282 - Characterizing High-Frequency MyChart Users for Better Care and Resource Allocation

Publication Number: 282.4428

Lauren C. Ferguson, University of Virginia School of Medicine, Charlottesville, VA, United States; Susan Gray, University of Virginia School of Medicine, Charlottesville, VA, United States; Martha Hellems, Virginia Commonwealth University School of Medicine, Charlottesville, VA, United States; Gabriel K. Lau, University of Virginia School of Medicine, Charlottesville, VA, United States

Background: Increasing numbers of electronic health record (EHR) messages from patients are problematic because they are associated with clinician burnout, and patient satisfaction can suffer when messages are not acknowledged quickly. Little has been studied about which pediatric patients and their proxies are most likely to utilize EHR messaging. This study aims to characterize high-frequency messaging users for optimized allocation of resources and improved patient care.

Objective: To determine associations between patient portal messaging frequency and other pediatric patient characteristics.

Design/Methods: This was a retrospective observational study of patients seen at any of three academic General Pediatric clinics and one academic Adolescent clinic from July 2023 to June 2024. Associations between frequency of messaging and other characteristics were analyzed using Chi-squared analysis and one-way ANOVA with post-hoc testing. We defined groups for comparison as follows: ultra-high-frequency (ultra-HF) users were in the top 1% of messaging, high-frequency (HF) users were in the top 1%-5%, typical users had any amount of messaging below the high-frequency cutoff, and non-users had no message encounters in the period analyzed.

Results: There was a total of 15,211 EHR message encounters in this time frame. Messages from ultra-HF users accounted for 10-26% of all EHR messages, and messages from HF-users accounted for 25-63% of all messages. Ultra-HF users were significantly older, had a higher BMI, and had more emergency department (ED) visits than non- and typical users (p < 0.001 for all associations) and HF users (p < 0.01). HF users had more ED visits than typical and non-users (p < 0.05). Ultra-HF users were also more likely to have a surgical history (OR=7.8, p< 0.001) than non-users. Ultra-HF use patients were more likely to be female, White, and use commercial insurance (p < 0.01 for all associations) compared to non-users. Notably, 0% of HF and ultra-HF users required interpreter services.

Conclusion(s): Preliminary analysis shows that a very small number of pediatric patients (and their proxies) accounted for a very large proportion of EHR messages in all four practices. This data suggest that interventions that focus on this small group of patients may significantly reduce inbox burden. Understanding characteristics of patients most likely to utilize messaging can help pediatric practices allocate case management resources appropriately and equitably.

Table 1. Demographic Data
P-values by one-way ANOVA for continuous variables and Chi2 test for binary/categorical variables. Continuous variables represented as mean (SD).