Influence of day of surgery and prediction of LOS > 2 days after fast-track hip and knee replacement

Applying models of care for total hip and knee arthroplasty: External validation of a published predictive model to identify extended stay risk prior to lower-limb arthroplasty

OBJECTIVE

This study aimed to externally validate a reported model for identifying patients requiring extended stay following lower limb arthroplasty in a new setting.

DESIGN

External validation of a previously reported prognostic model, using retrospective data.

SETTING

Medium-sized hospital orthopaedic department, Australia.

PARTICIPANTS

Electronic medical records were accessed for data collection between Sep-2019 and Feb-2020 and retrospective data extracted from 200 randomly selected total hip or knee arthroplasty patients.

INTERVENTION

Participants received total hip or knee replacement between 2-Feb-16 and 4-Apr-19. This study was a non-interventional retrospective study.

MAIN MEASURES

Model validation was assessed with discrimination, calibration on both original and adjusted forms of the candidate model. Decision curve analysis was conducted on the outputs of the adjusted model to determine net benefit at a predetermined decision threshold (0.5).

RESULTS

The original model performed poorly, grossly overestimating length of stay with mean calibration of -3.6 (95% confidence interval -3.9 to -3.2) and calibration slope of 0.52. Performance improved following adjustment of the model intercept and model coefficients (mean calibration 0.48, 95% confidence interval 0.16 to 0.80 and slope of 1.0), but remained poorly calibrated at low and medium risk threshold and net benefit was modest (three additional patients per hundred identified as at-risk) at the a-priori risk threshold.

CONCLUSIONS

External validation demonstrated poor performance when applied to a new patient population and would provide limited benefit for our institution. Implementation of predictive models for arthroplasty should include practical assessment of discrimination, calibration and net benefit at a clinically acceptable threshold.

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.

Patient characteristics and procedural variables are associated with length of stay and hospital cost among unilateral primary total hip arthroplasty patients: a single-center retrospective cohort study

BACKGROUND

Total hip arthroplasty (THA) is a successful treatment for many hip diseases. Length of stay (LOS) and hospital cost are crucial parameters to quantify the medical efficacy and quality of unilateral primary THA patients. Clinical variables associated with LOS and hospital costs haven't been investigated thoroughly.

METHODS

The present study retrospectively explored the contributors of LOS and hospital costs among a total of 452 unilateral primary THA patients from January 2019 to January 2020. All patients received conventional in-house rehabilitation services within our institute prior to discharge. Outcome parameters included LOS and hospital cost while clinical variables included patient characteristics and procedural variables. Multivariable linear regression analysis was performed to assess the association between outcome parameters and clinical variables by controlling confounding factors. Moreover, we analyzed patients in two groups according to their diagnosis with femur neck fracture (FNF) (confine THA) or non-FNF (elective THA) separately.

RESULTS

Among all 452 eligible participants (266 females and 186 males; age 57.05 ± 15.99 year-old), 145 (32.08%) patients diagnosed with FNF and 307 (67.92%) diagnosed with non-FNF were analyzed separately. Multivariable linear regression analysis revealed that clinical variables including surgery duration, transfusion, and comorbidity (stroke) among the elective THA patients while the approach and comorbidities (stoke, diabetes mellitus, coronary heart disease) among the confine THA patients were associated with a prolonged LOS (P < 0.05). Variables including the American Society of Anesthesiologists classification (ASA), duration, blood loss, and transfusion among the elective THA while the approach, duration, blood loss, transfusion, catheter, and comorbidities (stoke and coronary heart disease) among the confine THA were associated with higher hospital cost (P < 0.05). The results revealed that variables were associated with LOS and hospital cost at different degrees among both elective and confine THA.

CONCLUSIONS

Specific clinical variables of the patient characteristics and procedural variables are associated the LOS and hospital cost, which may be different between the elective and confine THA patients. The findings may indicate that evaluation and identification of detailed perioperative factors are beneficial in managing perioperative preparation, adjusting patients' anticipation, decreasing LOS, and reducing hospital cost.

Key Components, Current Practice and Clinical Outcomes of ERAS Programs in Patients Undergoing Orthopedic Surgery: A Systematic Review

Enhanced recovery after surgery (ERAS) protocols have led to improvements in outcomes in several surgical fields, through multimodal optimization of patient pathways, reductions in complications, improved patient experiences and reductions in the length of stay. However, their use has not been uniformly recognized in all orthopedic fields, and there is still no consensus on the best implementation process. Here, we evaluated pre-, peri-, and post-operative key elements and clinical evidence of ERAS protocols, measurements, and associated outcomes in patients undergoing different orthopedic surgical procedures. A systematic literature search on PubMed, Scopus, and Web of Science Core Collection databases was conducted to identify clinical studies, from 2012 to 2022. Out of the 1154 studies retrieved, 174 (25 on spine surgery, 4 on thorax surgery, 2 on elbow surgery and 143 on hip and/or knee surgery) were considered eligible for this review. Results showed that ERAS protocols improve the recovery from orthopedic surgery, decreasing the length of hospital stays (LOS) and the readmission rates. Comparative studies between ERAS and non-ERAS protocols also showed improvement in patient pain scores, satisfaction, and range of motion. Although ERAS protocols in orthopedic surgery are safe and effective, future studies focusing on specific ERAS elements, in particular for elbow, thorax and spine, are mandatory to optimize the protocols.

Same-day discharge to home is feasible and safe in up to 75% of unselected total hip and knee arthroplasty

PURPOSE

Though numerous studies highlighted benefits of ambulatory total joint arthroplasty (TJA), most had selected patients with age and comorbidities thresholds. We aimed to report proportions of unselected TJAs that could be scheduled for and operated in ambulatory settings, and to determine factors that hinder same-day discharge (SDD).

METHODS

We studied 1100 consecutive primary TJAs (644 THAs and 456 TKAs) that were prepared following a multidisciplinary protocol for patient education and logistical preparation. Data were stratified for THA vs TKA and for success vs failure of SDD to home and multivariable analysis was performed to determine factors associated with failure of scheduled SDD to home.

RESULTS

In total, 860 (78.2%) were scheduled for ambulatory surgery, but only 819 (74.5%) achieved SDD to home; 240 (21.8%) were scheduled for non-ambulatory surgery, but 103 (9.3%) achieved SDD to rehabilitation centre. Re-operations were required in 9 (1.0%) ambulatory TJAs vs 2 (0.8%) non-ambulatory TJAs (p = 0.769), while revisions were required in 13 (1.5%) ambulatory TJAs vs 1 (0.4%) non-ambulatory TJAs (p = 0.181). Multivariable analysis confirmed that failure of SDD to home was greater for women (OR 2.59; p = 0.011) and THA (vs TKA, OR 2.41; p = 0.023).

CONCLUSION

With appropriate education and preparation, 75% of unselected primary hip and knee arthroplasties achieved SDD to home without compromising risks of complications, re-operations, or revisions. A further 9% achieved SDD to rehabilitation centre, implying that 84% of patients did not require overnight stay. These findings suggest that ambulatory surgery is feasible and safe to implement in most unselected lower limb arthroplasties.

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.