Who Benefits From Hip Arthroplasty or Knee Arthroplasty? Preoperative Patient-reported Outcome Thresholds Predict Meaningful Improvement.

BACKGROUND: Hip arthroplasty (HA) and knee arthroplasty (KA) are high-volume procedures. However, there is a debate about the quality of indication; that is, whether surgery is truly indicated in all patients. Patient-reported outcome measures (PROMs) may be used to determine preoperative thresholds to differentiate patients who will likely benefit from surgery from those who will not. QUESTIONS/PURPOSES: (1) What were the minimum clinically important differences (MCIDs) for three commonly used PROMs in a large population of patients undergoing HA or KA treated in a general orthopaedic practice? (2) Do patients who reach the MCID differ in important ways from those who do not? (3) What preoperative PROM score thresholds best distinguish patients who achieve a meaningful improvement 12 months postsurgery from those who do not? (4) Do patients with preoperative PROM scores below thresholds still experience gains after surgery? METHODS: Between October 1, 2019, and December 31, 2020, 4182 patients undergoing HA and 3645 patients undergoing KA agreed to be part of the PROMoting Quality study and were hence included by study nurses in one of nine participating German hospitals. From a selected group of 1843 patients with HA and 1546 with KA, we derived MCIDs using the anchor-based change difference method to determine meaningful improvements. Second, we estimated which preoperative PROM score thresholds best distinguish patients who achieve an MCID from those who do not, using the preoperative PROM scores that maximized the Youden index. PROMs were Hip Disability and Osteoarthritis Outcome Score-Physical Function short form (HOOS-PS) (scored 0 to 100 points; lower indicates better health), Knee Injury and Osteoarthritis Outcome Score-Physical Function short form (KOOS-PS) (scored 0 to 100 points; lower indicates better health), EuroQol 5-Dimension 5-level (EQ-5D-5L) (scored -0.661 to 1 points; higher indicates better health), and a 10-point VAS for pain (perceived pain in the joint under consideration for surgery within the past 7 days) (scored 0 to 10 points; lower indicates better health). The performance of derived thresholds is reported using the Youden index, sensitivity, specificity, F1 score, geometric mean as a measure of central tendency, and area under the receiver operating characteristic curve. RESULTS: MCIDs for the EQ-5D-5L were 0.2 for HA and 0.2 for KA, with a maximum of 1 point, where higher values represented better health-related quality of life. For the pain scale, they were -0.9 for HA and -0.7 for KA, of 10 points (maximum), where lower scores represent lower pain. For the HOOS-PS, the MCID was -10, and for the KOOS-PS it was -5 of 100 points, where lower scores represent better functioning. Patients who reached the MCID differed from patients who did not reach the MCID with respect to baseline PROM scores across the evaluated PROMs and for both HA and KA. Patients who reached an MCID versus those who did not also differed regarding other aspects including education and comorbidities, but this was not consistent across PROMs and arthroplasty type. Preoperative PROM score thresholds for HA were 0.7 for EQ-5D-5L (Youden index: 0.55), 42 for HOOS-PS (Youden index: 0.27), and 3.5 for the pain scale (Youden index: 0.47). For KA, the thresholds were 0.6 for EQ-5D-5L (Youden index: 0.57), 39 for KOOS-PS (Youden index: 0.25), and 6.5 for the pain scale (Youden index: 0.40). A higher Youden index for EQ-5D-5L than for the other PROMs indicates that the thresholds for EQ-5D-5L were better for distinguishing patients who reached a meaningful improvement from those who did not. Patients who did not reach the thresholds could still achieve MCIDs, especially for functionality and the pain scale. CONCLUSION: We found that patients who experienced meaningful improvements (MCIDs) mainly differed from those who did not regarding their preoperative PROM scores. We further identified that patients undergoing HA or KA with a score above 0.7 or 0.6, respectively, on the EQ-5D-5L, below 42 or 39 on the HOOS-PS or KOOS-PS, or below 3.5 or 6.5 on a 10-point joint-specific pain scale presurgery had no meaningful benefit from surgery. The thresholds can support clinical decision-making. For example, when thresholds indicate that a meaningful improvement is not likely to be achieved after surgery, other treatment options may be prioritized. Although the thresholds can be used as support, patient preferences and medical expertise must supplement the decision. Future studies might evaluate the utility of using these thresholds in practice, examine how different thresholds can be combined as a multidimensional decision tool, and derive presurgery thresholds based on additional PROMs used in practice. CLINICAL RELEVANCE: Preoperative PROM score thresholds in this study will support clinicians in decision-making through objective measures that can improve the quality of the recommendation for surgery.

