Identifying antinuclear antibody positive individuals at risk for developing systemic autoimmune disease: development and validation of a real-time risk model.

Objective: Positive antinuclear antibodies (ANAs) cause diagnostic dilemmas for clinicians. Currently, no tools exist to help clinicians interpret the significance of a positive ANA in individuals without diagnosed autoimmune diseases. We developed and validated a risk model to predict risk of developing autoimmune disease in positive ANA individuals. Methods: Using a de-identified electronic health record (EHR), we randomly chart reviewed 2,000 positive ANA individuals to determine if a systemic autoimmune disease was diagnosed by a rheumatologist. A priori, we considered demographics, billing codes for autoimmune disease-related symptoms, and laboratory values as variables for the risk model. We performed logistic regression and machine learning models using training and validation samples. Results: We assembled training (n = 1030) and validation (n = 449) sets. Positive ANA individuals who were younger, female, had a higher titer ANA, higher platelet count, disease-specific autoantibodies, and more billing codes related to symptoms of autoimmune diseases were all more likely to develop autoimmune diseases. The most important variables included having a disease-specific autoantibody, number of billing codes for autoimmune disease-related symptoms, and platelet count. In the logistic regression model, AUC was 0.83 (95% CI 0.79-0.86) in the training set and 0.75 (95% CI 0.68-0.81) in the validation set. Conclusion: We developed and validated a risk model that predicts risk for developing systemic autoimmune diseases and can be deployed easily within the EHR. The model can risk stratify positive ANA individuals to ensure high-risk individuals receive urgent rheumatology referrals while reassuring low-risk individuals and reducing unnecessary referrals.

Why do users override alerts? Utilizing large language model to summarize comments and optimize clinical decision support.

OBJECTIVES: To evaluate the capability of using generative artificial intelligence (AI) in summarizing alert comments and to determine if the AI-generated summary could be used to improve clinical decision support (CDS) alerts. MATERIALS AND METHODS: We extracted user comments to alerts generated from September 1, 2022 to September 1, 2023 at Vanderbilt University Medical Center. For a subset of 8 alerts, comment summaries were generated independently by 2 physicians and then separately by GPT-4. We surveyed 5 CDS experts to rate the human-generated and AI-generated summaries on a scale from 1 (strongly disagree) to 5 (strongly agree) for the 4 metrics: clarity, completeness, accuracy, and usefulness. RESULTS: Five CDS experts participated in the survey. A total of 16 human-generated summaries and 8 AI-generated summaries were assessed. Among the top 8 rated summaries, five were generated by GPT-4. AI-generated summaries demonstrated high levels of clarity, accuracy, and usefulness, similar to the human-generated summaries. Moreover, AI-generated summaries exhibited significantly higher completeness and usefulness compared to the human-generated summaries (AI: 3.4 ± 1.2, human: 2.7 ± 1.2, P = .001). CONCLUSION: End-user comments provide clinicians' immediate feedback to CDS alerts and can serve as a direct and valuable data resource for improving CDS delivery. Traditionally, these comments may not be considered in the CDS review process due to their unstructured nature, large volume, and the presence of redundant or irrelevant content. Our study demonstrates that GPT-4 is capable of distilling these comments into summaries characterized by high clarity, accuracy, and completeness. AI-generated summaries are equivalent and potentially better than human-generated summaries. These AI-generated summaries could provide CDS experts with a novel means of reviewing user comments to rapidly optimize CDS alerts both online and offline.

Leveraging large language models for generating responses to patient messages-a subjective analysis.

