Understanding Primary Care Physicians’ Comprehension of Eye Care Notes to Improve Diabetes Care Coordination

Technical Summary: Understanding Primary Care Physicians' Comprehension of Eye Care Notes to Improve Diabetes Care Coordination

Problem Statement
Diabetic retinopathy (DR) is a leading cause of preventable blindness, requiring coordinated care among ophthalmologists, optometrists, primary care providers (PCPs), and endocrinologists. Despite the prevalence of shared electronic health records (EHRs), effective inter-specialty communication is often hindered by the structure and content of eye care documentation. Eye care notes frequently utilize specialty-specific jargon, abbreviations, and formatting that may be opaque to non-ophthalmologists. This study addresses the gap in understanding whether PCPs and endocrinologists can accurately interpret ophthalmology notes to manage diabetes care effectively, specifically regarding the identification of dilated fundus examinations and diabetic retinopathy status.

Methodology
The authors conducted a cross-sectional online survey involving 82 PCPs and endocrinologists across three academic tertiary health systems in California (University of California Davis, UCLA, and UCSD) between November 2022 and December 2024. The study was approved as an exempt protocol by the UCSD IRB.

• Participants: The cohort included family medicine physicians, internal medicine primary care physicians, and endocrinologists. Demographic data included age, sex, race, specialty, practice duration, and prior ophthalmology training (rotations in medical school or residency).
• Survey Instrument: The survey utilized a 5-point Likert scale to assess self-reported comprehension of eye care notes, including general understanding, interpretation of exam findings, and assessment/plan comprehension.
• Practical Assessment: A subset of 67 participants reviewed three sample eye care progress notes. For each note, they were tasked with identifying: (1) whether a dilated fundus examination was performed, and (2) the presence or status of diabetic retinopathy.
• Qualitative Analysis: A voluntary free-text section allowed respondents to provide suggestions for improving care coordination. These responses underwent thematic analysis to identify recurring barriers and proposed solutions.
• Data Analysis: Descriptive statistics were calculated for demographic and survey data. Free-text responses were analyzed thematically.

Key Results

• Self-Reported Confidence: Comprehension confidence was modest. While 76.4% agreed or strongly agreed they understood the "assessment and plan" section of notes, only 47.2% agreed they generally understood eye care notes, and merely 6% strongly agreed. Confidence in understanding specific eye exam findings was even lower (38.9%).
• Practical Interpretation Accuracy:
  • Dilation Status: Participants demonstrated high accuracy in identifying whether a dilated fundus exam was performed, with correct identification rates ranging from 94% to 96% across the three sample notes.
  • Diabetic Retinopathy Status: Accuracy varied significantly, ranging from 37.3% to 82.1% depending on the note. Notably, up to 58.2% of respondents were "unsure or blank" regarding the DR status in one of the sample notes.
• Confidence vs. Competence Gap: A critical finding was the disconnect between confidence and accuracy. 29% of respondents who answered fewer than half of the practical questions correctly still reported confidence in their general understanding of eye care notes. Conversely, those who answered all questions correctly were more likely to express uncertainty about their answers than those who answered 5 of 6 correctly.
• Barriers Identified: Thematic analysis of free-text responses highlighted three primary barriers:
  1. Inconsistent Coding and Formatting: Issues included unclear linkage between diagnoses and medications, lack of DR screening dates in notes, and scanned handwritten notes.
  2. Jargon and Abbreviations: The use of ophthalmology-specific abbreviations (e.g., NPDR) and terminology created interpretation challenges.
  3. Documentation Practices: Respondents noted that problem lists in shared EHRs were often not updated by eye specialists.
• Coping Mechanisms: When encountering unfamiliar terms, 68% of respondents reported using online search engines, while 26% consulted the referring eye care provider.

Key Contributions
The study provides quantitative evidence of a communication gap in inter-specialty diabetes care. It demonstrates that while PCPs and endocrinologists frequently review ophthalmology notes, their ability to extract critical clinical data—specifically diabetic retinopathy status—is inconsistent. The research identifies specific structural failures in current documentation practices, such as the over-reliance on non-standard abbreviations and the lack of standardized templates for DR screening results. Furthermore, the study highlights a "confidence-competence" gap, where providers may overestimate their ability to interpret specialized notes, potentially leading to gaps in care coordination.

Significance and Claims
The authors claim that the study reveals measurable gaps in the interpretation of key clinical elements within ophthalmology notes by non-ophthalmologists. They posit that these gaps pose a risk to the coordination of care for patients with diabetes.

The paper concludes that more standardized and structured documentation is necessary to improve inter-specialty communication. Specifically, the authors suggest that:

• Standardized note templates that clearly document DR screening status could reduce ambiguity.
• Reducing the use of ophthalmology-specific abbreviations or employing automated expansion within EHRs could improve interpretability.
• Linking medications directly to diagnoses and maintaining updated problem lists in shared EHRs would enhance clarity.

The authors maintain a modest tone regarding their findings, acknowledging limitations such as the sample size, potential response bias, and the restriction to academic medical centers in California. They do not claim that their findings represent a universal truth for all providers but suggest that the identified barriers and proposed structural improvements are viable strategies to enhance care coordination for diabetic patients. The study also notes the potential role of emerging technologies, such as Large Language Models (LLMs), in "translating" specialized terminology, though this is presented as a potential avenue rather than a tested solution within the study.