Faculty Bibliography

OBJECTIVES/OBJECTIVE:The study aimed to assess the usage and impact of a private and secure instance of a generative artificial intelligence (GenAI) application in a large academic health center. The goal was to understand how employees interact with this technology and the influence on their perception of skill and work performance. MATERIALS AND METHODS/METHODS:New York University Langone Health (NYULH) established a secure, private, and managed Azure OpenAI service (GenAI Studio) and granted widespread access to employees. Usage was monitored and users were surveyed about their experiences. RESULTS:Over 6 months, over 1007 individuals applied for access, with high usage among research and clinical departments. Users felt prepared to use the GenAI studio, found it easy to use, and would recommend it to a colleague. Users employed the GenAI studio for diverse tasks such as writing, editing, summarizing, data analysis, and idea generation. Challenges included difficulties in educating the workforce in constructing effective prompts and token and API limitations. DISCUSSION/CONCLUSIONS:The study demonstrated high interest in and extensive use of GenAI in a healthcare setting, with users employing the technology for diverse tasks. While users identified several challenges, they also recognized the potential of GenAI and indicated a need for more instruction and guidance on effective usage. CONCLUSION/CONCLUSIONS:The private GenAI studio provided a useful tool for employees to augment their skills and apply GenAI to their daily tasks. The study underscored the importance of workforce education when implementing system-wide GenAI and provided insights into its strengths and weaknesses.

OBJECTIVES/OBJECTIVE:To evaluate the proficiency of a HIPAA-compliant version of GPT-4 in identifying actionable, incidental findings from unstructured radiology reports of Emergency Department patients. To assess appropriateness of artificial intelligence (AI)-generated, patient-facing summaries of these findings. MATERIALS AND METHODS/METHODS:Radiology reports extracted from the electronic health record of a large academic medical center were manually reviewed to identify non-emergent, incidental findings with high likelihood of requiring follow-up, further sub-stratified as "definitely actionable" (DA) or "possibly actionable-clinical correlation" (PA-CC). Instruction prompts to GPT-4 were developed and iteratively optimized using a validation set of 50 reports. The optimized prompt was then applied to a test set of 430 unseen reports. GPT-4 performance was primarily graded on accuracy identifying either DA or PA-CC findings, then secondarily for DA findings alone. Outputs were reviewed for hallucinations. AI-generated patient-facing summaries were assessed for appropriateness via Likert scale. RESULTS:For the primary outcome (DA or PA-CC), GPT-4 achieved 99.3% recall, 73.6% precision, and 84.5% F-1. For the secondary outcome (DA only), GPT-4 demonstrated 95.2% recall, 77.3% precision, and 85.3% F-1. No findings were "hallucinated" outright. However, 2.8% of cases included generated text about recommendations that were inferred without specific reference. The majority of True Positive AI-generated summaries required no or minor revision. CONCLUSION/CONCLUSIONS:GPT-4 demonstrates proficiency in detecting actionable, incidental findings after refined instruction prompting. AI-generated patient instructions were most often appropriate, but rarely included inferred recommendations. While this technology shows promise to augment diagnostics, active clinician oversight via "human-in-the-loop" workflows remains critical for clinical implementation.

BACKGROUND:Healthcare crowdsourcing events (e.g. hackathons) facilitate interdisciplinary collaboration and encourage innovation. Peer-reviewed research has not yet considered a healthcare crowdsourcing event focusing on generative artificial intelligence (GenAI), which generates text in response to detailed prompts and has vast potential for improving the efficiency of healthcare organizations. Our event, the New York University Langone Health (NYULH) Prompt-a-thon, primarily sought to inspire and build AI fluency within our diverse NYULH community, and foster collaboration and innovation. Secondarily, we sought to analyze how participants' experience was influenced by their prior GenAI exposure and whether they received sample prompts during the workshop. METHODS:Executing the event required the assembly of an expert planning committee, who recruited diverse participants, anticipated technological challenges, and prepared the event. The event was composed of didactics and workshop sessions, which educated and allowed participants to experiment with using GenAI on real healthcare data. Participants were given novel "project cards" associated with each dataset that illuminated the tasks GenAI could perform and, for a random set of teams, sample prompts to help them achieve each task (the public repository of project cards can be found at https://github.com/smallw03/NYULH-Generative-AI-Prompt-a-thon-Project-Cards). Afterwards, participants were asked to fill out a survey with 7-point Likert-style questions. RESULTS:Our event was successful in educating and inspiring hundreds of enthusiastic in-person and virtual participants across our organization on the responsible use of GenAI in a low-cost and technologically feasible manner. All participants responded positively, on average, to each of the survey questions (e.g., confidence in their ability to use and trust GenAI). Critically, participants reported a self-perceived increase in their likelihood of using and promoting colleagues' use of GenAI for their daily work. No significant differences were seen in the surveys of those who received sample prompts with their project task descriptions. CONCLUSION/CONCLUSIONS:The first healthcare Prompt-a-thon was an overwhelming success, with minimal technological failures, positive responses from diverse participants and staff, and evidence of post-event engagement. These findings will be integral to planning future events at our institution, and to others looking to engage their workforce in utilizing GenAI.