OBJECTIVE: This study aimed to develop and assess the performance of fine-tuned large language models for generating responses to patient messages sent via an electronic health record patient portal. MATERIALS AND METHODS: Utilizing a dataset of messages and responses extracted from the patient portal at a large academic medical center, we developed a model (CLAIR-Short) based on a pre-trained large language model (LLaMA-65B). In addition, we used the OpenAI API to update physician responses from an open-source dataset into a format with informative paragraphs that offered patient education while emphasizing empathy and professionalism. By combining with this dataset, we further fine-tuned our model (CLAIR-Long). To evaluate fine-tuned models, we used 10 representative patient portal questions in primary care to generate responses. We asked primary care physicians to review generated responses from our models and ChatGPT and rated them for empathy, responsiveness, accuracy, and usefulness. RESULTS: The dataset consisted of 499 794 pairs of patient messages and corresponding responses from the patient portal, with 5000 patient messages and ChatGPT-updated responses from an online platform. Four primary care physicians participated in the survey. CLAIR-Short exhibited the ability to generate concise responses similar to provider's responses. CLAIR-Long responses provided increased patient educational content compared to CLAIR-Short and were rated similarly to ChatGPT's responses, receiving positive evaluations for responsiveness, empathy, and accuracy, while receiving a neutral rating for usefulness. CONCLUSION: This subjective analysis suggests that leveraging large language models to generate responses to patient messages demonstrates significant potential in facilitating communication between patients and healthcare providers.

Leveraging explainable artificial intelligence to optimize clinical decision support.

OBJECTIVE: To develop and evaluate a data-driven process to generate suggestions for improving alert criteria using explainable artificial intelligence (XAI) approaches. METHODS: We extracted data on alerts generated from January 1, 2019 to December 31, 2020, at Vanderbilt University Medical Center. We developed machine learning models to predict user responses to alerts. We applied XAI techniques to generate global explanations and local explanations. We evaluated the generated suggestions by comparing with alert's historical change logs and stakeholder interviews. Suggestions that either matched (or partially matched) changes already made to the alert or were considered clinically correct were classified as helpful. RESULTS: The final dataset included 2 991 823 firings with 2689 features. Among the 5 machine learning models, the LightGBM model achieved the highest Area under the ROC Curve: 0.919 [0.918, 0.920]. We identified 96 helpful suggestions. A total of 278 807 firings (9.3%) could have been eliminated. Some of the suggestions also revealed workflow and education issues. CONCLUSION: We developed a data-driven process to generate suggestions for improving alert criteria using XAI techniques. Our approach could identify improvements regarding clinical decision support (CDS) that might be overlooked or delayed in manual reviews. It also unveils a secondary purpose for the XAI: to improve quality by discovering scenarios where CDS alerts are not accepted due to workflow, education, or staffing issues.

Development and Validation of a Machine Learning Algorithm Using Clinical Pages to Predict Imminent Clinical Deterioration.

BACKGROUND: Early detection of clinical deterioration among hospitalized patients is a clinical priority for patient safety and quality of care. Current automated approaches for identifying these patients perform poorly at identifying imminent events. OBJECTIVE: Develop a machine learning algorithm using pager messages sent between clinical team members to predict imminent clinical deterioration. DESIGN: We conducted a large observational study using long short-term memory machine learning models on the content and frequency of clinical pages. PARTICIPANTS: We included all hospitalizations between January 1, 2018 and December 31, 2020 at Vanderbilt University Medical Center that included at least one page message to physicians. Exclusion criteria included patients receiving palliative care, hospitalizations with a planned intensive care stay, and hospitalizations in the top 2% longest length of stay. MAIN MEASURES: Model classification performance to identify in-hospital cardiac arrest, transfer to intensive care, or Rapid Response activation in the next 3-, 6-, and 12-hours. We compared model performance against three common early warning scores: Modified Early Warning Score, National Early Warning Score, and the Epic Deterioration Index. KEY RESULTS: There were 87,783 patients (mean [SD] age 54.0 [18.8] years; 45,835 [52.2%] women) who experienced 136,778 hospitalizations. 6214 hospitalized patients experienced a deterioration event. The machine learning model accurately identified 62% of deterioration events within 3-hours prior to the event and 47% of events within 12-hours. Across each time horizon, the model surpassed performance of the best early warning score including area under the receiver operating characteristic curve at 6-hours (0.856 vs. 0.781), sensitivity at 6-hours (0.590 vs. 0.505), specificity at 6-hours (0.900 vs. 0.878), and F-score at 6-hours (0.291 vs. 0.220). CONCLUSIONS: Machine learning applied to the content and frequency of clinical pages improves prediction of imminent deterioration. Using clinical pages to monitor patient acuity supports improved detection of imminent deterioration without requiring changes to clinical workflow or nursing documentation.

POINT: Pipeline for Offline Conversion and Integration of Geocodes and Neighborhood Data.

