Implementation of Diabetic Retinopathy Screening in Adult Patients With Type 2 Diabetes in a Primary Care Setting Free

By 2025, the prevalence of type 2 diabetes in U.S. adults is expected to increase by 50%, while the availability of trained optometry and ophthalmology providers is expected to increase by only 1–2% (10). Given these projections, the demand for DR screening will likely outpace optometry/ophthalmology accessibility. The ADA’s Standards of Care in Diabetes—2023 and its abridged version for PCPs recommend remote DR screening with validated retinal imaging technology to increase DR screening rates. Such programs should allow for prompt referral to an optometrist or ophthalmologist if warranted (9,11). By providing DR screening in the primary care setting, PCPs can enhance the person-centered approach to managing diabetes by increasing convenience, access, and educational opportunities (7). Therefore, primary care settings must create innovative methods to meet the needs of patients with diabetes and alleviate this potential eye health crisis (6).

There is an abundance of literature regarding provider- and patient-level factors associated with barriers to DR screening. These barriers include limited accessibility, time constraints, financial burden, and inadequate patient education (7,12–14). Identifying barriers is an essential step in creating strategies to mitigate suboptimal DR screening adherence. For example, intervention studies focused on improving access to care demonstrated an association with increased DR screening adherence when access was increased within the primary care setting (12–14). Furthermore, one systematic review found accessibility to be a recurring problem for patients, hindering DR screening compliance (12). In response, researchers proposed four key recommendations: improve convenience of DR screening, increase awareness/education of DR screening, reduce anxieties, and improve communication (12).

DR screening is traditionally performed within an optometry or ophthalmology clinic. However, in recent years, screening has begun shifting from the traditional specialty care approach to nontraditional approaches such as remote screenings offered in various settings (13). Remote DR screening, sometimes referred to as screening beyond an optometry/ophthalmology practice, is not a new concept. Although remote DR screening has not been widely implemented in the United States, other countries have successfully established remote DR screening in primary care settings and have hosted city-wide DR screening events that are open to the general public (13,15). For example, since initiation of a national DR screening program in England in 2003, DR is no longer the dominant cause for blindness in working-aged adults in this country (15). Additionally, remote DR screening is gaining global acceptance because of its high sensitivity and specificity, easy access, reduced cost, consistent patient follow-up, and convenience (4,10,16–20).

Various forms of remote DR screening include use of handheld cameras (4,10,16,18,20), artificial intelligence (15,17,19,21), teleophthalmology (6,15,20), and mobile screening units. These multiple forms of remote DR screening have successfully demonstrated improved DR screening compliance, particularly in the primary care setting.

A remote screening mobile unit is a vehicle equipped to perform DR screenings remotely in any practice setting and/or provide DR screening equipment to a clinic. Remote screening provides patients with diabetes in rural and low-income communities access to DR screening (12,15,22). Research findings demonstrated that mobile DR screening units have been broadly successful for improving DR screening compliance and also for early identification of treatable DR (12,15,22).

Although studies have demonstrated the benefit of DR screening within the primary care setting, a common theme of initial reluctance and negativity on the part of providers was also noted. For example, several studies reported that PCPs are hesitant to accept the concept of remote DR screening because of perceived increased provider workload, additional responsibilities, and increased clinic costs. Moreover, PCPs reported that interpreting DR screening results was better suited for their optometrist/ophthalmologist colleagues because PCPs lacked confidence and formal training in performing and interpreting DR screening exams (7,13,18). Interestingly, several of these studies followed PCPs through the process of training for and interpreting DR eye exams. Once trained and appraised by ophthalmologists, PCPs were not only observed to sufficiently interpret exam findings, but also formed supportive professional relationships with their optometry/ophthalmology colleagues (15,16,18,23). Furthermore, these studies noted that PCPs felt confident with DR screening, enjoyed discussing results with their patients, and found DR screening cameras to be advantageous for evaluating and diagnosing other eye conditions (18). The partnership between optometrists/ophthalmologists and PCPs for DR screening is consistent with the ADA’s recommendations regarding person-centered, team-based diabetes care (9,11).

The ADA recommends annual (or biennial, in some cases) DR screenings for patients with type 2 diabetes (9). The overarching goal of this innovative quality improvement (QI) project was to increase DR screening in adult patients with type 2 diabetes by implementing a mobile DR screening unit at a primary care clinic in the southeastern United States.

