical information to predict patient satisfaction three months postoperatively following RA-TKA.

MATERIALS AND METHODS

After careful selection, data from 142 patients were randomly divided into a training set (n = 99) and a test set (n = 43), approximately in a 7:3 ratio. A total of 1329 radiomic features were extracted from the regions of interest delineated in CT scans. The features were standardized using normalization algorithms, and the least absolute shrinkage and selection operator regression model was employed to select radiomic features with ICC > 0.75 and P < 0.05, generating the Rad-score as feature markers. Univariate and multivariate logistic regression was then used to screen clinical information (age, body mass index, operation time, gender, surgical side, comorbidities, preoperative KSS score, preoperative range of motion (ROM), preoperative and postoperative HKA angle, preoperative and postoperative VAS score) as potential predictive factors. The satisfaction scale ≥ 20 indicates patient satisfaction. Finally, three prediction models were established, focusing on radiomic features, clinical features, and their fusion. Model performance was evaluated using Receiver Operating Characteristic curves and decision curve analysis.

RESULTS

In the training set, the area under the curve (AUC) of the clinical model was 0.793 (95% CI 0.681-0.906), the radiomic model was 0.854 (95% CI 0.743-0.964), and the combined radiomic-clinical model was 0.899 (95% CI 0.804-0.995). In the test set, the AUC of the clinical model was 0.908 (95% CI 0.814-1.000), the radiomic model was 0.709 (95% CI 0.541-0.878), and the combined radiomic-clinical model was 0.928 (95% CI 0.842-1.000). The AUC of the radiomic-clinical model was significantly higher than the other two models. The decision curve analysis indicated its clinical application value.

CONCLUSION

We developed a radiomic-based nomogram model using CT imaging to predict the satisfaction of RA-TKA patients at 3 months postoperatively. This model integrated clinical and radiomic features and demonstrated good predictive performance and excellent clinical application potential.

Patient-Reported Outcomes of Total Hip Arthroplasty at an Ambulatory Surgery Center Versus a Hospital-Based Center

INTRODUCTION

The utilization of ambulatory surgery centers (ASCs) and same-day discharge (SDD) from hospital-based centers (HBCs) after total hip arthroplasty (THA) continues to increase. There remains a paucity of literature directly comparing patient-reported outcomes by surgery site. We sought to compare outcomes between patients undergoing THA at an ASC versus HBC while controlling for medical comorbidities.

METHODS

Patients undergoing primary THA with SDD (postoperative day 0) from a single HBC (1,015 patients) or stand-alone ASC (170 patients) from December 2020 to 2021 were identified. Patient demographics, comorbidities, and 90-day complications were collected. Hip Osteoarthritis Outcome Score (HOOS JR), VR-12, and procedural satisfaction scores were collected preoperatively and at 3, 6, and 12 months. Patients were matched by age and American Society of Anesthesiologists (ASA). Chi-squared analysis was conducted to compare categorical variables, and a Wilcoxon rank-sum test was used for continuous variables. Linear regression models were conducted considering age, sex, and presence of comorbidities.

RESULTS

Patients undergoing THA at an ASC had markedly higher VR-12 Physical Component Scores at all time points and improved VR-12 Mental Component Scores at preoperative visit and 6 months. These patients had increased procedural satisfaction at 3 months, although there was no difference at 1 year. No notable difference was observed in 90-day complication rates between groups. After matching by age and ASA, each group had 170 patients. In the matched analysis, preoperative HOOS JR scores were markedly lower in the HBC group. However, there was no notable difference in HOOS JR scores, change in HOOS JR scores, and procedural satisfaction, at any postoperative time point.

CONCLUSIONS

No notable difference was observed in patient-reported outcomes at any time point for SDD after THA performed at an ASC or an HBC when controlling for age and comorbidities. This study suggests noninferiority of stand-alone ASCs for outpatient THA, regarding patient satisfaction and patient-reported outcomes.

The influence of body mass index on patient-reported outcome measures following total hip arthroplasty: a retrospective study of 3,903 Cases

BACKGROUND

The influence of obesity on patient-reported outcome measures (PROMs) following total hip arthroplasty (THA) is currently controversial. This study aimed to compare PROM scores for pain, functional status, and global physical/mental health based on body mass index (BMI) classification.

METHODS

Primary, elective THA procedures at a single institution between 2018 and 2021 were retrospectively reviewed, and patients were stratified into four groups based on BMI: normal weight (18.5-24.99 kg/m2), overweight (25-29.99 kg/m2), obese (30-39.99 kg/m2), and morbidly obese (> 40 kg/m2). Patient-Reported Outcome Measurement Information System (PROMIS) and Hip Disability and Osteoarthritis Outcome Score for Joint Replacement (HOOS, JR) scores were collected. Preoperative, postoperative, and pre/post- changes (pre/post-Δ) in scores were compared between groups. Multiple linear regression was used to assess for confounders.

RESULTS

We analyzed 3,404 patients undergoing 3,903 THAs, including 919 (23.5%) normal weight, 1,374 (35.2%) overweight, 1,356 (35.2%) obese, and 254 (6.5%) morbidly obese cases. HOOS, JR scores were worse preoperatively and postoperatively for higher BMI classes, however HOOS, JR pre/post-Δ was comparable between groups. All PROMIS measures were worse preoperatively and postoperatively in higher BMI classes, though pre/post-Δ were comparable for all groups. Clinically significant improvements for all BMI classes were observed in all PROM metrics except PROMIS mental health. Regression analysis demonstrated that obesity, but not morbid obesity, was independently associated with greater improvement in HOOS, JR.

CONCLUSIONS

Obese patients undergoing THA achieve lower absolute scores for pain, function, and self-perceived health, despite achieving comparable relative improvements in pain and function with surgery. Denying THA based on BMI restricts patients from clinically beneficial improvements comparable to those of non-obese patients, though morbidly obese patients may benefit from additional weight loss to achieve maximal functional improvement.

Discordance between self-report and performance-based outcomes: Contribution of psychosocial factors

The purpose of this study was to examine the role of psychosocial factors in the discordance between perceived and observed physical disability in patients with osteoarthritis of the hip or knee joint. This was a cross-sectional study of patients seen for consideration of joint arthroplasty surgery. Patients completed a psychosocial outcome measure, a patient self-reported functional scale, and two performance-based tests. Data of 121 patients, mean age, 67 (8), 81 (67%) females were used for analysis. The fear avoidance and positive affect domains had the strongest association with the discordance between the self-report and both performance outcome measures. Age, gender, and severity of osteoarthritis were associated with discordance in relation to walking. Fear avoidance beliefs and positive affect play important roles in perception of pain and function. Age, gender, and severity of arthritis should be taken into consideration for a more holistic approach to arthritis care.