OBJECTIVES: Geocoding, the process of converting addresses into precise geographic coordinates, allows researchers and health systems to obtain neighborhood-level estimates of social determinants of health. This information supports opportunities to personalize care and interventions for individual patients based on the environments where they live. We developed an integrated offline geocoding pipeline to streamline the process of obtaining address-based variables, which can be integrated into existing data processing pipelines. METHODS: POINT is a web-based, containerized, application for geocoding addresses that can be deployed offline and made available to multiple users across an organization. Our application supports use through both a graphical user interface and application programming interface to query geographic variables, by census tract, without exposing sensitive patient data. We evaluated our application's performance using two datasets: one consisting of 1 million nationally representative addresses sampled from Open Addresses, and the other consisting of 3,096 previously geocoded patient addresses. RESULTS: A total of 99.4 and 99.8% of addresses in the Open Addresses and patient addresses datasets, respectively, were geocoded successfully. Census tract assignment was concordant with reference in greater than 90% of addresses for both datasets. Among successful geocodes, median (interquartile range) distances from reference coordinates were 52.5 (26.5-119.4) and 14.5 (10.9-24.6) m for the two datasets. CONCLUSION: POINT successfully geocodes more addresses and yields similar accuracy to existing solutions, including the U.S. Census Bureau's official geocoder. Addresses are considered protected health information and cannot be shared with common online geocoding services. POINT is an offline solution that enables scalability to multiple users and integrates downstream mapping to neighborhood-level variables with a pipeline that allows users to incorporate additional datasets as they become available. As health systems and researchers continue to explore and improve health equity, it is essential to quickly and accurately obtain neighborhood variables in a Health Insurance Portability and Accountability Act (HIPAA)-compliant way.

Leveraging natural language processing to identify eligible lung cancer screening patients with the electronic health record.

OBJECTIVE: To develop and validate an approach that identifies patients eligible for lung cancer screening (LCS) by combining structured and unstructured smoking data from the electronic health record (EHR). METHODS: We identified patients aged 50-80 years who had at least one encounter in a primary care clinic at Vanderbilt University Medical Center (VUMC) between 2019 and 2022. We fine-tuned an existing natural language processing (NLP) tool to extract quantitative smoking information using clinical notes collected from VUMC. Then, we developed an approach to identify patients who are eligible for LCS by combining smoking information from structured data and clinical narratives. We compared this method with two approaches to identify LCS eligibility only using smoking information from structured EHR. We used 50 patients with a documented history of tobacco use for comparison and validation. RESULTS: 102,475 patients were included. The NLP-based approach achieved an F1-score of 0.909, and accuracy of 0.96. The baseline approach could identify 5,887 patients. Compared to the baseline approach, the number of identified patients using all structured data and the NLP-based algorithm was 7,194 (22.2 %) and 10,231 (73.8 %), respectively. The NLP-based approach identified 589 Black/African Americans, a significant increase of 119 %. CONCLUSION: We present a feasible NLP-based approach to identify LCS eligible patients. It provides a technical basis for the development of clinical decision support tools to potentially improve the utilization of LCS and diminish healthcare disparities.

Leveraging Large Language Models for Generating Responses to Patient Messages.

Objective: This study aimed to develop and assess the performance of fine-tuned large language models for generating responses to patient messages sent via an electronic health record patient portal. Methods: Utilizing a dataset of messages and responses extracted from the patient portal at a large academic medical center, we developed a model (CLAIR-Short) based on a pre-trained large language model (LLaMA-65B). In addition, we used the OpenAI API to update physician responses from an open-source dataset into a format with informative paragraphs that offered patient education while emphasizing empathy and professionalism. By combining with this dataset, we further fine-tuned our model (CLAIR-Long). To evaluate the fine-tuned models, we used ten representative patient portal questions in primary care to generate responses. We asked primary care physicians to review generated responses from our models and ChatGPT and rated them for empathy, responsiveness, accuracy, and usefulness. Results: The dataset consisted of a total of 499,794 pairs of patient messages and corresponding responses from the patient portal, with 5,000 patient messages and ChatGPT-updated responses from an online platform. Four primary care physicians participated in the survey. CLAIR-Short exhibited the ability to generate concise responses similar to provider's responses. CLAIR-Long responses provided increased patient educational content compared to CLAIR-Short and were rated similarly to ChatGPT's responses, receiving positive evaluations for responsiveness, empathy, and accuracy, while receiving a neutral rating for usefulness. Conclusion: Leveraging large language models to generate responses to patient messages demonstrates significant potential in facilitating communication between patients and primary care providers.