IMPORTANCE/UNASSIGNED:By law, patients have immediate access to discharge notes in their medical records. Technical language and abbreviations make notes difficult to read and understand for a typical patient. Large language models (LLMs [eg, GPT-4]) have the potential to transform these notes into patient-friendly language and format. OBJECTIVE/UNASSIGNED:To determine whether an LLM can transform discharge summaries into a format that is more readable and understandable. DESIGN, SETTING, AND PARTICIPANTS/UNASSIGNED:This cross-sectional study evaluated a sample of the discharge summaries of adult patients discharged from the General Internal Medicine service at NYU (New York University) Langone Health from June 1 to 30, 2023. Patients discharged as deceased were excluded. All discharge summaries were processed by the LLM between July 26 and August 5, 2023. INTERVENTIONS/UNASSIGNED:A secure Health Insurance Portability and Accountability Act-compliant platform, Microsoft Azure OpenAI, was used to transform these discharge summaries into a patient-friendly format between July 26 and August 5, 2023. MAIN OUTCOMES AND MEASURES/UNASSIGNED:Outcomes included readability as measured by Flesch-Kincaid Grade Level and understandability using Patient Education Materials Assessment Tool (PEMAT) scores. Readability and understandability of the original discharge summaries were compared with the transformed, patient-friendly discharge summaries created through the LLM. As balancing metrics, accuracy and completeness of the patient-friendly version were measured. RESULTS/UNASSIGNED:Discharge summaries of 50 patients (31 female [62.0%] and 19 male [38.0%]) were included. The median patient age was 65.5 (IQR, 59.0-77.5) years. Mean (SD) Flesch-Kincaid Grade Level was significantly lower in the patient-friendly discharge summaries (6.2 [0.5] vs 11.0 [1.5]; P < .001). PEMAT understandability scores were significantly higher for patient-friendly discharge summaries (81% vs 13%; P < .001). Two physicians reviewed each patient-friendly discharge summary for accuracy on a 6-point scale, with 54 of 100 reviews (54.0%) giving the best possible rating of 6. Summaries were rated entirely complete in 56 reviews (56.0%). Eighteen reviews noted safety concerns, mostly involving omissions, but also several inaccurate statements (termed hallucinations). CONCLUSIONS AND RELEVANCE/UNASSIGNED:The findings of this cross-sectional study of 50 discharge summaries suggest that LLMs can be used to translate discharge summaries into patient-friendly language and formats that are significantly more readable and understandable than discharge summaries as they appear in electronic health records. However, implementation will require improvements in accuracy, completeness, and safety. Given the safety concerns, initial implementation will require physician review.

Background: The introduction of electronic workflows has allowed for the flow of raw uncontextualized clinical data into medical documentation. As a result, many electronic notes have become replete of "noise" and deplete clinically significant "signals." There is an urgent need to develop and implement innovative approaches in electronic clinical documentation that improve note quality and reduce unnecessary bloating. Objective: This study aims to describe the development and impact of a novel set of templates designed to change the flow of information in medical documentation. Methods: This is a multihospital nonrandomized prospective improvement study conducted on the inpatient general internal medicine service across 3 hospital campuses at the New York University Langone Health System. A group of physician leaders representing each campus met biweekly for 6 months. The output of these meetings included (1) a conceptualization of the note bloat problem as a dysfunction in information flow, (2) a set of guiding principles for organizational documentation improvement, (3) the design and build of novel electronic templates that reduced the flow of extraneous information into provider notes by providing link outs to best practice data visualizations, and (4) a documentation improvement curriculum for inpatient medicine providers. Prior to go-live, pragmatic usability testing was performed with the new progress note template, and the overall user experience was measured using the System Usability Scale (SUS). Primary outcome measures after go-live include template utilization rate and note length in characters. Results: In usability testing among 22 medicine providers, the new progress note template averaged a usability score of 90.6 out of 100 on the SUS. A total of 77% (17/22) of providers strongly agreed that the new template was easy to use, and 64% (14/22) strongly agreed that they would like to use the template frequently. In the 3 months after template implementation, general internal medicine providers wrote 67% (51,431/76,647) of all inpatient notes with the new templates. During this period, the organization saw a 46% (2768/6191), 47% (3505/7819), and 32% (3427/11,226) reduction in note length for general medicine progress notes, consults, and history and physical notes, respectively, when compared to a baseline measurement period prior to interventions. Conclusions: A bundled intervention that included the deployment of novel templates for inpatient general medicine providers significantly reduced average note length on the clinical service. Templates designed to reduce the flow of extraneous information into provider notes performed well during usability testing, and these templates were rapidly adopted across all hospital campuses. Further research is needed to assess the impact of novel templates on note quality, provider efficiency, and patient outcomes.