Gender health gap pre- and post-joint arthroplasty: identifying affected patient-reported health domains.

BACKGROUND: As patient-reported outcomes (PROs) gain prominence in hip and knee arthroplasty (HA and KA), studies indicate PRO variations between genders. Research on the specific health domains particularly impacted is lacking. Hence, we aim to quantify the gender health gap in PROs for HA/KA patients, differentiating between general health, health-related quality of life (HrQoL), physical functioning, pain, fatigue, and depression. METHODS: The study included 3,693 HA patients (1,627 men, 2,066 women) and 3,110 KA patients (1,430 men, 1,680 women) receiving surgery between 2020 to 2021 in nine German hospitals, followed up until March 2022. Questionnaires used were: EQ-VAS, EQ-5D-5L, HOOS-PS, KOOS-PS, PROMIS-F-SF, PROMIS-D-SF, and a joint-specific numeric pain scale. PROs at admission, discharge, 12-months post-surgery, and the change from admission to 12-months (PRO-improvement) were compared by gender, tested for differences, and assessed using multivariate linear regressions. To enable comparability, PROs were transformed into z-scores (standard deviations from the mean). RESULTS: Observed differences between genders were small in all health domains and differences reduced over time. Men reported significantly better health versus women pre-HA (KA), with a difference of 0.252 (0.224) standard deviations from the mean for pain, 0.353 (0.243) for fatigue (PROMIS-F-SF), 0.327 (0.310) for depression (PROMIS-D-SF), 0.336 (0.273) for functionality (H/KOOS-PS), 0.177 (0.186) for general health (EQ-VAS) and 0.266 (0.196) for HrQoL (EQ-5D-5L). At discharge, the gender health gap reduced and even disappeared for some health dimensions since women improved in health to a greater extent than men. No gender health gap was observed in most PRO-improvements and at month 12. CONCLUSIONS: Men experiencing slightly better health than women in all health dimensions before surgery while experiencing similar health benefits 12-months post-surgery, might be an indicator of men receiving surgery inappropriately early, women unnecessarily late or both. As studies often investigate the PRO-improvement, they miss pre-surgery gender differences, which could be an important target for improvement initiatives in patient-centric care. Moreover, future research on cutoffs for meaningful between-group PRO differences per measurement time would aid the interpretation of gender health disparities. TRIAL REGISTRATION: German Register for Clinical Trials, DRKS00019916, 26 November 2019.

Stability of hospital quality indicators over time: A multi-year observational study of German hospital data.

BACKGROUND: Retrospective hospital quality indicators can only be useful if they are trustworthy signals of current or future quality. Despite extensive longitudinal quality indicator data and many hospital quality public reporting initiatives, research on quality indicator stability over time is scarce and skepticism about their usefulness widespread. OBJECTIVE: Based on aggregated, widely available hospital-level quality indicators, this paper sought to determine whether quality indicators are stable over time. Implications for health policy were drawn and the limited methodological foundation for stability assessments of hospital-level quality indicators enhanced. METHODS: Two longitudinal datasets (self-reported and routine data), including all hospitals in Germany and covering the period from 2004 to 2017, were analysed. A logistic regression using Generalized Estimating Equations, a time-dependent, graphic quintile representation of risk-adjusted rates and Spearman's rank correlation coefficient were used. RESULTS: For a total of eight German quality indicators significant stability over time was demonstrated. The probability of remaining in the best quality cluster in the future across all hospitals reached from 46.9% (CI: 42.4-51.6%) for hip replacement reoperations to 80.4% (CI: 76.4-83.8%) for decubitus. Furthermore, graphical descriptive analysis showed that the difference in adverse event rates for the 20% top performing compared to the 20% worst performing hospitals in the two following years is on average between 30% for stroke and AMI and 79% for decubitus. Stability over time has been shown to vary strongly between indicators and treatment areas. CONCLUSION: Quality indicators were found to have sufficient stability over time for public reporting. Potentially, increasing case volumes per hospital, centralisation of medical services and minimum-quantity regulations may lead to more stable and reliable quality of care indicators. Finally, more robust policy interventions such as outcome-based payment, should only be applied to outcome indicators with a higher level of stability over time. This should be subject to future research.