Prediction of 30-Day Mortality Following Revision Total Hip and Knee Arthroplasty: Machine Learning Algorithms Outperform CARDE-B, 5-Item, and 6-Item Modified Frailty Index Risk Scores

BACKGROUND

Although risk calculators are used to prognosticate postoperative outcomes following revision total hip and knee arthroplasty (total joint arthroplasty [TJA]), machine learning (ML) based predictive tools have emerged as a promising alternative for improved risk stratification. This study aimed to compare the predictive ability of ML models for 30-day mortality following revision TJA to that of traditional risk-assessment indices such as the CARDE-B score (congestive heart failure, albumin (< 3.5 mg/dL), renal failure on dialysis, dependence for daily living, elderly (> 65 years of age), and body mass index (BMI) of < 25 kg/m2), 5-item modified frailty index (5MFI), and 6MFI.

METHODS

Adult patients undergoing revision TJA between 2013 and 2020 were selected from the American College of Surgeons National Surgical Quality Improvement Program database and randomly split 80:20 to compose the training and validation cohorts. There were 3 ML models - extreme gradient boosting, random forest, and elastic-net penalized logistic regression (NEPLR) - that were developed and evaluated using discrimination, calibration metrics, and accuracy. The discrimination of CARDE-B, 5MFI, and 6MFI scores was assessed individually and compared to that of ML models.

RESULTS

All models were equally accurate (Brier score = 0.005) and demonstrated outstanding discrimination with similar areas under the receiver operating characteristic curve (AUCs, extreme gradient boosting = 0.94, random forest = NEPLR = 0.93). The NEPLR was the best-calibrated model overall (slope = 0.54, intercept = -0.004). The CARDE-B had the highest discrimination among the scores (AUC = 0.89), followed by 6MFI (AUC = 0.80), and 5MFI (AUC = 0.68). Albumin < 3.5 mg/dL and BMI (< 30.15) were the most important predictors of 30-day mortality following revision TJA.

CONCLUSIONS

The ML models outperform traditional risk-assessment indices in predicting postoperative 30-day mortality after revision TJA. Our findings highlight the utility of ML for risk stratification in a clinical setting. The identification of hypoalbuminemia and BMI as prognostic markers may allow patient-specific perioperative optimization strategies to improve outcomes following revision TJA.

Bibliometric analysis of machine learning trends and hotspots in arthroplasty literature over 31 years

Background

Artificial intelligence has demonstrated utility in orthopedic research. Algorithmic models derived from machine learning have demonstrated adaptive learning with predictive application towards outcomes, leading to increased traction in the literature. This study aims to identify machine learning arthroplasty research trends and anticipate emerging key terms.

Methods

Published literature focused on machine learning in arthroplasty from 1992 to 2023 was selected through the Web of Science Core Collection of Clarivate Analytics. Following that, bibliometric indicators were attained and brought in to perform an additional examination using Bibliometrix and VOSviewer to identify historical and present patterns within the literature.

Results

A total of 235 documents were obtained through bibliometric sourcing based on machine learning applications within the arthroplasty literature. Thirty-four countries published articles on the topic, and the United States was demonstrated to be the largest global contributor. Four hundred-five institutions internationally contributed articles, with Harvard Medical School and the University of California system as the most relevant institutes, with 75 and 44 articles produced, respectively. Kwon YM was the most productive author, while Haeberle HS and Ramkumar PN were the most impactful based on h-index. The Thematic map and Co-occurrence visualization helped identify both major and niche themes present in the scientific databases.

Conclusions

Machine learning in arthroplasty research continues to gain traction with a growing annual production rate and contributions from international authors and institutions. Institutions and authors based in the United States are the leading contributors to machine learning applications within arthroplasty research. This research discerns trends that have occurred, are presently ongoing, and are emerging within this field, aiming to inform future hotspot development.

Using Unsupervised Machine Learning to Predict Quality of Life After Total Knee Arthroplasty

BACKGROUND

Patient-reported outcome measures (PROMs) are an important metric to assess total knee arthroplasty (TKA) patients. The purpose of this study was to use a machine learning (ML) algorithm to identify patient features that impact PROMs after TKA.

METHODS

Data from 636 TKA patients enrolled in our patient database between 2018 and 2022, were retrospectively reviewed. Their mean age was 68 years (range, 39 to 92), 56.7% women, and mean body mass index of 31.17 (range, 16 to 58). Patient demographics and the Functional Comorbidity Index were collected alongside Patient-Reported Outcome Measures Information System Global Health v1.2 (PROMIS GH-P) physical component scores preoperatively, at 3 months, and 1 year after TKA. An unsupervised ML algorithm (spectral clustering) was used to identify patient features impacting PROMIS GH-P scores at the various time points.

RESULTS

The algorithm identified 5 patient clusters that varied by demographics, comorbidities, and pain scores. Each cluster was associated with predictable trends in PROMIS GH-P scores across the time points. Notably, patients who had the worst preoperative PROMIS GH-P scores (cluster 5) had the most improvement after TKA, whereas patients who had higher global health rating preoperatively had more modest improvement (clusters 1, 2, and 3). Two out of Five patient clusters (cluster 4 and 5) showed improvement in PROMIS GH-P scores that met a minimally clinically important difference at 1-year postoperative.

CONCLUSIONS

The unsupervised ML algorithm identified patient clusters that had predictable changes in PROMs after TKA. It is a positive step toward providing precision medical care for each of our arthroplasty patients.

Predicting Functional Outcomes of Total Hip Arthroplasty Using Machine Learning: A Systematic Review

The aim of this review was to assess the reliability of machine learning (ML) techniques to predict the functional outcome of total hip arthroplasty. The literature search was performed up to October 2023, using MEDLINE/PubMed, Embase, Web of Science, and NIH Clinical Trials. Level I to IV evidence was included. Seven studies were identified that included 44,121 patients. The time to follow-up varied from 3 months to more than 2 years. Each study employed one to six ML techniques. The best-performing models were for health-related quality of life (HRQoL) outcomes, with an area under the curve (AUC) of more than 84%. In contrast, predicting the outcome of hip-specific measures was less reliable, with an AUC of between 71% to 87%. Random forest and neural networks were generally the best-performing models. Three studies compared the reliability of ML with traditional regression analysis: one found in favour of ML, one was not clear and stated regression closely followed the best-performing ML model, and one showed a similar AUC for HRQoL outcomes but did show a greater reliability for ML to predict a clinically significant change in the hip-specific function. ML offers acceptable-to-excellent discrimination of predicting functional outcomes and may have a marginal advantage over traditional regression analysis, especially in relation to hip-specific hip functional outcomes.

