Lawrence M. Lewis

Lawrence M. Lewis

Auto-generated profile, based on publicly available information

Lawrence M. Lewis

Department of Emergency Medicine, Washington University

Research Focus & Metrics

Academic Career Length

15Years (2012-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
25 May 2012 in Academic Emergency Medicine
Verified
Journal Article
Open access
PDF file
DHDavid A. Horwitz
David A. Horwitz
ESEvan S. Schwarz
Evan S. Schwarz
MSMitchell G. Scott
Mitchell G. Scott
LLLawrence M. Lewis
Lawrence M. Lewis
ACADEMIC EMERGENCY MEDICINE 2012; 19:00–00 © 2012 by the Society for Academic Emergency Medicine Abstract Objectives: The objective was to determine if emergency department (ED) patients with diabetes mellitus (DM) who have primary care providers (PCPs) have better control of their DM than patients with no PCPs. Methods: This was a prospective, cross-sectional, observation study at a large, adult, urban, academic ED with 85,000 annual visits. ED patients with a history of DM were eligible. Patients with severe systemic disease, diabetic ketoacidosis (DKA), sepsis, active steroid use, pregnancy, or cognitive impairment were excluded. Consenting patients had hemoglobin A1c (HgbA1c) analysis and completed a questionnaire regarding demographics, lifestyle, medication usage, educational level attained, and health care access, including whether or not they had PCPs. HgbA1c levels were compared between subjects with and without PCPs using medians with interquartile ranges (IQRs). A continuous plot was developed to demonstrate the proportion of patients without PCPs (PCP−) compared to those with PCPs (PCP+) at every level of %HgbA1c across the entire measured range. Multivariate logistic regression analysis was used to determine which clinical and demographic factors obtained from the questionnaire were associated with improved glycemic control (increased relative risk [RR] of having a %HgbA1c < 8%). Results: A total of 284 patients were screened; 227 were enrolled, had HgbA1c analysis performed, and had complete PCP, race, and sex information. Complete demographic data (insurance status, employment status, etc.) were available on 209 subjects. Sixty-four of the 227 patients (28.2%) denied having PCPs. Median HgbA1c was 7.7% (IQR = 6.5% to 9.68%) in PCP+ versus 8.9% (IQR = 6.8% to 11.3%) in PCP− patients (p = 0.01). Ninety-one of 163 (55.8%) PCP+ subjects had a median HgbA1c < 8% versus 25 of 64 (39.1%) in the PCP− group (p = 0.02). After adjusting for multiple clinical and demographic variables, having a PCP remained significantly associated with a median HgbA1c value less than 8% (RR = 1.43; p = 0.04). Conclusions: Diabetes control was significantly better in patients with PCPs, even after adjusting for a number of potentially confounding social and demographic factors.

5 August 2020 in Academic Emergency Medicine
Verified
Journal Article
Closed access
PDF file
KBKyla Bauer
Kyla Bauer
OSOmolade O Sogade
Omolade O Sogade
BGBrian F Gage
Brian F Gage
BRBrent Ruoff
Brent Ruoff
LLLawrence Lewis
Lawrence Lewis

1 November 2012 in The American Journal of Emergency Medicine
Verified
Journal Article
Closed access
JRJennifer Prah Ruger
Jennifer Prah Ruger
LLLawrence M. Lewis
Lawrence M. Lewis
CRChristopher J. Richter
Christopher J. Richter

1 October 2015 in Respiratory Medicine
Verified
Journal Article
Closed access
SHStacey House
Stacey House
KMKazuko Matsuda
Kazuko Matsuda
GOGeoffrey O'BRien
Geoffrey O'BRien
MMMalath Makhay
Malath Makhay
YIYuichi Iwaki
Yuichi Iwaki
LLLawrence M. Lewis
Lawrence M. Lewis

30 November 2023 in medRxiv
Verified
Preprint
Open access
PDF file
Thomas F HestonThomas F Heston
Thomas F Heston
LLLawrence M. Lewis
Lawrence M. Lewis
BACKGROUND ChatGPT is a large language model with promising healthcare applications. However, its ability to analyze complex clinical data and provide consistent results is poorly known. This study evaluated ChatGPT-4’s risk stratification of simulated patients with acute nontraumatic chest pain compared to validated tools. METHODS Three datasets of simulated case studies were created: one based on the TIMI score variables, another on HEART score variables, and a third comprising 44 randomized variables related to non-traumatic chest pain presentations. ChatGPT independently scored each dataset five times. Its risk scores were compared to calculated TIMI and HEART scores. A model trained on 44 clinical variables was evaluated for consistency. RESULTS ChatGPT showed a high correlation with TIMI and HEART scores (r = 0.898 and 0.928, respectively), but the distribution of individual risk assessments was broad. ChatGPT gave a different risk 45-48% of the time for a fixed TIMI or HEART score. On the 44 variable model, a majority of the five ChatGPT models agreed on a diagnosis category only 56% of the time, and risk scores were poorly correlated (r = 0.605). ChatGPT assigned higher risk scores to males and African Americans. CONCLUSION While ChatGPT correlates closely with established risk stratification tools regarding mean scores, its inconsistency when presented with identical patient data on separate occasions raises concerns about its reliability. The findings suggest that while large language models like ChatGPT hold promise for healthcare applications, further refinement and customization are necessary, particularly in the clinical risk assessment of atraumatic chest pain patients.