The Impact of Clinical Decision Support on Health Disparities and the Digital Divide.

OBJECTIVES: This literature review summarizes relevant studies from the last three years (2020-2022) related to clinical decision support (CDS) and CDS impact on health disparities and the digital divide. This survey identifies current trends and synthesizes evidence-based recommendations and considerations for future development and implementation of CDS tools. METHODS: We conducted a search in PubMed for literature published between 2020 and 2022. Our search strategy was constructed as a combination of the MEDLINE®/PubMed® Health Disparities and Minority Health Search Strategy and relevant CDS MeSH terms and phrases. We then extracted relevant data from the studies, including priority population when applicable, domain of influence on the disparity being addressed, and the type of CDS being used. We also made note of when a study discussed the digital divide in some capacity and organized the comments into general themes through group discussion. RESULTS: Our search yielded 520 studies, with 45 included at the conclusion of screening. The most frequent CDS type in this review was point-of-care alerts/reminders (33.3%). Health Care System was the most frequent domain of influence (71.1%), and Blacks/African Americans were the most frequently included priority population (42.2%). Throughout the literature, we found four general themes related to the technology divide: inaccessibility of technology, access to care, trust of technology, and technology literacy.This survey revealed the diversity of CDS being used to address health disparities and several barriers which may make CDS less effective or potentially harmful to certain populations. Regular examinations of literature that feature CDS and address health disparities can help to reveal new strategies and patterns for improving healthcare.

Rates and Correlates of Uptake of Continuous Glucose Monitors Among Adults with Type 2 Diabetes in Primary Care and Endocrinology Settings.

BACKGROUND: Clinical trials indicate continuous glucose monitor (CGM) use may benefit adults with type 2 diabetes, but CGM rates and correlates in real-world care settings are unknown. OBJECTIVE: We sought to ascertain prevalence and correlates of CGM use and to examine rates of new CGM prescriptions across clinic types and medication regimens. DESIGN: Retrospective cohort using electronic health records in a large academic medical center in the Southeastern US. PARTICIPANTS: Adults with type 2 diabetes and a primary care or endocrinology visit during 2021. MAIN MEASURES: Age, gender, race, ethnicity, insurance, clinic type, insulin regimen, hemoglobin A1c values, CGM prescriptions, and prescribing clinic type. KEY RESULTS: Among 30,585 adults with type 2 diabetes, 13% had used a CGM. CGM users were younger and more had private health insurance (p < .05) as compared to non-users; 72% of CGM users had an intensive insulin regimen, but 12% were not taking insulin. CGM users had higher hemoglobin A1c values (both most recent and most proximal to the first CGM prescription) than non-users. CGM users were more likely to receive endocrinology care than non-users, but 23% had only primary care visits in 2021. For each month in 2021, a mean of 90.5 (SD 12.5) people started using CGM. From 2020 to 2021, monthly rates of CGM prescriptions to new users grew 36% overall, but 125% in primary care. Most starting CGM in endocrinology had an intensive insulin regimen (82% vs. 49% starting in primary care), whereas 28% starting CGM in primary care were not using insulin (vs. 5% in endocrinology). CONCLUSION: CGM uptake for type 2 diabetes is increasing rapidly, with most growth in primary care. These trends present opportunities for healthcare system adaptations to support CGM use and related workflows in primary care to support growth in uptake.

Using AI-generated suggestions from ChatGPT to optimize clinical decision support.