Association of affected and non-affected side ability with postoperative outcomes in patients undergoing total hip arthroplasty

BACKGROUND

Although several reports have examined the association between preoperative function and postoperative outcomes in patients undergoing total hip arthroplasty (THA), it is unclear whether the ability of the affected or non-affected side particularly impacts on outcomes. We aimed to investigate the association between affected and non-affected side ability and walking independence.

METHODS

We prospectively enrolled 721 consecutive patients who underwent THA. Preoperatively, quadriceps isometric strength (QIS) and one-leg standing time (OLST) were measured. The endpoints were walking independence within 3, 5, 7, 10, and 14 days postoperatively. The associations between preoperative abilities and outcomes were examined using multivariate Cox hazard model, and the area under the curves (AUCs) for outcomes were compared.

RESULTS

We analysed 540 patients after excluding patients who met the exclusion criteria. Both affected and non-affected QIS predicted walking independence within 3 (p = 0.006 and 0.001, respectively), 5, 7, 10, and 14 (both p < 0.001) days postoperatively. For OLST, only the affected side did not predict walking independence within 3 days postoperatively (p = 0.154 and 0.012, respectively), and both sides did at days 5 (p = 0.019 and <0.001, respectively), 7, 10, and 14 (both p < 0.001). The AUCs of the non-affected side ability for walking independence were significantly greater than those of the affected side on postoperative days 3 (0.66 vs. 0.73; p = 0.021) and 5 (0.67 vs. 0.71; p = 0.040), with no significant difference after day 7.

CONCLUSIONS

Both sides abilities were associated with walking independence after THA, but non-affected side was found to be particularly crucial for early walking independence.

Differences in Outcomes Between Initial Responders and Subsequent Responders to Health Questionnaires for Total Hip and Knee Arthroplasty: An Australian Tertiary Institutional Registry Study

BACKGROUND

Patient-reported outcome measure (PROM) questionnaires in national arthroplasty registries often have low response rates leading to questions about data reliability. In Australia, the SMART (St. Vincent's Melbourne Arthroplasty Outcomes) registry captures all elective total hip (THA) and total knee (TKA) arthroplasty patients with an approximate 98% response rate for preoperative and 12-month PROM scores. This high response rate is due to dedicated registry staff following up patients who do not initially respond (subsequent responders). This study compared initial responders to subsequent responders to find differences in 12-month PROM outcomes for THA and TKA.

METHODS

All elective THA and TKA patients for osteoarthritis from 2012 to 2021 captured by the SMART registry were included. In total, 1,333 THA and 1,340 TKA patients were included. The PROM scores were assessed using the Veterans-RAND 12 (VR12) and Western Ontario and McMasters Universities Arthritis Index (WOMAC) questionnaires. The primary outcome was differences in mean 12-month PROM scores between initial and subsequent responders.

RESULTS

Baseline characteristics and PROM scores were similar between initial and subsequent responders. However, 12-month PROM scores varied significantly. The adjusted mean difference showed that for the WOMAC pain score, subsequent responders scored 3.4 points higher in the THA cohort and 7.4 points higher in the TKA cohort compared to initial responders. Significant differences were also found in other WOMAC and VR12 scores for both THA and TKA cohorts at the 12-month timepoint.

CONCLUSION

This study found that significant differences in PROM outcomes postsurgery occurred in THA and TKA patients based on response to PROM questionnaires, suggesting that loss to follow-up in PROM outcomes should not be treated as missing completely at random (MCAR).

Machine-learning vs. logistic regression for preoperative prediction of medical morbidity after fast-track hip and knee arthroplasty-a comparative study

BACKGROUND

Machine-learning models may improve prediction of length of stay (LOS) and morbidity after surgery. However, few studies include fast-track programs, and most rely on administrative coding with limited follow-up and information on perioperative care. This study investigates potential benefits of a machine-learning model for prediction of postoperative morbidity in fast-track total hip (THA) and knee arthroplasty (TKA).

METHODS

Cohort study in consecutive unselected primary THA/TKA between 2014-2017 from seven Danish centers with established fast-track protocols. Preoperative comorbidity and prescribed medication were recorded prospectively and information on length of stay and readmissions was obtained through the Danish National Patient Registry and medical records. We used a machine-learning model (Boosted Decision Trees) based on boosted decision trees with 33 preoperative variables for predicting "medical" morbidity leading to LOS > 4 days or 90-days readmissions and compared to a logistical regression model based on the same variables. We also evaluated two parsimonious models, using the ten most important variables in the full machine-learning and logistic regression models. Data collected between 2014-2016 (n:18,013) was used for model training and data from 2017 (n:3913) was used for testing. Model performances were analyzed using precision, area under receiver operating (AUROC) and precision recall curves (AUPRC), as well as the Mathews Correlation Coefficient. Variable importance was analyzed using Shapley Additive Explanations values.

RESULTS

Using a threshold of 20% "risk-patients" (n:782), precision, AUROC and AUPRC were 13.6%, 76.3% and 15.5% vs. 12.4%, 74.7% and 15.6% for the machine-learning and logistic regression model, respectively. The parsimonious machine-learning model performed better than the full logistic regression model. Of the top ten variables, eight were shared between the machine-learning and logistic regression models, but with a considerable age-related variation in importance of specific types of medication.

CONCLUSION

A machine-learning model using preoperative characteristics and prescriptions slightly improved identification of patients in high-risk of "medical" complications after fast-track THA and TKA compared to a logistic regression model. Such algorithms could help find a manageable population of patients who may benefit most from intensified perioperative care.

The Impact of Machine Learning on Total Joint Arthroplasty Patient Outcomes: A Systemic Review

BACKGROUND

Supervised machine learning techniques have been increasingly applied to predict patient outcomes after hip and knee arthroplasty procedures. The purpose of this study was to systematically review the applications of supervised machine learning techniques to predict patient outcomes after primary total hip and knee arthroplasty.