Electronic Patient-Reported Outcome Monitoring to Improve Quality of Life After Joint Replacement: Secondary Analysis of a Randomized Clinical Trial.

Importance: Although remote patient-reported outcome measure (PROM) monitoring has shown promising results in cancer care, there is a lack of research on PROM monitoring in orthopedics. Objective: To determine whether PROM monitoring can improve health outcomes for patients with joint replacement compared with the standard of care. Design, Setting, and Participants: A 2-group, patient-level randomized clinical trial (PROMoting Quality) across 9 German hospitals recruited patients aged 18 years or older with primary hip or knee replacement from October 1, 2019, to December 31, 2020, with follow-up until March 31, 2022. Interventions: Intervention and control groups received the standard of care and PROMs at hospital admission, discharge, and 12 months after surgery. In addition, the intervention group received PROMs at 1, 3, and 6 months after surgery. Based on prespecified PROM score thresholds, at these times, an automated alert signaled critical recovery paths to hospital study nurses. On notification, study nurses contacted patients and referred them to their physicians if necessary. Main Outcomes and Measures: The prespecified outcomes were the mean change in PROM scores (European Quality of Life 5-Dimension 5-Level version [EQ-5D-5L; range, -0.661 to 1.0, with higher values indicating higher levels of health-related quality of life (HRQOL)], European Quality of Life Visual Analogue Scale [EQ-VAS; range, 0-100, with higher values indicating higher levels of HRQOL], Hip Disability and Osteoarthritis Outcome Score-Physical Function Shortform [HOOS-PS; range, 0-100, with lower values indicating lower physical impairment] or Knee Injury and Osteoarthritis Outcome Score-Physical Function Shortform [KOOS-PS; range, 0-100, with lower values indicating lower physical impairment], Patient-Reported Outcomes Measurement Information System [PROMIS]-fatigue [range, 33.7-75.8, with lower values indicating lower levels of fatigue], and PROMIS-depression [range, 41-79.4, with lower values indicating lower levels of depression]) from baseline to 12 months after surgery. Analysis was on an intention-to-treat basis. Results: The study included 3697 patients with hip replacement (mean [SD] age, 65.8 [10.6] years; 2065 women [55.9%]) and 3110 patients with knee replacement (mean [SD] age, 66.0 [9.2] years; 1669 women [53.7%]). Exploratory analyses showed significantly better health outcomes in the intervention group on all PROMs except the EQ-5D-5L among patients with hip replacement, with a 2.10-point increase on the EQ-VAS in the intervention group compared with the control group (HOOS-PS, -1.86 points; PROMIS-fatigue, -0.69 points; PROMIS-depression, -0.57 points). Patients in the intervention group with knee replacement had a 1.24-point increase on the EQ-VAS, as well as significantly better scores on the KOOS-PS (-0.99 points) and PROMIS-fatigue (-0.84 points) compared with the control group. Mixed-effect models showed a significant difference in improvement on the EQ-VAS (hip replacement: effect estimate [EE], 1.66 [95% CI, 0.58-2.74]; knee replacement: EE, 1.71 [95% CI, 0.53-2.90]) and PROMIS-fatigue (hip replacement: EE, -0.65 [95% CI, -1.12 to -0.18]; knee replacement: EE, -0.71 [95% CI, -1.23 to -0.20]). The PROMIS-depression score was significantly reduced in the hip replacement group (EE, -0.60 [95% CI, -1.01 to -0.18]). Conclusions and Relevance: In this randomized clinical trial, the PROM-based monitoring intervention led to a small improvement in HRQOL and fatigue among patients with hip or knee replacement, as well as in depression among patients with hip replacement. Trial registration: Deutsches Register Klinischer Studien ID: DRKS00019916.

Predicting whether patients will achieve minimal clinically important differences following hip or knee arthroplasty.

