Background Overdose fatality review (OFR) is a public health process in which cases of fatal overdose are carefully reviewed to identify prevention strategies. Current OFR requires review of multiple unconnected data sources, which is a manually intensive process. The Substance Misuse Data Commons (SMDC) was created to link electronic health record data with data from local and state agencies into a single, cloud-based e-platform but does not currently have a data visualization tool. Objective We aimed to use human factors design principles to develop a comprehensive dashboard for the SMDC that could facilitate enhanced processes to support OFR. Methods We first surveyed OFR leaders in Wisconsin using the National Aeronautics and Space Administration-Task Load Index to understand the cognitive workload of 3 tasks: (1) analysis of population-level overdose trends, (2) selection and preparation of individual cases for review, and (3) abstraction of data from individual causes. We then conducted semistructured interviews to identify targets for workflow optimization. Next, we developed a prototype dashboard for evaluation using a synthetic dataset built with GPT-4. We subsequently performed iterative design sessions with heuristic evaluations and collected end-user feedback on the final prototype via a second round of semistructured interviews and targeted surveys, including the Unified Theory of Acceptance and Use of Technology and the Perceived Usefulness Questionnaire. Results The National Aeronautics and Space Administration-Task Load Index revealed a moderately high mental workload with the current workflow for all 3 tasks, with mean scores of 12.60 (SD 3.31), 11.90 (SD 3.57), and 12.43 (SD 5.41) for tasks 1, 2, and 3, respectively. Interviews pointed to causes including technological challenges and a reliance on manual processes. The prototype dashboard addressed these concerns by integrating multiple data sources to generate population-level visualizations and patient-level event timelines. End users reported the potential for improved efficiency and data accessibility compared to antecedent processes. The Unified Theory of Acceptance and Use of Technology results indicated the dashboard would likely be adopted if made available, with a mean of 4.07 out of 5.00 (SD 0.65). The Perceived Usefulness Questionnaire results suggested moderate usefulness for both the aggregate and individual-level data, with means of 3.61 (SD 0.82) and 3.64 (SD 0.85) out of 5.00, respectively. Conclusions OFR is a data-intensive process that traditionally demands substantial cognitive and manual effort, and there are multiple barriers to efficiently collecting data and presenting them for review. The dashboard offers a user-centered, informatics-based approach to streamline data aggregation and presentation, potentially enhancing the efficiency of case reviews. Implementing a dashboard that consolidates and visualizes disparate data sources has the potential to alleviate the manual workload in OFR. Ultimately, our aim is to deliver a finalized data dashboard with real-world SMDC data, giving OFR leaders additional tools to aid in their rigorous work shaping interventions to reduce overdose fatalities.

Objectives:Patients with substance misuse are at high risk for clinical deterioration, and pre-hospital encounters constitute important risk factors. We sought to incorporate these risk factors into a novel prediction model using linked electronic health record, emergency medical services (EMSs), and claims data. Materials and methods:Using 23 454 hospital encounters, we developed machine-learning models to predict mechanical ventilation, vasopressor initiation, or death within 12 hours of a vital sign or laboratory measurement. Models were compared to the modified early warning score. Results:Extreme gradient boosting achieved an area under the receiver operating characteristic curve of 0.93, significantly outperforming modified early warning score (MEWS 0.79). Notably, removing EMS and claims data did not reduce predictive performance. Discussion:The model demonstrated strong predictive performance of clinical deterioration in this high-risk population. Incorporation of pre-hospital risk factors did not improve performance beyond EHR data alone. Conclusion:Electronic health record-based machine learning models can support accurate short-term prediction of clinical deterioration among this population at risk for delayed recognition.