OBJECTIVE: To determine if ChatGPT can generate useful suggestions for improving clinical decision support (CDS) logic and to assess noninferiority compared to human-generated suggestions. METHODS: We supplied summaries of CDS logic to ChatGPT, an artificial intelligence (AI) tool for question answering that uses a large language model, and asked it to generate suggestions. We asked human clinician reviewers to review the AI-generated suggestions as well as human-generated suggestions for improving the same CDS alerts, and rate the suggestions for their usefulness, acceptance, relevance, understanding, workflow, bias, inversion, and redundancy. RESULTS: Five clinicians analyzed 36 AI-generated suggestions and 29 human-generated suggestions for 7 alerts. Of the 20 suggestions that scored highest in the survey, 9 were generated by ChatGPT. The suggestions generated by AI were found to offer unique perspectives and were evaluated as highly understandable and relevant, with moderate usefulness, low acceptance, bias, inversion, redundancy. CONCLUSION: AI-generated suggestions could be an important complementary part of optimizing CDS alerts, can identify potential improvements to alert logic and support their implementation, and may even be able to assist experts in formulating their own suggestions for CDS improvement. ChatGPT shows great potential for using large language models and reinforcement learning from human feedback to improve CDS alert logic and potentially other medical areas involving complex, clinical logic, a key step in the development of an advanced learning health system.

Assessing the Value of ChatGPT for Clinical Decision Support Optimization.

Objective: To determine if ChatGPT can generate useful suggestions for improving clinical decision support (CDS) logic and to assess noninferiority compared to human-generated suggestions. Methods: We supplied summaries of CDS logic to ChatGPT, an artificial intelligence (AI) tool for question answering that uses a large language model, and asked it to generate suggestions. We asked human clinician reviewers to review the AI-generated suggestions as well as human-generated suggestions for improving the same CDS alerts, and rate the suggestions for their usefulness, acceptance, relevance, understanding, workflow, bias, inversion, and redundancy. Results: Five clinicians analyzed 36 AI-generated suggestions and 29 human-generated suggestions for 7 alerts. Of the 20 suggestions that scored highest in the survey, 9 were generated by ChatGPT. The suggestions generated by AI were found to offer unique perspectives and were evaluated as highly understandable and relevant, with moderate usefulness, low acceptance, bias, inversion, redundancy. Conclusion: AI-generated suggestions could be an important complementary part of optimizing CDS alerts, can identify potential improvements to alert logic and support their implementation, and may even be able to assist experts in formulating their own suggestions for CDS improvement. ChatGPT shows great potential for using large language models and reinforcement learning from human feedback to improve CDS alert logic and potentially other medical areas involving complex, clinical logic, a key step in the development of an advanced learning health system.

Beyond chest pain: Incremental value of other variables to identify patients for an early ECG.

BACKGROUND: Chest pain (CP) is the hallmark symptom for acute coronary syndrome (ACS) but is not reported in 20-30% of patients, especially women, elderly, non-white patients, presenting to the emergency department (ED) with an ST-segment elevation myocardial infarction (STEMI). METHODS: We used a retrospective 5-year adult ED sample of 279,132 patients to explore using CP alone to predict ACS, then we incrementally added other ACS chief complaints, age, and sex in a series of multivariable logistic regression models. We evaluated each model's identification of ACS and STEMI. RESULTS: Using CP alone would recommend ECGs for 8% of patients (sensitivity, 61%; specificity, 92%) but missed 28.4% of STEMIs. The model with all variables identified ECGs for 22% of patients (sensitivity, 82%; specificity, 78%) but missed 14.7% of STEMIs. The model with CP and other ACS chief complaints had the highest sensitivity (93%) and specificity (55%), identified 45.1% of patients for ECG, and only missed 4.4% of STEMIs. CONCLUSION: CP alone had highest specificity but lacked sensitivity. Adding other ACS chief complaints increased sensitivity but identified 2.2-fold more patients for ECGs. Achieving an ECG in 10 min for patients with ACS to identify all STEMIs will be challenging without introducing more complex risk calculation into clinical care.

A patient education intervention improved rates of successful video visits during rapid implementation of telehealth.