Aims: A substantial fraction of patients undergoing knee arthroplasty (KA) or hip arthroplasty (HA) do not achieve an improvement as high as the minimal clinically important difference (MCID), i.e. do not achieve a meaningful improvement. Using three patient-reported outcome measures (PROMs), our aim was: 1) to assess machine learning (ML), the simple pre-surgery PROM score, and logistic-regression (LR)-derived performance in their prediction of whether patients undergoing HA or KA achieve an improvement as high or higher than a calculated MCID; and 2) to test whether ML is able to outperform LR or pre-surgery PROM scores in predictive performance. Methods: MCIDs were derived using the change difference method in a sample of 1,843 HA and 1,546 KA patients. An artificial neural network, a gradient boosting machine, least absolute shrinkage and selection operator (LASSO) regression, ridge regression, elastic net, random forest, LR, and pre-surgery PROM scores were applied to predict MCID for the following PROMs: EuroQol five-dimension, five-level questionnaire (EQ-5D-5L), EQ visual analogue scale (EQ-VAS), Hip disability and Osteoarthritis Outcome Score-Physical Function Short-form (HOOS-PS), and Knee injury and Osteoarthritis Outcome Score-Physical Function Short-form (KOOS-PS). Results: Predictive performance of the best models per outcome ranged from 0.71 for HOOS-PS to 0.84 for EQ-VAS (HA sample). ML statistically significantly outperformed LR and pre-surgery PROM scores in two out of six cases. Conclusion: MCIDs can be predicted with reasonable performance. ML was able to outperform traditional methods, although only in a minority of cases.

Exploring treatment effect heterogeneity of a PROMs alert intervention in knee and hip arthroplasty patients: A causal forest application.

Patient reported outcome measures (PROMs) experience an uptake in use for hip (HA) and knee arthroplasty (KA) patients. As they may be used for patient monitoring interventions, it remains unclear whether their use in HA/KA patients is effective, and which patient groups benefit the most. Nonetheless, knowledge about treatment effect heterogeneity is crucial for decision makers to target interventions towards specific subgroups that benefit to a greater extend. Therefore, we evaluate the treatment effect heterogeneity of a remote PROM monitoring intervention that includes ∼8000 HA/KA patients from a randomized controlled trial conducted in nine German hospitals. The study setting gave us the unique opportunity to apply a causal forest, a recently developed machine learning method, to explore treatment effect heterogeneity of the intervention. We found that among both HA and KA patients, the intervention was especially effective for patients that were female, >65 years of age, had a blood pressure disease, were not working, reported no backpain and were adherent. When transferring the study design into standard care, policy makers should make use of the knowledge obtained in this study and allocate the treatment towards subgroups for which the treatment is especially effective.

Machine learning as a tool to identify inpatients who are not at risk of adverse drug events in a large dataset of a tertiary care hospital in the USA.

AIMS: Adverse drug events (ADEs) are a major threat to inpatients in the United States of America (USA). It is unknown how well machine learning (ML) is able to predict whether or not a patient will suffer from an ADE during hospital stay based on data available at hospital admission for emergency department patients of all ages (binary classification task). It is further unknown whether ML is able to outperform logistic regression (LR) in doing so, and which variables are the most important predictors. METHODS: In this study, 5 ML models- namely a random forest, gradient boosting machine (GBM), ridge regression, least absolute shrinkage and selection operator (LASSO) regression, and elastic net regression-as well as a LR were trained and tested for the prediction of inpatient ADEs identified using ICD-10-CM codes based on comprehensive previous work in a diverse population. In total, 210 181 observations from patients who were admitted to a large tertiary care hospital after emergency department stay between 2011 and 2019 were included. The area under the receiver operating characteristics curve (AUC) and AUC-precision-recall (AUC-PR) were used as primary performance indicators. RESULTS: Tree-based models performed best with respect to AUC and AUC-PR. The gradient boosting machine (GBM) reached an AUC of 0.747 (95% confidence interval (CI): 0.735 to 0.759) and an AUC-PR of 0.134 (95% CI: 0.131 to 0.137) on unforeseen test data, while the random forest reached an AUC of 0.743 (95% CI: 0.731 to 0.755) and an AUC-PR of 0.139 (95% CI: 0.135 to 0.142), respectively. ML statistically significantly outperformed LR both on AUC and AUC-PR. Nonetheless, overall, models did not differ much with respect to their performance. Most important predictors were admission type, temperature and chief complaint for the best performing model (GBM). CONCLUSIONS: The study demonstrated a first application of ML to predict inpatient ADEs based on ICD-10-CM codes, and a comparison with LR. Future research should address concerns arising from low precision and related problems.