BACKGROUND: We previously developed and validated a predictive model to help clinicians identify hospitalized adults with coronavirus disease 2019 (COVID-19) who may be ready for discharge given their low risk of adverse events. Whether this algorithm can prompt more timely discharge for stable patients in practice is unknown. OBJECTIVES/OBJECTIVE: The aim of the study is to estimate the effect of displaying risk scores on length of stay (LOS). METHODS: We integrated model output into the electronic health record (EHR) at four hospitals in one health system by displaying a green/orange/red score indicating low/moderate/high-risk in a patient list column and a larger COVID-19 summary report visible for each patient. Display of the score was pseudo-randomized 1:1 into intervention and control arms using a patient identifier passed to the model execution code. Intervention effect was assessed by comparing LOS between intervention and control groups. Adverse safety outcomes of death, hospice, and re-presentation were tested separately and as a composite indicator. We tracked adoption and sustained use through daily counts of score displays. RESULTS: Enrolling 1,010 patients from May 15, 2020 to December 7, 2020, the trial found no detectable difference in LOS. The intervention had no impact on safety indicators of death, hospice or re-presentation after discharge. The scores were displayed consistently throughout the study period but the study lacks a causally linked process measure of provider actions based on the score. Secondary analysis revealed complex dynamics in LOS temporally, by primary symptom, and hospital location. CONCLUSION/CONCLUSIONS: An AI-based COVID-19 risk score displayed passively to clinicians during routine care of hospitalized adults with COVID-19 was safe but had no detectable impact on LOS. Health technology challenges such as insufficient adoption, nonuniform use, and provider trust compounded with temporal factors of the COVID-19 pandemic may have contributed to the null result. TRIAL REGISTRATION/BACKGROUND: ClinicalTrials.gov identifier: NCT04570488.

Predictive models may be particularly beneficial to clinicians when they face uncertainty and seek to develop a mental model of disease progression, but we know little about the post-implementation effects of predictive models on clinicians' experience of their work. Combining survey and interview methods, we found that providers using a predictive algorithm reported being significantly less uncertain and better able to anticipate, plan and prepare for patient discharge than non-users. The tool helped hospitalists form and develop confidence in their mental models of a novel disease (Covid-19). Yet providers' attention to the predictive tool declined as their confidence in their own mental models grew. Predictive algorithms that not only offer data but also provide feedback on decisions, thus supporting providers' motivation for continuous learning, hold promise for more sustained provider attention and cognition augmentation.

BACKGROUND:Clinical decision support (CDS) is a valuable feature of electronic health records (EHRs) designed to improve quality and safety. However, due to the complexities of system design and inconsistent results, CDS tools may inadvertently increase alert fatigue and contribute to physician burnout. A/B testing, or rapid-cycle randomized tests, is a useful method that can be applied to the EHR in order to rapidly understand and iteratively improve design choices embedded within CDS tools. OBJECTIVE:This paper describes how rapid randomized controlled trials (RCTs) embedded within EHRs can be used to quickly ascertain the superiority of potential CDS design changes to improve their usability, reduce alert fatigue, and promote quality of care. METHODS:A multistep process combining tools from user-centered design, A/B testing, and implementation science was used to understand, ideate, prototype, test, analyze, and improve each candidate CDS. CDS engagement metrics (alert views, acceptance rates) were used to evaluate which CDS version is superior. RESULTS:To demonstrate the impact of the process, 2 experiments are highlighted. First, after multiple rounds of usability testing, a revised CDS influenza alert was tested against usual care CDS in a rapid (~6 weeks) RCT. The new alert text resulted in minimal impact on reducing firings per patients per day, but this failure triggered another round of review that identified key technical improvements (ie, removal of dismissal button and firings in procedural areas) that led to a dramatic decrease in firings per patient per day (23.1 to 7.3). In the second experiment, the process was used to test 3 versions (financial, quality, regulatory) of text supporting tobacco cessation alerts as well as 3 supporting images. Based on 3 rounds of RCTs, there was no significant difference in acceptance rates based on the framing of the messages or addition of images. CONCLUSIONS:These experiments support the potential for this new process to rapidly develop, deploy, and rigorously evaluate CDS within an EHR. We also identified important considerations in applying these methods. This approach may be an important tool for improving the impact of and experience with CDS. TRIAL REGISTRATION/BACKGROUND:Flu alert trial: ClinicalTrials.gov NCT03415425; https://clinicaltrials.gov/ct2/show/NCT03415425. Tobacco alert trial: ClinicalTrials.gov NCT03714191; https://clinicaltrials.gov/ct2/show/NCT03714191.