INTRODUCTION: The need to rapidly implement telehealth at large scale during the COVID-19 pandemic led to many patients using telehealth for the first time. We assessed the effect of structured pre-visit preparatory telephone calls on success of telehealth visits and examined risk factors for unsuccessful visits. METHODS: A retrospective cohort study was carried out of 45,803 adult patients scheduled for a total of 64,447 telehealth appointments between March and July 2020 at an academic medical center. A subset of patients received a structured pre-visit phone call. Demographic factors and inclusion of a pre-visit call were analysed by logistic regression. Primary outcomes were non-completion of any visit and completion of phone-only versus audio-visual telehealth visits. RESULTS: A pre-visit telephone call to a subset of patients significantly increased the likelihood of a successful telehealth visit (OR 0.54; 95% CI: 0.48-0.60). Patients aged 18-30 years, those with non-commercial insurance or those of Black race were more likely to have incomplete visits. Compared to age 18-30, increasing age increased likelihood of a failed video visit: 31-50 years (OR 1.31; 95% CI: 1.13-1.51), 51-70 years (OR 2.98; 2.60-3.42) and >70 years (OR 4.16; 3.58-4.82). Those with non-commercial insurance and those of Black race (OR 1.8; 95% CI 1.67-1.92) were more likely to have a failed video visit. DISCUSSION: A structured pre-call to patients improved the likelihood of a successful video visit during widespread adoption of telehealth. Structured pre-calls to patients may be an important tool to help reduce gaps in utilization among groups.

New onset delirium prediction using machine learning and long short-term memory (LSTM) in electronic health record.

OBJECTIVE: To develop and test an accurate deep learning model for predicting new onset delirium in hospitalized adult patients. METHODS: Using electronic health record (EHR) data extracted from a large academic medical center, we developed a model combining long short-term memory (LSTM) and machine learning to predict new onset delirium and compared its performance with machine-learning-only models (logistic regression, random forest, support vector machine, neural network, and LightGBM). The labels of models were confusion assessment method (CAM) assessments. We evaluated models on a hold-out dataset. We calculated Shapley additive explanations (SHAP) measures to gauge the feature impact on the model. RESULTS: A total of 331 489 CAM assessments with 896 features from 34 035 patients were included. The LightGBM model achieved the best performance (AUC 0.927 [0.924, 0.929] and F1 0.626 [0.618, 0.634]) among the machine learning models. When combined with the LSTM model, the final model's performance improved significantly (P = .001) with AUC 0.952 [0.950, 0.955] and F1 0.759 [0.755, 0.765]. The precision value of the combined model improved from 0.497 to 0.751 with a fixed recall of 0.8. Using the mean absolute SHAP values, we identified the top 20 features, including age, heart rate, Richmond Agitation-Sedation Scale score, Morse fall risk score, pulse, respiratory rate, and level of care. CONCLUSION: Leveraging LSTM to capture temporal trends and combining it with the LightGBM model can significantly improve the prediction of new onset delirium, providing an algorithmic basis for the subsequent development of clinical decision support tools for proactive delirium interventions.

Real-Time User Feedback to Support Clinical Decision Support System Improvement.

OBJECTIVES: To improve clinical decision support (CDS) by allowing users to provide real-time feedback when they interact with CDS tools and by creating processes for responding to and acting on this feedback. METHODS: Two organizations implemented similar real-time feedback tools and processes in their electronic health record and gathered data over a 30-month period. At both sites, users could provide feedback by using Likert feedback links embedded in all end-user facing alerts, with results stored outside the electronic health record, and provide feedback as a comment when they overrode an alert. Both systems are monitored daily by clinical informatics teams. RESULTS: The two sites received 2,639 Likert feedback comments and 623,270 override comments over a 30-month period. Through four case studies, we describe our use of end-user feedback to rapidly respond to build errors, as well as identifying inaccurate knowledge management, user-interface issues, and unique workflows. CONCLUSION: Feedback on CDS tools can be solicited in multiple ways, and it contains valuable and actionable suggestions to improve CDS alerts. Additionally, end users appreciate knowing their feedback is being received and may also make other suggestions to improve the electronic health record. Incorporation of end-user feedback into CDS monitoring, evaluation, and remediation is a way to improve CDS.

Decreasing pre-procedural fasting times in hospitalized children.

