RESUMO
This study aimed to explore the potential of gait analysis coupled with supervised machine learning models as a predictive tool for assessing post-injury complications such as infection, malunion, or hardware irritation among individuals with lower extremity fractures. We prospectively identified participants with lower extremity fractures at a tertiary academic center. These participants underwent gait analysis with a chest-mounted inertial measurement unit device. Using customized software, the raw gait data were preprocessed, emphasizing 12 essential gait variables. The data were standardized, and several machine learning models, including XGBoost, logistic regression, support vector machine, LightGBM, and Random Forest, were trained, tested, and evaluated. Special attention was given to class imbalance, addressed using the synthetic minority oversampling technique (SMOTE). Additionally, we introduced a novel methodology to compute the post-injury recovery rate for gait variables, which operates independently of the time difference between the gait analyses of different participants. XGBoost was identified as the optimal model both before and after the application of SMOTE. Before using SMOTE, the model achieved an average test area under the ROC curve (AUC) of 0.90, with a 95% confidence interval (CI) of [0.79, 1.00], and an average test accuracy of 86%, with a 95% CI of [75%, 97%]. Through feature importance analysis, a pivotal role was attributed to the duration between the occurrence of the injury and the initial gait analysis. Data patterns over time revealed early aggressive physiological compensations, followed by stabilization phases, underscoring the importance of prompt gait analysis. χ2 analysis indicated a statistically significant higher readmission rate among participants with underlying medical conditions (p = 0.04). Although the complication rate was also higher in this group, the association did not reach statistical significance (p = 0.06), suggesting a more pronounced impact of medical conditions on readmission rates rather than on complications. This study highlights the transformative potential of integrating advanced machine learning techniques like XGBoost with gait analysis for orthopedic care. The findings underscore a shift toward a data-informed, proactive approach in orthopedics, enhancing patient outcomes through early detection and intervention. The χ2 analysis added crucial insights into the broader clinical implications, advocating for a comprehensive treatment strategy that accounts for the patient's overall health profile. The research paves the way for personalized, predictive medical care in orthopedics, emphasizing the importance of timely and tailored patient assessments.
Assuntos
Análise da Marcha , Humanos , Masculino , Análise da Marcha/métodos , Feminino , Pessoa de Meia-Idade , Adulto , Idoso , Aprendizado de Máquina , Estudos Prospectivos , Fraturas Ósseas , MarchaRESUMO
OBJECTIVES: To evaluate the work relative value units (RVUs) attributed per minute of operative time (wRVU/min) in fixation of acetabular fractures, evaluate surgical factors that influence wRVU/min, and compare wRVU/min with other procedures. DESIGN: Retrospective. SETTING: Level 1 academic center. PATIENT SELECTION CRITERIA: Two hundred fifty-one operative acetabular fractures (62 A, B, C) from 2015 to 2021. OUTCOME MEASURES AND COMPARISONS: Work relative value unit per minute of operative time for each acetabular current procedural terminology (CPT) code. Surgical approach, patient positioning, total room time, and surgeon experience were collected. Comparison wRVU/min were collected from the literature. RESULTS: The mean wRVU per surgical minute for each CPT code was (1) CPT 27226 (isolated wall fracture): 0.091 wRVU/min, (2) CPT 27227 (isolated column or transverse fracture): 0.120 wRVU/min, and (3) CPT 27228 (associated fracture types): 0.120 wRVU/min. Of fractures with single approaches, anterior approaches generated the least wRVU/min (0.091 wRVU/min, P = 0.0001). Average nonsurgical room time was 82.1 minutes. Surgeon experience ranged from 3 to 26 years with operative time decreasing as surgeon experience increased ( P = 0.03). As a comparison, the wRVU/min for primary and revision hip arthroplasty have been reported as 0.26 and 0.249 wRVU/min, respectively. CONCLUSIONS: The wRVUs allocated per minute of operative time for acetabular fractures is less than half of other reported hip procedures and lowest for isolated wall fractures. There was a significant amount of nonsurgical room time that should be accounted for in compensation models. This information should be used to ensure that orthopaedic trauma surgeons are being appropriately supported for managing these fractures. LEVEL OF EVIDENCE: Economic Level IV. See Instructions for Authors for a complete description of levels of evidence.
Assuntos
Fraturas do Quadril , Ortopedia , Fraturas da Coluna Vertebral , Cirurgiões , Humanos , Duração da Cirurgia , Estudos RetrospectivosRESUMO
OBJECTIVE: The objective is to quantify the rate of opioid and benzodiazepine prescribing for the diagnosis of shoulder osteoarthritis across a large healthcare system and to describe the impact of a clinical decision support intervention on prescribing patterns. DESIGN: A prospective observational study. SETTING: One large healthcare system. PATIENTS AND PARTICIPANTS: Adult patients presenting with shoulder osteoarthritis. INTERVENTIONS: A clinical decision support intervention that presents an alert to prescribers when patients meet criteria for increased risk of opioid use disorder. MAIN OUTCOME MEASURE: The percentage of patients receiving an opioid or benzodiazepine, the percentage who had at least one risk factor for misuse, and the percent of encounters in which the prescribing decision was influenced by the alert were the main outcome measures. RESULTS: A total of 5,380 outpatient encounters with a diagnosis of shoulder osteoarthritis were included. Twenty-nine percent (n = 1,548) of these encounters resulted in an opioid or benzodiazepine prescription. One-third of those who received a prescription had at least one risk factor for prescription misuse. Patients were more likely to receive opioids from the emergency department or urgent care facilities (40 percent of encounters) compared to outpatient facilities (28 percent) (p < .0001). Forty-four percent of the opioid prescriptions were for "potent opioids" (morphine milliequivalent conversion factor > 1). Of the 612 encounters triggering an alert, the prescribing decision was influenced (modified or not prescribed) in 53 encounters (8.7 percent). All but four (0.65 percent) of these encounters resulted in an opioid prescription. CONCLUSION: Despite evidence against routine opioid use for osteoarthritis, one-third of patients with a primary diagnosis of glenohumeral osteoarthritis received an opioid prescription. Of those who received a prescription, over one-third had a risk factor for opioid misuse. An electronic clinic decision support tool influenced the prescription in less than 10 percent of encounters.