METHODS

A comprehensive literature search using the electronic databases MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials, and Cochrane Database of Systematic Reviews was conducted in July of 2021. The inclusion criteria were studies that utilized supervised machine learning techniques to predict patient outcomes after primary total hip or knee arthroplasty.

RESULTS

Search criteria yielded n = 30 relevant studies. Topics of study included patient complications (n = 6), readmissions (n = 1), revision (n = 2), patient-reported outcome measures (n = 4), patient satisfaction (n = 4), inpatient status and length of stay (LOS) (n = 9), opioid usage (n = 3), and patient function (n = 1). Studies involved TKA (n = 12), THA (n = 11), or a combination (n = 7). Less than 35% of predictive outcomes had an area under the receiver operating characteristic curve (AUC) in the excellent or outstanding range. Additionally, only 9 of the studies found improvement over logistic regression, and only 9 studies were externally validated.

CONCLUSION

Supervised machine learning algorithms are powerful tools that have been increasingly applied to predict patient outcomes after total hip and knee arthroplasty. However, these algorithms should be evaluated in the context of prognostic accuracy, comparison to traditional statistical techniques for outcome prediction, and application to populations outside the training set. While machine learning algorithms have been received with considerable interest, they should be critically assessed and validated prior to clinical adoption.

Depression trajectories during the COVID-19 pandemic: a secondary analysis of the impact of cognitive-appraisal processes

PURPOSE

This study characterized depression trajectories during the COVID pandemic and investigated how appraisal and changes in appraisal over time related to these depression trajectories.

METHODS

This longitudinal study of the psychosocial impact of the COVID-19 pandemic included 771 people with data at three timepoints over 15.5 months. The depression index was validated using item-response-theory methods and receiver-operating-characteristic curve analysis. The Quality of Life (QOL) Appraisal Profile_v2 Short-Form assessed cognitive-appraisal processes. Sequence analysis characterized depression-trajectory groups, and random effects models examined appraisal main effects, appraisal-by-group, and appraisal-by-group-by-time interactions.

RESULTS

Sequence analysis generated six trajectory groups: Stably Well (n = 241), Stably Depressed (n = 299), Worsening (n = 79), Improving (n = 83), Fluctuating Pattern 1 (No-Yes-No; n = 41), and Fluctuating Pattern 2 (Yes-No-Yes; n = 28). While all groups engaged in negative appraisal processes when they were depressed, the Stably Depressed group consistently focused on negative aspects of their life. Response-shift effects were revealed such that there were differences in the appraisal-depression relationship over time for standards of comparison and recent changes for the Stably Depressed, and in health goals for those Getting Better.

CONCLUSION

The present work is, to our knowledge, the first study of response-shift effects in depression. During these first 15.5 pandemic months, group differences highlighted the connection between negative appraisals and depression, and response-shift effects in these relationships over time. Egregious life circumstances may play a lesser role for the Stably Depressed but a greater role for people who have transient periods of depression as well as for those with improving trajectories (i.e., endogenous vs. reactive depression). How one thinks about QOL is intrinsically linked to mental health, with clear clinical implications.

Assessing the predictive capacity of machine learning models using patient-specific variables in determining in-hospital outcomes after THA

Background

Machine learning is a subset of artificial intelligence using algorithmic modeling to progressively learn and create predictive models. Clinical application of machine learning can aid physicians through identification of risk factors and implications of predicted patient outcomes.

Aims

The aim of this study was to compare patient-specific and situation perioperative variables through optimized machine learning models to predict postoperative outcomes.

Methods

Data from 2016 to 2017 from the National Inpatient Sample was used to identify 177,442 discharges undergoing primary total hip arthroplasty, which were included in the training, testing, and validation of 10 machine learning models. 15 predictive variables consisting of 8 patient-specific and 7 situational specific variables were utilized to predict 3 outcome variables: length of stay, discharge, and mortality. The machine learning models were assessed in responsiveness via area under the curve and reliability.

Results

For all outcomes, Linear Support Vector Machine had the highest responsiveness among all models when using all variables. When utilizing patient-specific variables only, responsiveness of the top 3 models ranged between 0.639 and 0.717 for length of stay, 0.703-0.786 for discharge disposition, and 0.887-0.952 for mortality. The top 3 models utilizing situational variables only produced responsiveness of 0.552-0.589 for length of stay, 0.543-0.574 for discharge disposition, and 0.469-0.536 for mortality.

Conclusions

Linear Support Vector Machine was the most responsive machine learning model of the 10 algorithms trained, while decision list was most reliable. Responsiveness was observed to be consistently higher with patient-specific variables than situational variables, emphasizing the predictive capacity and value of patient-specific variables. The current practice in machine learning literature generally deploys a single model, it is suboptimal to develop optimized models for application into clinical practice. The limitation of other algorithms may prohibit potential more reliable and responsive models.Level of Evidence III.

Osteoarthritis year in review 2022: epidemiology & therapy

This 'Year in Review' provides a synopsis of key research themes and individual studies from the clinical osteoarthritis (OA) field, focused on epidemiology and therapy. The electronic database search for the review was adapted from the 2021 year in review search, to increase search specificity for relevant study designs, and was conducted in Medline, Embase and medRxiv (31^st March 2021 to 4^th March 2022). Following screening for eligibility, studies were grouped according to their key research design, including reviews, cohorts and randomised trials. 11 key themes emerged, including the importance of several comorbidities in predicting OA incidence and prevalence, surgical approaches that can reduce the risk of post-traumatic OA, the heterogenous but nevertheless relatively stable nature of OA subgroup trajectories, the paucity of robust studies particularly of surgery for OA and the very modest benefit of many therapies under evaluation in trials. A particular interest of the authors was to consider whether new studies are helping determine how to better ensure the right patient with OA is matched to the right treatment at the right time. There are several new studies developing improved predictive models through big data analytics and machine learning which show promise, need validation, and may support new approaches to stratified care.