OBJECTIVE: Prolonged pre-procedural fasting in children is associated with decreased patient and family satisfaction and increased patient hemodynamic instability. Practice guidelines recommend clear liquid fasting times of 2 h. We aimed to decrease pre-procedural clear liquid fasting time from 10 h 13 min to 5 h for pediatric hospital medicine (PHM) patients. METHODS: All children admitted to the PHM service at a quaternary care children's hospital with an NPO (nil per os) order associated with a procedure requiring general anesthesia or sedation from November 2, 2017 to September 19, 2021 were included. The primary outcome measure was the average time from clear liquid fasting end time to anesthesia start time. The process measure was the percent of NPO orders including a documented clear liquid fasting end time. Balancing measures were aspiration events and case delays/cancellations. Statistical process control charts were used to analyze outcomes. RESULTS: Shortly after implementation of a SmartPhrase in the NPO order, there was special cause variation resulting in a centerline shift from a mean of 10 h 13 min to 6 h 37 min and an increase in the process measure from a baseline of 2%-52%. Following implementation of a hospital-wide change to the NPO order format, another centerline shift to 6 h 7 min occurred which has been sustained for 6 months. No aspiration events and four NPO violations occurred during the intervention period. CONCLUSION: Quality improvement methodology and higher reliability interventions safely decreased the average pre-procedural fasting time in hospitalized children.

Integration of Face-to-Face Screening With Real-time Machine Learning to Predict Risk of Suicide Among Adults.

Importance: Understanding the differences and potential synergies between traditional clinician assessment and automated machine learning might enable more accurate and useful suicide risk detection. Objective: To evaluate the respective and combined abilities of a real-time machine learning model and the Columbia Suicide Severity Rating Scale (C-SSRS) to predict suicide attempt (SA) and suicidal ideation (SI). Design, Setting, and Participants: This cohort study included encounters with adult patients (aged ≥18 years) at a major academic medical center. The C-SSRS was administered during routine care, and a Vanderbilt Suicide Attempt and Ideation Likelihood (VSAIL) prediction was generated in the electronic health record. Encounters took place in the inpatient, ambulatory surgical, and emergency department settings. Data were collected from June 2019 to September 2020. Main Outcomes and Measures: Primary outcomes were the incidence of SA and SI, encoded as International Classification of Diseases codes, occurring within various time periods after an index visit. We evaluated the retrospective validity of the C-SSRS, VSAIL, and ensemble models combining both. Discrimination metrics included area under the receiver operating curve (AUROC), area under the precision-recall curve (AUPR), sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Results: The cohort included 120 398 unique index visits for 83 394 patients (mean [SD] age, 51.2 [20.6] years; 38 107 [46%] men; 45 273 [54%] women; 13 644 [16%] Black; 63 869 [77%] White). Within 30 days of an index visit, the combined models had higher AUROC (SA: 0.874-0.887; SI: 0.869-0.879) than both the VSAIL (SA: 0.729; SI: 0.773) and C-SSRS (SA: 0.823; SI: 0.777) models. In the highest risk-decile, ensemble methods had PPV of 1.3% to 1.4% for SA and 8.3% to 8.7% for SI and sensitivity of 77.6% to 79.5% for SA and 67.4% to 70.1% for SI, outperforming VSAIL (PPV for SA: 0.4%; PPV for SI: 3.9%; sensitivity for SA: 28.8%; sensitivity for SI: 35.1%) and C-SSRS (PPV for SA: 0.5%; PPV for SI: 3.5%; sensitivity for SA: 76.6%; sensitivity for SI: 68.8%). Conclusions and Relevance: In this study, suicide risk prediction was optimal when leveraging both in-person screening (for acute measures of risk in patient-reported suicidality) and historical EHR data (for underlying clinical factors that can quantify a patient's passive risk level). To improve suicide risk classification, prediction systems could combine pretrained machine learning with structured clinician assessment without needing to retrain the original model.

Clinical Decision Support Stewardship: Best Practices and Techniques to Monitor and Improve Interruptive Alerts.

Interruptive clinical decision support systems, both within and outside of electronic health records, are a resource that should be used sparingly and monitored closely. Excessive use of interruptive alerting can quickly lead to alert fatigue and decreased effectiveness and ignoring of alerts. In this review, we discuss the evidence for effective alert stewardship as well as practices and methods we have found useful to assess interruptive alert burden, reduce excessive firings, optimize alert effectiveness, and establish quality governance at our institutions. We also discuss the importance of a holistic view of the alerting ecosystem beyond the electronic health record.