The first aim of this project was to increase DR screening quality metrics in adult patients with type 2 diabetes by 15%, in collaboration with an accountable care organization, to improve access to and compliance with DR screening at the end of 5 months using DR screening mobile units. The second aim was to increase patients’ knowledge and awareness of the importance of DR screening, as measured by a pre- and post-educational questionnaire.

This QI project used the Model for Improvement, developed by Associates in Process Improvement (24). This model was chosen because it measures specific times and changes and selects ideas from the experiences of the clinical team throughout the QI project. A pre-/post-design was used to measure quality metric percentages and knowledge/awareness questionnaire scores. Descriptive data included patient demographics (age, sex, and race), diabetes duration, A1C at the time of DR screening, and last retinal exam.

The project was implemented at Cone Health LeBauer Healthcare Stoney Creek, a primary care clinic located in the small town of Whitsett, NC, which encompasses several diverse communities. LeBauer Healthcare is part of Cone Health and Cone Health Medical Group, which is a private, not-for-profit health care delivery system based in Greensboro, NC, within the Triad, a metropolitan region of north-central North Carolina that comprises the cities of Greensboro, High Point, and Winston-Salem.

The participating clinic provides primary care to ∼15,000 unique patients and sees 120–150 patients per day. The clinic population primarily consists of insured adults from diverse socioeconomic backgrounds. More than half of the patients (52%) have commercial insurance, 27% are enrolled in Medicare, and 21% are self-paying. The clinic team is composed of 10 PCPs (three nurse practitioners and seven physicians) and their respective certified medical assistants, two licensed practical nurses, the practice administrator, and 10 clerical staff members.

The clinic partnered with a local accountable care organization, Triad Healthcare Network (THN), which offers patients assistance in meeting guideline-driven standards of care (e.g., for mammograms, colonoscopies, and DR screenings). THN contracts with select commercial insurance providers and Medicare to provide patients with standard of care services. Approximately 2,000 (13%) of the adult patients within the participating primary care clinic have an established diagnosis of type 2 diabetes, and nearly 1,000 (50%) of these patients are insured and contracted with THN.

Inclusion criteria for this project were 1) established patients within the clinic, 2) ≥18 years of age with an established diagnosis of type 2 diabetes, 3) with an insurance provider that was contracted with THN (all patients with a contracted insurance provider receive complimentary health care maintenance services), and 4) due or overdue for an annual DR screening according to ADA recommendations. Patients were excluded if they 1) were pregnant adults with gestational diabetes; 2) had type 1 diabetes; 3) had ocular disease such as glaucoma, macular degeneration, optic neuropathy, retinal vessel occlusion, or choroidal disease; and 4) had already completed their annual DR screening. Patients with type 1 diabetes within the participating primary clinic are managed by endocrinologists for diabetes treatment and routine screening exams.

A convenience sample of ∼100 adult patients with type 2 diabetes who met the inclusion criteria were selected from the participating primary care clinic. Qualified patients were identified by THN, which keeps records of all health maintenance compliance, including diabetes recommendations. THN representatives contacted patients who met the inclusion criteria and offered them the option to complete their DR screening exam at their PCP’s clinic. Upon accepting the offer, participants received information regarding DR, the importance of annual screening, and the mobile screening unit program offered at the participating clinic. The THN representatives then scheduled patients on one of eight reserved screening dates between August and December 2022, with approximately 13–16 patients scheduled per screening day. Screenings took place between 9:00 a.m. and 3:00 p.m., with each lasting ∼20 minutes. The 5-month project duration was necessary to ensure that enough patients were served and sufficient data were accrued.

On each screening day, two THN representatives who were trained to administer dilated eye exams arrived with all of the necessary equipment to complete DR screening exams and operated out of two designated exam rooms within the clinic. One day before each screening day, these representatives provided the clinic nurse practitioner who served as the project champion with a list of patients expected to come in for screening the next day. The screening dates were selected based on THN mobile unit availability.

The THN representatives completed DR screenings using handheld cameras. A dark room without windows was the method used for eye dilation. Patients scheduled for DR screening notified clerical staff at check-in regarding the purpose of their visit and were checked in using a paper attendance log. Clerical staff provided patients with two forms, including 1) an educational handout adapted from Centers for Disease Control and Prevention materials (25) that explained DR and the significance of DR screening (Supplementary Appendix A) and 2) a two-sided questionnaire developed by the project champion based on the standard of care for diabetes that assessed patients’ knowledge and awareness of the importance of DR screening (Supplementary Appendix B). In addition, patients received a small, bright sign on which “THN” was displayed. This sign was used as a mechanism to differentiate patients who were presenting for DR screening from those presenting for routine office, laboratory, or nurse visits.