Artificial intelligence and machine learning as a viable solution for hip implant failure diagnosis-Review of literature and in vitro case study

The digital health industry is experiencing fast-paced research which can provide digital care programs and technologies to enhance the competence of healthcare delivery. Orthopedic literature also confirms the applicability of artificial intelligence (AI) and machine learning (ML) models to medical diagnosis and clinical decision-making. However, implant monitoring after primary surgery often happens with a wellness visit or when a patient complains about it. Neglecting implant design and other technical errors in this scenario, unmonitored circumstances, and lack of post-surgery monitoring may ultimately lead to the implant system's failure and leave us with the only option of high-risk revision surgery. Preventive maintenance seems to be a good choice to identify the onset of an irreversible prosthesis failure. Considering all these aspects for hip implant monitoring, this paper explores existing studies linking ML models and intelligent systems for hip implant diagnosis. This paper explores the feasibility of an alternative continuous monitoring technique for post-surgery implant monitoring backed by an in vitro ML case study. Tribocorrosion and acoustic emission (AE) data are considered based on their efficacy in determining irreversible alteration of implant material to prevent total failures. This study also facilitates the relevance of developing an artificially intelligent implant monitoring methodology that can function with daily patient activities and how it can influence the digital orthopedic diagnosis. AI-based non-invasive hip implant monitoring system enabling point-of-care testing.

Are contextual factors associated with activities and participation after total hip arthroplasty? A systematic review

OBJECTIVES

After total hip arthroplasty (THA), over 30% of individuals report activity limitations and participation restrictions. This systematic review aimed to determine the association between contextual factors and outcomes in the activity and participation domain after THA for hip osteoarthritis (OA).

METHODS

This systematic review was developed according to the PRISMA guidelines for systematic reviews. PubMed, Web of Science, Embase and Scopus were searched until August 2022. Risk of bias was assessed with the Quality in Prognosis Studies tool (QUIPS).

RESULTS

Twenty-nine articles were included. Eighteen had a high risk of bias, 3 had a low risk of bias, and 8 had a moderate risk of bias. Anxiety was only investigated in studies with high risk of bias but showed a consistent negative association with activities and participation after THA across multiple studies. Evidence was inconsistent regarding the associations between depression, trait anxiety, sense of coherence, big 5 personality traits, educational level, marital status, employment status, job position, expectations and social support, and the activity and participation domain. Optimism, general self-efficacy, cognitive appraisal processes, illness perception, ethnicity, and positive life events were associated with activities and participation but were only investigated in 1 study. No associations were identified across multiple studies for living or smoking status. Control beliefs, kinesiophobia, race, discharge location, level of poverty in neighbourhood, negative life events and occupational factors, were not associated with the activity and participation domain but were only investigated in 1 study.

CONCLUSION

Methodological quality of the included studies was low. Anxiety was the only factor consistently associated with worse outcomes in the activity and participation domain after THA but was only investigated in studies with high risk of bias. Further research is needed to confirm relationships between other contextual factors and activities and participation after THA.

REGISTRATION

PROSPERO CRD42020199070.

Epidemiology of total hip arthroplasty: demographics, comorbidities and outcomes

BACKGROUND

Primary THA (THA) is a successful procedure for end-stage hip osteoarthritis. In the setting of a failed THA, revision total hip arthroplasty (rTHA) acts as a salvage procedure. This procedure has increased risks, including sepsis, infection, prolonged surgery time, blood loss, and increased length of stay. Increasing focus on understanding of demographics, comorbidities, and inpatient outcomes can lead to better perioperative optimization and post-operative outcomes. This epidemiological registry study aimed to compare the demographics, comorbidity profiles, and outcomes of patients undergoing THA and rTHA.

METHODS

A retrospective review of discharge data reported from 2006 to the third quarter of 2015 using the National Inpatient Sample registry was performed. The study included adult patients aged 40 and older who underwent either THA or rTHA. A total of 2,838,742 THA patients and 400,974 rTHA patients were identified.

RESULTS

The primary reimbursement for both THA and rTHA was dispensed by Medicare at 53.51% and 65.36% of cases respectively. Complications arose in 27.32% of THA and 39.46% of rTHA cases. Postoperative anemia was the most common complication in groups (25.20% and 35.69%). Common comorbidities in both groups were hypertension and chronic pulmonary disease. rTHA indications included dislocation/instability (21.85%) followed by mechanical loosening (19.74%), other mechanical complications (17.38%), and infection (15.10%).

CONCLUSION

Our data demonstrated a 69.50% increase in patients receiving THA and a 28.50% increase in rTHA from the years 2006 to 2014. The data demonstrated 27.32% and 39.46% complication rate with THA and rTHA, with postoperative anemia as the most common cause. Common comorbidities were hypertension and chronic pulmonary disease. Future analyses into preoperative optimizations, such as prior consultation with medical specialists or improved primary hip protocol, should be considered to prevent/reduce postoperative complications amongst a progressive expansion in patients receiving both THA and rTHA.

Applying machine learning methods to predict the hospital re-admission within 30 days of total hip arthroplasty and hemiarthroplasty

BACKGROUND

Total hip arthroplasty (THA) and hemiarthroplasty are common treatments for severe hip joint disease. To predict the probability of re-admission after discharge when patients are hospitalized will support providing appropriate health education and guidance.

METHODS

The research aims to use logistic regression (LR), decision trees (DT), random forests (RF), and artificial neural networks (ANN) to establish predictive models and compare their performances on re-admissions within 30 days after THA or hemiarthroplasty. The data of this study includes patient demographics, physiological measurements, disease history, and clinical laboratory test results.

RESULTS

There were 508 and 309 patients in the THA and hemiarthroplasty studies respectively from September 2016 to December 2018. The accuracies of the four models LR, DT, RF, and ANN in the THA experiment are 94.3%, 93.2%, 97.3%, and 93.9%, respectively. In the hemiarthroplasty experiment, the accuracies of the four models are 92.4%, 86.1%, 94.2%, and 94.8%, respectively. Among these, we found that the RF model has the best sensitivity and ANN model has the best area under the receiver operating characteristic (AUROC) score in both experiments.

CONCLUSIONS

The THA experiment confirmed that the performance of the RF model is better than the other models. The key factors affecting the prognosis after THA surgery are creatinine, sodium, anesthesia duration, and dialysis. In the hemiarthroplasty experiment, the ANN model showed more accurate results. Poor kidney function increases the risk of hospital re-admission. This research highlights that RF and ANN model perform well on the hip replacement surgery outcome prediction.

