Machine Learning Approaches for Hospital Acquired Pressure Injuries: A Retrospective Study of Electronic Medical Records.

Background: Many machine learning heuristics integrate well with Electronic Medical Record (EMR) systems yet often fail to surpass traditional statistical models for biomedical applications. Objective: We sought to compare predictive performances of 12 machine learning and traditional statistical techniques to predict the occurrence of Hospital Acquired Pressure Injuries (HAPI). Methods: EMR information was collected from 57,227 hospitalizations acquired from Dartmouth Hitchcock Medical Center (April 2011 to December 2016). Twelve classification algorithms, chosen based upon classic regression and recent machine learning techniques, were trained to predict HAPI incidence and performance was assessed using the Area Under the Receiver Operating Characteristic Curve (AUC). Results: Logistic regression achieved a performance (AUC = 0.91 ± 0.034) comparable to the other machine learning approaches. We report discordance between machine learning derived predictors compared to the traditional statistical model. We visually assessed important patient-specific factors through Shapley Additive Explanations. Conclusions: Machine learning models will continue to inform clinical decision-making processes but should be compared to traditional modeling approaches to ensure proper utilization. Disagreements between important predictors found by traditional and machine learning modeling approaches can potentially confuse clinicians and need to be reconciled. These developments represent important steps forward in developing real-time predictive models that can be integrated into EMR systems to reduce unnecessary harm.

African American Women's Experiences with Birth After a Prior Cesarean Section.

OBJECTIVES: Women who have had a cesarean section (C-section) and become pregnant again may choose to have a planned repeat cesarean delivery (RCD) or vaginal birth after a cesarean (VBAC). This study aimed to characterize the pregnancy and birth experiences of African American (AA) women who had a successful VBAC, failed VBAC, or RCD. METHODS: Eligible participants (N = 25) self-identified as AA, had a C-section and a subsequent birth(s) in the past 12 years, and were educated past high school. Each participant was individually interviewed via phone call. The Sort and Sift, Think and Shift method was used to evaluate interview transcripts to minimize researcher bias and emphasize the voices of the participants. RESULTS: The resulting themes included the impact of providers on pregnancy and childbirth satisfaction, the value of autonomy in maternal health decision-making, and the role that racism plays in AA women's birth experiences. Although some participants recalled a positive experience, the presence of limited autonomy, lack of support, and negative experiences with providers indicate that birth after a prior C-section for AA women can be improved. CONCLUSIONS: Providers should address their own racial biases and utilize the shared decision-making approach when their patients decide between a VBAC and RCD to improve patient satisfaction.

Patient-specific factors associated with pressure injuries revealed by electronic health record analyses.

Hospital-acquired pressure injuries (HAPIs) are a major source of unintended patient harm and unnecessary costs. The Braden Scale is widely used for risk assessment, yet it lacks specificity and clinical applications. This study used the electronic health record to examine associations between patient-specific factors and pressure injury. Adult patients (age >18) with 3-day length of stay from April 2011 to December 2016 were included. Pressure injuries were identified by ICD-9/ICD-10 codes. Longitudinal multivariate logistic regression was used to evaluate the association between patient-specific factors and HAPIs. This included 57,227 hospital encounters and 241 HAPIs. We observed 2-3 times increased likelihood of acquiring a pressure injury among patients who were malnourished or who had increased intraoperative time. The Braden subscales of nutrition, mobility, and friction showed significant predictive value. Future work is needed to assess the clinical applicability of this work.

Reduction of Hospital-acquired Pressure Injuries Using a Multidisciplinary Team Approach: A Descriptive Study.

INTRODUCTION: Hospital-acquired pressure injuries (HAPIs) increase patient mortality and length of stay (LOS). Twenty-eight full-thickness HAPIs occurred in fiscal year 2015 (FY15), and that trend continued into FY16 with 14 injuries on multiple units throughout a tertiary acute care center with 400 beds. To address this trend, a multidisciplinary Pressure Injury Prevention (PIP) team was created. OBJECTIVE: This report is a description of ongoing, hospital-wide efforts to understand the common factors of HAPI causality and to establish corrective action plans institutionally to prevent similar events in the future. METHODS: The team goals were to document the occurrence of HAPIs across all hospital units, reduce preventable full-thickness PIs to zero, and recommend institution-wide changes as those opportunities were recognized. RESULTS: Since the committee's inception in July 2015, an 89% reduction of full-thickness HAPIs, with only 2 full-thickness HAPIs in FY17 and 3 in FY18, has been seen. This effort has been hospital wide with involvement of all inpatient units and perioperative areas (including the operating rooms). Opportunities remain for improvement around the prevention of deep tissue and partial-thickness HAPIs. CONCLUSIONS: The data demonstrate that the formation of a multidisciplinary PIP team of engaged clinicians can reduce the number of preventable full-thickness HAPIs.