OBJECTIVES:The National Emergency Medical Services Information System (NEMSIS) is a large repository of de-identified Emergency Medical Services (EMS) data supported by the National Highway Traffic Safety Administration and is a cornerstone of large database EMS research. A known concern with large EMS datasets like NEMSIS is that multiple records may exist for a single episode of patient care. However, the extent of potentially related records is not well described. We evaluated potential EMS episodes of care contained within NEMSIS. METHODS:We used an observational study design and conducted a retrospective analysis of the 2023 NEMSIS dataset. We identified hierarchical clusters of likely related episodes of care for the same patient using tiered criteria of temporal (dispatch call times of 30 s to 2 min) and Census Division, EMS agency, urbanicity, patient demographics, dispatch characteristics, clinical impression, and location type. We evaluated all possible combinations of these criteria and reported the percentage of episodes of care out of all included records overall and for five specific EMS subpopulations: cardiac arrest, trauma, stroke, airway intervention, and pediatric patients. RESULTS:We included 36,757,562 EMS records. When applying all criteria, the final algorithm identified 36,552,901 likely episodes of care, representing 99.4% of all records. Most episodes of care (36,353,329, 99.5%) consisted of a single EMS record, with fewer episodes having 2 (195,786, 0.5%), 3 (3,088, <0.1%), 4 (428, <0.1%), and 5 or more (270, <0.1%) records. When evaluating all possible alternative combinations of the criteria, most resulted in a similar number of episodes of care. Among the subpopulations, the number of episodes of care approached the total number of encounters for that condition: cardiac arrest (89.2%-98.5%), pediatric (73.8%-99.3%), stroke (89.5%-99.3%), airway procedure (95.4%-99.4%), and trauma (84.1%-99.6%). Findings were robust to sensitivity analyses which relaxed the time- and age-based criteria. CONCLUSIONS:Nearly all included EMS records in NEMSIS appeared to represent unique patient-care episodes. Approximately 0.5% of episodes involved more than one EMS record, suggesting that related records or multi-unit documentation is uncommon but should be considered in analyses, particularly for selected clinical subgroups.

RATIONALE: A rising proportion of hospitalizations involve patients with substance misuse, who are more likely to require intensive care than the general population. Although these patients are generally younger, they suffer worse outcomes, suggesting that risk prediction models may need to account for health characteristics unique to this group. The objective of this study was to develop a machine learning model to predict critical illness or death among hospitalized patients with substance misuse, using data from electronic health records, emergency medical services (EMS), insurance claims, and socioeconomic indicators. METHODS: A retrospective study was conducted using data from 2017 to 2021, extracted from the Substance Misuse Data Commons, which includes all adult patients with an International Classification of Disease Code for substance misuse who presented to two University of Wisconsin hospitals. The primary outcome was defined as initiation of vasopressors or invasive mechanical ventilation (IMV) or death in the next 12 hours. Predictors included demographics, comorbidities, hospital transport, prior EMS use, prior hospitalizations, insurance type, location in the hospital, vital signs, laboratory results, and time since hospital presentation. The dataset was randomly split, with 70% of encounters assigned to the training cohort and 30% to the testing cohort. An extreme gradient boosted machine (XGBoost) classification model was trained, and the area under the curve (AUC) was evaluated in the test cohort for the primary outcome. The model's performance was compared to the Modified Early Warning Score (MEWS) prediction of the primary outcome using the Delong test. RESULTS: Of the 23,454 hospital encounters, 2,665 (11.3%) experienced the primary outcome. In the test cohort, the XGBoost model achieved an AUC of 0.93 (95% CI: 0.93-0.93) for predicting the initiation of vasopressors or IMV or death within 12 hours. The most important predictors, shown in Figure 1, were level of consciousness, fraction of inspired oxygen and location in the hospital. The model outperformed MEWS, which had an AUC of 0.79 (95% CI: 0.78-0.79; p <0.001) in the test cohort. CONCLUSION: We developed a machine learning model that accurately predicts critical illness in hospitalized patients with a history of substance misuse and offers the potential for facilitating timelier interventions. The most important predictors demonstrate face validity, and the model outperforms existing generalized indices for predicting critical illness or death. Future studies should evaluate whether this model prospectively accelerates the detection of critical illness and improves outcomes for this patient group.

Patients with substance misuse who are admitted to the hospital are at heightened risk for adverse outcomes, such as readmission and death. This study aims to develop methods to identify at-risk patients to facilitate timely interventions that can improve outcomes and optimize healthcare resources. To accomplish this, we leveraged the Substance Misuse Data Commons to predict 30-day death or readmission from hospital discharge in patients with substance misuse. We explored several machine learning algorithms and approaches to integrate information from multiple data sources, such as structured features from a patient's electronic health record (EHR), unstructured clinical notes, socioeconomic data, and emergency medical services (EMS) data. Our gradient-boosted machine model, which combined structured EHR data, socioeconomic status, and EMS data, was the best-performing model (c-statistic 0.746 [95% CI: 0.732-0.759]), outperforming other machine learning methods and structured data source combinations. The addition of unstructured text did not improve performance, suggesting a need for further exploration of how to leverage unstructured data effectively. Feature importance plots highlighted the importance of prior hospital and EMS encounters and discharge disposition in predicting our primary outcome. In conclusion, we integrated multiple data sources that offer complementary information from data sources beyond the typically used EHRs for risk assessment in patients with substance misuse.