Concerns surrounding application of artificial intelligence in hip and knee arthroplasty

Literature surrounding artificial intelligence (AI)-related applications for hip and knee arthroplasty has proliferated. However, meaningful advances that fundamentally transform the practice and delivery of joint arthroplasty are yet to be realized, despite the broad range of applications as we continue to search for meaningful and appropriate use of AI. AI literature in hip and knee arthroplasty between 2018 and 2021 regarding image-based analyses, value-based care, remote patient monitoring, and augmented reality was reviewed. Concerns surrounding meaningful use and appropriate methodological approaches of AI in joint arthroplasty research are summarized. Of the 233 AI-related orthopaedics articles published, 178 (76%) constituted original research, while the rest consisted of editorials or reviews. A total of 52% of original AI-related research concerns hip and knee arthroplasty (n = 92), and a narrative review is described. Three studies were externally validated. Pitfalls surrounding present-day research include conflating vernacular (“AI/machine learning”), repackaging limited registry data, prematurely releasing internally validated prediction models, appraising model architecture instead of inputted data, withholding code, and evaluating studies using antiquated regression-based guidelines. While AI has been applied to a variety of hip and knee arthroplasty applications with limited clinical impact, the future remains promising if the question is meaningful, the methodology is rigorous and transparent, the data are rich, and the model is externally validated. Simple checkpoints for meaningful AI adoption include ensuring applications focus on: administrative support over clinical evaluation and management; necessity of the advanced model; and the novelty of the question being answered.

Cite this article: Bone Joint J 2022;104-B(12):1292–1303.

Comparative Analysis of the Ability of Machine Learning Models in Predicting In-hospital Postoperative Outcomes After Total Hip Arthroplasty

BACKGROUND

Machine learning (ML) methods have shown promise in a wide range of applications including the development of patient-specific predictive models before surgical interventions. The purpose of this study was to develop, test, and compare four distinct ML models to predict postoperative parameters after primary total hip arthroplasty.

METHODS

Data from the Nationwide Inpatient Sample were used to identify patients undergoing total hip arthroplasty from 2016 to 2017. Linear support vector machine (LSVM), random forest (RF), neural network (NN), and extreme gradient boost trees (XGBoost) predictive of mortality, length of stay, and discharge disposition were developed and validated using 15 predictive patient-specific and hospital-specific factors. Area under the curve of the receiver operating characteristic (AUCROC) curve and accuracy were used as validity metrics, and the strongest predictive variables under each model were assessed.

RESULTS

A total of 177,442 patients were included in this analysis. For mortality, the XGBoost, NN, and LSVM models all had excellent responsiveness during validation while RF had fair responsiveness. LSVM had the highest responsiveness with an AUCROC of 0.973 during validation. For the length of stay, the LSVM and NN models had fair responsiveness while the XGBoost and random forest models had poor responsiveness. LSVM had the highest responsiveness with an AUCROC of 0.744 during validation. For the discharge disposition outcome, LSVM had good responsiveness while the XGBoost, NN, and RF models all had fair responsiveness. LSVM had the highest responsiveness with an AUCROC of 0.801.

DISCUSSION

The ML methods tested demonstrated a range of poor-to-excellent responsiveness and accuracy in the prediction of the assessed metrics, with LSVM being the best performer. Such models should be further developed, with eventual integration into clinical practice to inform patient discussions and management decision making, with the potential for integration into tiered bundled payment models.

Cognitive Processes during Recovery: Moving toward Personalized Spine Surgery Outcomes

This paper focuses on a novel application of personalized medicine: the ways one thinks about health (i.e., appraisal processes) as relevant predictors of spine-surgery response. This prospective longitudinal cohort study (n = 235) investigated how appraisal processes relate to outcomes of spinal decompression and/or fusion surgery, from pre-surgery through one-year post-surgery. Patient-reported outcomes assessed spine-specific disability (Oswestry Disability Index (ODI)), mental health functioning (Rand-36 Mental Component Score (MCS)), and cognitive appraisal processes (how people recall past experiences and to whom they compare themselves). Analysis of Variance examined the appraisal-outcomes association in separate models at pre-surgery, 3 months, and 12 months. We found that appraisal processes explained less variance at pre-surgery than later and were differentially relevant to health outcomes at different times in the spine-surgery recovery trajectory. For the ODI, recall of the seriousness of their condition was most prominent early in recovery, and comparing themselves to positive standards was most prominent later. For the MCS, not focusing on the negative aspects of their condition and/or on how others see them was associated with steady improvement and higher scores at 12 months. Appraisal processes are relevant to both spine-specific disability and mental-health functioning. Such processes are modifiable objects of attention for personalizing spine-surgery outcomes.

Patient reported outcomes measures (PROMs) trajectories after elective hip arthroplasty: a latent class and growth mixture analysis

Background

Patient-reported outcome measures (PROMs) are an extensively used tool to assess and improve the quality of healthcare services. PROMs can be related to individual demographic and clinical characteristics in patients undergoing hip arthroplasty (HA). The aim of this study is to identify distinct subgroups of patients with unique trajectories of PROMS scores and to determine patients’ features associated with these subgroups.

Methods

We conducted a prospective, cohort study in which PROMs questionnaires (Euro Quality 5 Dimensions 3L, EQ-5D-3L, Euro-Quality-Visual-Analytic-Score, EQ-VAS, Hip disability and Osteoarthritis Outcome Score, HOOS-PS) were administered to patients undergoing elective HA pre-operatively, and at 6 and 12 months after surgery. For each measure, latent class growth analysis and growth mixture models were used to identify subgroups of patients with distinct trajectories of scores. Demografic and clinical predictors of the latent classes in growth mixture model were identified using a 3-step approach.

Results

We found three distinct trajectories for each PROM score. These trajectories indicated a response heterogeneity to the HA among the patients (n = 991). Patient’s gender, ASA score, and obesity were significantly associated with different PROMs trajectories.

Conclusions

We identified three distinct trajectories for each of the three PROMs indicators. Several demographic and clinical characteristics are associated with the different trajectories of PROMs at 6 and 12 months after HA and could be used to identify groups of patients with different outcomes following HA surgery. These findings underline the importance of patient-centered care, supporting the usefulness of integrating PROMs data alongside routinely collected healthcare records for guiding clinical care and maximizing patients’ positive outcomes.

Trial registration: Protocol version (1.0) and trial registration data are available on the platform www.clinicaltrial.gov with the identifier NCT03790267, posted on December 31, 2018.

Supplementary Information

The online version contains supplementary material available at 10.1186/s41687-022-00503-5.

Endoscopic Surgical Treatment of Osteoarthritis and Prognostic Model Construction

Purpose

Osteoarthritis (OA) is a degenerative disease of joints. Currently, there is still a lack of effective tools to predict the long-term efficacy of surgical treatment of OA. The purpose of this study was to explore the prognostic factors of endoscopic surgery for OA and to predict the long-term efficacy of this type of surgery for OA by establishing a prognostic model.

