Search

Background Robotic-assisted total knee arthroplasty (TKA) is a growing technique in adult reconstruction. The variations between robotic-assisted and conventional TKA could lead to changes in immediate postoperative outcomes. We aimed to evaluate for differences in postoperative pain, discharge day, as well as post-hospital disposition (home vs subacute rehabilitation facility [SAR]) between robotic-assisted and conventional TKA. Methods We retrospectively identified 2 cohorts of patients who underwent either conventional or robotic-assisted TKA between January 2019 and July 2019. Their average pain scores from postoperative day 0, day 1, and day 2 were recorded. Their postoperative discharge day was recorded, as well as their disposition to either home or a SAR. Preoperatively, all patients are offered robotic-assisted TKA, and only those who want the procedure and undergo a preoperative CT scan receive the robotic-assisted surgery. Statistical analysis was conducted using SPSS. Results One hundred sixty-six patients were identified with 83 in each cohort. No differences between age, race, and gender were found. Despite minor variations in pain levels, the overall postoperative pain score analysis did not strongly favor one technique over the other. The robotic-assisted group had a significantly higher amount of patients discharged to home instead of a SAR and also had a shorter time to discharge than the conventional group. Conclusions Robotic-assisted TKA has similar postoperative pain scores compared with conventional TKA. The robotic-assisted cohort demonstrated other benefits including earlier discharge and are more likely to be discharged home instead of a SAR.

A 75-year-old man with dilated cardiomyopathy status post-implantable cardioverter-defibrillator (ICD) presented with progressive dyspnea and ascites refractory to medical therapy. He underwent paracentesis for symptom relief. The patient was discharged to a local rehabilitation facility and

Summary Healthcare costs due to unplanned readmissions are high and negatively affect health and wellness of patients. Hospital readmission is an undesirable outcome for elderly patients. Here, we present readmission risk prediction using five machine learning approaches for predicting 30-day unplanned readmission for elderly patients (age ≥ 50 years). We use a comprehensive and curated set of variables that include frailty, comorbidities, high-risk medications, demographics, hospital, and insurance utilization to build these models. We conduct a large-scale study with electronic health record (her) data with over 145,000 observations from 76,000 patients. Findings indicate that the category boost (CatBoost) model outperforms other models with a mean area under the curve (AUC) of 0.79. We find that prior readmissions, discharge to a rehabilitation facility, length of stay, comorbidities, and frailty indicators were all strong predictors of 30-day readmission. We present in-depth insights using Shapley additive explanations (SHAP), the state of the art in machine learning explainability., Highlights • We report a 0.79 AUROC for 30-day readmission prediction • We use frailty, comorbidity, high-risk medications, and demographics for improved accuracy • We identify clusters of high-risk patients based on sets of patient features • Explainability is a prime focus for model predictions at different levels of granularity, The bigger picture Unplanned readmission currently costs the United States millions of dollars. Predicting whether an incoming patient is at a high risk of readmission can help target healthcare efforts better to reduce this risk. In this age of big data, we can use machine learning to analyze a cohort of variables to pinpoint the risk of readmission. Our work does exactly that. One of the hindrances for adoption of artificial intelligence in healthcare is the lack of explainability. To combat that, we provide several mechanisms for understanding reasons for the model's predictions, starting at a global level across the entire dataset and down to individual patient observations. These explanations enhance confidence in the model's decision making., Automated prediction of readmission risk has the potential to save millions of dollars in healthcare costs and can improve patient care. Our work presents machine learning models that take into account various facets of patients, such as demographics, comorbidities, and frailty parameters, to accurately estimate their risk of being readmitted within 30 days. We place high importance on explainability, thereby enhancing confidence in the automated models.