Gupta, Preeti; Gruenloh, Tim; Oguss, Madeline; Safipour Afshar, Askar; Spigner, Michael; Churpek, Matthew; Lee, Todd; Mayampurath, Anoop; Afshar, Majid Author Information

Substance misuse poses a significant public health challenge, characterized by premature morbidity and mortality, and heightened healthcare utilization. While studies have demonstrated that previous hospitalizations and emergency department visits are associated with increased mortality in patients with substance misuse, it is unknown whether prior utilization of emergency medical service (EMS) is similarly associated with poor outcomes among this population. The objective of this study is to determine the association between EMS utilization in the 30 days before a hospitalization or emergency department visit and in-hospital outcomes among patients with substance misuse. We conducted a retrospective analysis of adult emergency department visits and hospitalizations (referred to as a hospital encounter) between 2017 and 2021 within the Substance Misuse Data Commons, which maintains electronic health records from substance misuse patients seen at two University of Wisconsin hospitals, linked with state agency, claims, and socioeconomic datasets. Using regression models, we examined the association between EMS use and the outcomes of in-hospital death, hospital length of stay, intensive care unit (ICU) admission, and critical illness events, defined by invasive mechanical ventilation or vasoactive drug administration. Models were adjusted for age, comorbidities, initial severity of illness, substance misuse type, and socioeconomic status. Among 19,402 encounters, individuals with substance misuse who had at least one EMS incident within 30 days of a hospital encounter experienced a higher likelihood of in-hospital mortality (OR 1.52, 95

The duration of low flow prior to initiation of extracorporeal cardiopulmonary resuscitation (eCPR) appears to influence survival. Strategies to reduce the low-flow interval for out-of-hospital cardiac arrest have been focused on expediting patient transport to the hospital or initiating extracorporeal support in the prehospital setting. To date, a direct comparison of low-flow interval between these strategies has not been made. To attempt this comparison, a model was created to predict low-flow intervals for each strategy at different locations across the city of Albuquerque, New Mexico. The data, specific to Albuquerque, suggest that a prehospital cannulation strategy consistently outperforms an expedited transport strategy, with an estimated difference in low-flow interval of 34.3 to 37.2 minutes, depending on location. There is no location within the city in which an expedited transport strategy results in a shorter low-flow interval than prehospital cannulation. It would be rare to successfully initiate eCPR by either strategy in fewer than 30 minutes from the time of patient collapse. Using a prehospital cannulation strategy, the entire coverage area could be eligible for eCPR within 60 minutes of patient collapse. The use of predictive modeling can be a low-cost solution to assist with strategic deployment of prehospital resources and may have potential for real-time decision support for prehospital clinicians.

Introduction: Up to 40% of out-of-hospital cardiac arrest patients will re-arrest in the immediate post-return of spontaneous circulation (post-ROSC) period, and re-arrest is associated with decreased survival. Cardiac arrest guidelines are equivocal regarding what post-ROSC care should be provided in the prehospital setting and when hospital transport should occur. Prehospital protocols must balance the benefit of time-dependent hospital-based care with the risk of early re-arrest. We sought to describe current prehospital protocols for post-ROSC care. Hypothesis: Prehospital protocols for post-ROSC care will be variable. Methods: A single trained abstractor systematically reviewed a purposive sample of prehospital protocols for adult non-traumatic cardiac arrest from the United States using an a priori standardized data abstraction form. Protocols were either stand-alone or integrated into intra-arrest care. Exclusion criteria were non-911 ground transport agencies and protocols not revised since the 2015 guideline update. All protocols were publically available via the Internet. Data abstraction was conducted in May 2019. The number of protocols that met pre-defined criteria were counted and summarized. Results: We identified and reviewed 82 prehospital protocols from 46 states and the District of Columbia. Seven protocols were excluded due to the revision date, leaving 75 protocol included in the study (Table). Six protocols (8%) provide no guidance on prehospital post-ROSC care. ECG acquisition (63/75 [84%]) and transport to specific post-ROSC hospitals (overall 55/75 [73%], but 22/55 [40%] are only if STEMI present) are common but not universal. Only 9 (12%) provide any guidance on when to initiate transport post-ROSC, with 4 (5%) requiring a post-ROSC stabilization period of at least 5 minutes prior to transport. Conclusion: Prehospital treatment and transport protocols for post-ROSC care are highly variable across the United States.