Methods

Baseline and follow-up data on 236 OA patients who underwent surgery in our hospital from January 2017 to December 2021 were selected and patients were randomly assigned to a training set (n = 165) and a test set (n = 71). The Pearson correlation coefficient was used to analyze the correlation between features. Feature selection was performed by recursive feature elimination (RFE) and linear regression. K-means clustering analysis was performed on the selected features to obtain the number of output layers. Finally, a single hidden layer error backpropagation (BP)-artificial neural network (ANN) model was established on the training set, and receiver operating characteristic (ROC) curve was drawn on the test set for verification.

Results

Correlation analysis revealed no redundancy among features. RFE and linear regression screened out the features associated with postoperative prognosis under endoscopic surgery: sex, age, BMI, region, morning stiffness time, step count, and osteophyte area. K-means clustering yielded that the optimal number of categories was three, the same as the number of categories for the outcome variable. Therefore, a 7-1-3 BP neural network model was established based on these 7 features, and this model could predict the postoperative situation within one year to a relatively accurate extent: area under curve values (AUC) were 0.814, 0.700, and 0.761 in patients with worse, unchanged, and improved conditions one year after surgery, respectively, higher than the multiclass AUC value (0.646).

Conclusion

The prognostic model of endoscopic surgery for OA constructed in this study can well predict the disease progression of patients within one year after surgery.

Appraisal and patient-reported outcomes following total hip arthroplasty: a longitudinal cohort study

Background

Total hip arthroplasty (THA) is a successful procedure that provides pain relief, restores function, and improves quality of life (QOL) for patients with advanced arthritis in their hip joint. To date, little research has examined the role of cognitive appraisal processes in THA outcomes. This study examined the role of cognitive appraisal processes in THA outcomes in the first year post-surgery.

Methods

This longitudinal cohort study collected data at pre-surgery, 6 weeks post-surgery, 3 months post-surgery, and 12 months post-surgery. Adults (n = 189) with a primary diagnosis of osteoarthritis were consecutively recruited from an active THA practice at a Canadian academic teaching hospital. Measures included the Hip Disability and Osteoarthritis Outcome Score (HOOS), the Mental Component Score (MCS) of the Rand-36, and the Brief Appraisal Inventory (BAI). Analysis of Variance examined the association between BAI items and the HOOS or MCS scores. Random effects models investigated appraisal main effects and appraisal-by-time interactions for selected BAI items.

Results

HOOS showed great improvement over the first 12 months after THA, and was mitigated by three appraisal processes in particular: focusing on problems with healthcare or living situation, and preparing one’s family for health changes. MCS was stable and low over time, and the following appraisal processes were implicated by very large effect sizes: not comparing themselves to healthier people, focusing on money problems, preparing their family for their health changes, or trying to shed responsibilities.

Conclusions

Appraisal processes are relevant to health outcomes after THA, with different processes coming into play at different points in the recovery trajectory.

Supplementary Information

The online version contains supplementary material available at 10.1186/s41687-022-00498-z.

The utilization of artificial neural networks for the prediction of 90-day unplanned readmissions following total knee arthroplasty

BACKGROUND

A reliable predictive tool to predict unplanned readmissions has the potential to lower readmission rates through targeted pre-operative counseling and intervention with respect to modifiable risk factors. This study aimed to develop and internally validate machine learning models for the prediction of 90-day unplanned readmissions following total knee arthroplasty.

METHODS

A total of 10,021 consecutive patients underwent total knee arthroplasty. Patient charts were manually reviewed to identify patient demographics and surgical variables that may be associated with 90-day unplanned hospital readmissions. Four machine learning algorithms (artificial neural networks, support vector machine, k-nearest neighbor, and elastic-net penalized logistic regression) were developed to predict 90-day unplanned readmissions following total knee arthroplasty and these models were evaluated using ROC AUC statistics as well as calibration and decision curve analysis.

RESULTS

Within the study cohort, 644 patients (6.4%) were readmitted within 90 days. The factors most significantly associated with 90-day unplanned hospital readmissions included drug abuse, surgical operative time, and American Society of Anaesthesiologist Physical Status (ASA) score. The machine learning models all achieved excellent performance across discrimination (AUC > 0.82), calibration, and decision curve analysis.

CONCLUSION

This study developed four machine learning models for the prediction of 90-day unplanned hospital readmissions in patients following total knee arthroplasty. The strongest predictors for unplanned hospital readmissions were drug abuse, surgical operative time, and ASA score. The study findings show excellent model performance across all four models, highlighting the potential of these models for the identification of high-risk patients prior to surgery for whom coordinated care efforts may decrease the risk of subsequent hospital readmission.

LEVEL OF EVIDENCE

Level III, case-control retrospective analysis.

Machine learning models accurately predict recurrent infection following revision total knee arthroplasty for periprosthetic joint infection

PURPOSE

This study aimed to develop and validate machine-learning models for the prediction of recurrent infection in patients following revision total knee arthroplasty for periprosthetic joint infection.

METHODS

A total of 618 consecutive patients underwent revision total knee arthroplasty for periprosthetic joint infection. The patient cohort included 165 patients with confirmed recurrent periprosthetic joint infection (PJI). Potential risk factors including patient demographics and surgical characteristics served as input to three machine-learning models which were developed to predict recurrent periprosthetic joint. The machine-learning models were assessed by discrimination, calibration and decision curve analysis.

RESULTS

The factors most significantly associated with recurrent PJI in patients following revision total knee arthroplasty for PJI included irrigation and debridement with/without modular component exchange (p < 0.001), > 4 prior open surgeries (p < 0.001), metastatic disease (p < 0.001), drug abuse (p < 0.001), HIV/AIDS (p < 0.01), presence of Enterococcus species (p < 0.01) and obesity (p < 0.01). The machine-learning models all achieved excellent performance across discrimination (AUC range 0.81-0.84).

CONCLUSION

This study developed three machine-learning models for the prediction of recurrent infections in patients following revision total knee arthroplasty for periprosthetic joint infection. The strongest predictors were previous irrigation and debridement with or without modular component exchange and prior open surgeries. The study findings show excellent model performance, highlighting the potential of these computational tools in quantifying increased risks of recurrent PJI to optimize patient outcomes.

LEVEL OF EVIDENCE

IV.