The application of machine learning to predict high-cost patients: A performance-comparison of different models using healthcare claims data.

Our aim was to predict future high-cost patients with machine learning using healthcare claims data. We applied a random forest (RF), a gradient boosting machine (GBM), an artificial neural network (ANN) and a logistic regression (LR) to predict high-cost patients in the following year. Therefore, we exploited routinely collected sickness funds claims and cost data of the years 2016, 2017 and 2018. Various specifications of each algorithm were trained and cross-validated on training data (n = 20,984) with claims and cost data from 2016 and outcomes from 2017. The best performing specifications of each algorithm were selected based on validation dataset performance. For performance comparison, selected models were applied to unforeseen data with features of the year 2017 and outcomes of the year 2018 (n = 21,146). The RF was the best performing algorithm measured by the area under the receiver operating curve (AUC) with a value of 0.883 (95% confidence interval (CI): 0.872-0.893) on test data, followed by the GBM (AUC = 0.878; 95% CI: 0.867-0.889). The ANN (AUC = 0.846; 95% CI: 0.834-0.857) and LR (AUC = 0.839; 95% CI: 0.826-0.852) were significantly outperformed by the GBM and the RF. All ML algorithms and the LR performed ´good´ (i.e. 0.9 > AUC ≥ 0.8). We were able to develop machine learning models that predict high-cost patients with 'good' performance facilitating routinely collected sickness fund claims and cost data. We found that tree-based models performed best and outperformed the ANN and LR.

Can minimal clinically important differences in patient reported outcome measures be predicted by machine learning in patients with total knee or hip arthroplasty? A systematic review.

OBJECTIVES: To systematically review studies using machine learning (ML) algorithms to predict whether patients undergoing total knee or total hip arthroplasty achieve an improvement as high or higher than the minimal clinically important differences (MCID) in patient reported outcome measures (PROMs) (classification problem). METHODS: Studies were eligible to be included in the review if they collected PROMs both pre- and postintervention, reported the method of MCID calculation and applied ML. ML was defined as a family of models which automatically learn from data when selecting features, identifying nonlinear relations or interactions. Predictive performance must have been assessed using common metrics. Studies were searched on MEDLINE, PubMed Central, Web of Science Core Collection, Google Scholar and Cochrane Library. Study selection and risk of bias assessment (ROB) was conducted by two independent researchers. RESULTS: 517 studies were eligible for title and abstract screening. After screening title and abstract, 18 studies qualified for full-text screening. Finally, six studies were included. The most commonly applied ML algorithms were random forest and gradient boosting. Overall, eleven different ML algorithms have been applied in all papers. All studies reported at least fair predictive performance, with two reporting excellent performance. Sample size varied widely across studies, with 587 to 34,110 individuals observed. PROMs also varied widely across studies, with sixteen applied to TKA and six applied to THA. There was no single PROM utilized commonly in all studies. All studies calculated MCIDs for PROMs based on anchor-based or distribution-based methods or referred to literature which did so. Five studies reported variable importance for their models. Two studies were at high risk of bias. DISCUSSION: No ML model was identified to perform best at the problem stated, nor can any PROM said to be best predictable. Reporting standards must be improved to reduce risk of bias and improve comparability to other studies.

The detection and prevention of adverse drug events in nursing home and home care patients: Study protocol of a quasi-experimental study.

AIM: To estimate the cost-effectiveness of an intervention facilitating the early detection of adverse drug events through the means of health professional training and the application of a digital screening tool. DESIGN: Multi-centred non-randomized controlled trial from August 2018 to March 2020 including 65 nursing homes or home care providers. METHODS: We aim to estimate the effect of the intervention on the rate of adverse drug events as primary outcome through a quasi-experimental empirical study design. As secondary outcomes, we use hospital admissions and falls. All outcomes will be measured on patient-month level. Once the causal effect of the intervention is estimated, cost-effectiveness will be calculated. For cost-effectiveness, we include all patient costs observed by the German statutory health insurance. RESULTS: The results of this study will inform about the cost-effectiveness of the optimized drug supply intervention and provide evidence for potential reimbursement within the German statutory health insurance system.