Factors contributing to 1-year dissatisfaction after total knee arthroplasty: a nomogram prediction model

Background

Identifying risk factors and early intervention are critical for improving the satisfaction rate of total knee arthroplasty (TKA). Our study aimed to identify patient-specific variables and establish a nomogram model to predict dissatisfaction at 1 year after TKA.

Methods

This prospective cohort study involved 208 consecutive primary TKA patients with end-stage arthritis who completed self-reported measures preoperatively and at 1 year postoperatively. All participants were randomized into a training cohort (n = 154) and validation cohort (n = 54). Multiple regression models with preoperative and postoperative factors were used to establish the nomogram model for dissatisfaction at 1 year postoperatively. The least absolute shrinkage and selection operator method was used to screen the suitable and effective risk factors (demographic variables, preoperative variables, surgical variable, and postoperative variables) collected. These variables were compared between the satisfied and dissatisfied groups in the training cohort. The receiver operating characteristic (ROC) curve, calibration plot, and decision curve analysis were used to validate the discrimination, calibration, and clinical usefulness of the model. Results were evaluated by internal validation of the validation cohort.

Results

The overall satisfaction rate 1 year after TKA was 77.8%. The nomogram prediction model included the following risk factors: gender; primary diagnosis; postoperative residual pain; poor postoperative range of motion; wound healing; and the rate of change in the degree of coronal lower limb alignment (hip–knee–ankle angle, HKA).The ROC curves of the training and validation cohorts were 0.9206 (95% confidence interval [CI], 0.8785–0.9627) and 0.9662 (0.9231, 1.0000) (95% CI, 0.9231, 1.0000), respectively. The Hosmer–Lemeshow test showed good calibration of the nomogram (training cohort, p = 0.218; validation cohort, p = 0.103).

Conclusion

This study developed a prediction nomogram model based on partially modifiable risk factors for predicting dissatisfaction 1 year after TKA. This model demonstrated good discriminative capacity for identifying those at greatest risk for dissatisfaction and may help surgeons and patients identify and evaluate the risk factors for dissatisfaction and optimize TKA outcomes.

The Utility of Machine Learning Algorithms for the Prediction of Early Revision Surgery After Primary Total Hip Arthroplasty

BACKGROUND

Revision total hip arthroplasty (THA) is associated with increased morbidity, mortality, and healthcare costs due to a technically more demanding surgical procedure when compared with primary THA. Therefore, a better understanding of risk factors for early revision THA is essential to develop strategies for mitigating the risk of patients undergoing early revision. This study aimed to develop and validate novel machine learning (ML) models for the prediction of early revision after primary THA.

METHODS

A total of 7,397 consecutive patients who underwent primary THA were evaluated, including 566 patients (6.6%) with confirmed early revision THA (<2 years from index THA). Electronic patient records were manually reviewed to identify patient demographics, implant characteristics, and surgical variables that may be associated with early revision THA. Six ML algorithms were developed to predict early revision THA, and these models were assessed by discrimination, calibration, and decision curve analysis.

RESULTS

The strongest predictors for early revision after primary THA were Charlson Comorbidity Index, body mass index >35 kg/m2, and depression. The six ML models all achieved excellent performance across discrimination (area under the curve >0.80), calibration, and decision curve analysis.

CONCLUSION

This study developed ML models for the prediction of early revision surgery for patients after primary THA. The study findings show excellent performance on discrimination, calibration, and decision curve analysis for all six candidate models, highlighting the potential of these models to assist in clinical practice patient-specific preoperative quantification of increased risk of early revision THA.

Machine learning classifiers do not improve prediction of hospitalization > 2 days after fast-track hip and knee arthroplasty compared with a classical statistical risk model

Background and purpose: Prediction of postoperative outcomes and length of hospital stay (LOS) of patients is vital for allocation of healthcare resources. We investigated the performance of prediction models based on machinelearning algorithms compared with a previous risk stratification model using traditional multiple logistic regression, for predicting the risk of a LOS of > 2 days after fast-track total hip and knee replacement. Patients and methods: 3 different machine learning classifiers were trained on data from the Lundbeck Centre for Fast-track Hip and Knee Replacement Database (LCDB) collected from 9,512 patients between 2016 and 2017. The chosen classifiers were a random forest classifier (RF), a support vector machine classifier with a polynomial kernel (SVM), and a multinomial Naïve-Bayes classifier (NB). Results: Comparing performance measures of the classifiers with the traditional model revealed that all the models had a similar performance in terms of F1 score, accuracy, sensitivity, specificity, area under the receiver operating curve (AUC), and area under the precision-recall curve (AUPRC). A feature importance analysis of the RF classifier found hospital, age, use of walking aid, living alone, and joint operated on to be the most relevant input features. None of the classifiers reached a clinically relevant performance with the input data from the LCDB. Interpretation: Despite the promising prospects of machine-learning practices for disease and risk prediction, none of the machine learning models tested outperformed the traditional multiple regression model in predicting which patients in this cohort had a LOS > 2 days.

Periprosthetic Joint Infection Prediction via Machine Learning: Comprehensible Personalized Decision Support for Diagnosis

BACKGROUND

The criteria outlined in the International Consensus Meeting (ICM) in 2018, which were prespecified and fixed, have been commonly practiced by clinicians to diagnose periprosthetic joint infection (PJI). We developed a machine learning (ML) system for PJI diagnosis and compared it with the ICM scoring system to verify the feasibility of ML.

METHODS

We designed an ensemble meta-learner, which combined 5 learning algorithms to achieve superior performance by optimizing their synergy. To increase the comprehensibility of ML, we developed an explanation generator that produces understandable explanations of individual predictions. We performed stratified 5-fold cross-validation on a cohort of 323 patients to compare the ML meta-learner with the ICM scoring system.

RESULTS

Cross-validation demonstrated ML's superior predictive performance to that of the ICM scoring system for various metrics, including accuracy, precision, recall, F1 score, Matthews correlation coefficient, and area under receiver operating characteristic curve. Moreover, the case study showed that ML was capable of identifying personalized important features missing from ICM and providing interpretable decision support for individual diagnosis.

CONCLUSION

Unlike ICM, ML could construct adaptive diagnostic models from the available patient data instead of making diagnoses based on prespecified criteria. The experimental results suggest that ML is feasible and competitive for PJI diagnosis compared with the current widely used ICM scoring criteria. The adaptive ML models can serve as an auxiliary system to ICM for diagnosing PJI.