All patients who participated in the QI project were asked to complete a pre- and post-screening questionnaire. The questionnaire contained two parts; the first listed five true-or-false questions assessing knowledge and awareness of DR, and the second listed two true-or-false questions regarding satisfaction of their DR screening experience (Supplementary Appendix B). Figure 1 illustrates the DR screening process flow.

Scheduled patients who participated in this QI project were escorted to the screening examination room. Eye dilation was completed by darkening the room and allowing the patient to sit for 10 minutes. Once dilation occurred, retinal images were captured by handheld cameras, uploaded to a partnering ophthalmology group, and remotely interpreted by an ophthalmologist.

All data were collected by the project champion. Diabetes quality metric data collected were the percentages of patients who completed DR screening. Data for DR screening quality metrics were obtained from a secure health system server that tracks several categories of quality metric data for the participating primary care clinic. Data tracking for DR screening is an automated process. The server communicates with the electronic medical record (EMR) system by tracking each DR screening completion and by uploading these data into the quality metric dashboard once per month.

For the server to recognize a DR screening completion, several interventions occurred. Upon completion of a DR screening, the THN representatives uploaded the captured retinal images and sent them via secure e-mail to a remote ophthalmologist for evaluation. The ophthalmologist interpreted the retinal images and provided recommendations for either continued annual/biennial screening or for a referral to a local ophthalmologist for a comprehensive examination and further treatment, as appropriate. Interpretated DR screening reports, including results and recommendations, were sent via secure e-mail to the project champion, who delivered a printed copy of the report to the office inbox of the patient’s PCP. The PCP then reviewed the final interpreted report and entered the screening date into the EMR as completed. However, this data entry neither confirmed nor concluded DR screening as complete. Once reviewed by the PCP and abstracted in the EMR, the interpreted report was forwarded via interoffice mail to the primary care clinic’s medical records department. When medical records staff uploaded the report into the EMR under the DR screening requirement, this action prompted a completion status and manifested within the diabetes quality metric dashboard under DR screening. Credit for DR screening quality metrics was not official until the medical records department received and scanned the screening report into the EMR system.

The percentages of completed screenings (i.e., the quality metric data) were input into a Microsoft Excel worksheet for the 3 months before project implementation and project’s 5-month duration. Patients’ demographics, diabetes duration, A1C at the time of DR screening, and last retinal exam were also collected on the Excel worksheet for all patients participating in the project. All identifying information (i.e., patients’ name, medical record number, and date of birth) was excluded.

Completed knowledge, awareness, and satisfaction questionnaires were collected from each patient by the THN representatives and returned to the project champion at the end of each screening day. Scores for knowledge and awareness and satisfaction were compiled into an Excel worksheet after each screening day (Figure 1).

Fisher exact and paired t tests were used to analyze the data. A Fisher exact test was conducted to evaluate pre- and post-project DR screening compliance. A paired sample t test was used to determine the mean difference between the pre- and post-screening knowledge scores and to evaluate patient satisfaction scores regarding completing DR screening in the primary care setting. Data analyses were conducted with SPSS, v. 29.0, statistical software (26).

Forty-six patients participated in DR screening throughout the 5-month QI project. Most participants were older adults (mean age 65.6 ± 9.0 years, range 41–88 years) and male (n = 27 [58.7%]). Most participants were Caucasian (n = 32 [69.6%]), followed by African American (n = 13 [28.3%]) and Asian (n = 1 [2.2%]) (Table 1). The mean type 2 diabetes duration was 9.6 ± 8.4 years (range 1–31 years). Mean A1C at the time of DR screening was 7.4 ± 1.5% (range 5.5–12.4%). Among the 26 participants who provided a response, time since the last retinal exam ranged from 1 to 10 years; eight of the participants had never completed a DR screening, nine did not provide a response, and three were unsure (Table 2).

The first aim of this project was to increase DR screening quality metrics by 15% in adult patients with type 2 diabetes in the participating primary care clinic by the end of 5 months through the use of DR screening mobile units. A target DR screening rate of 35% was preset by the clinic health system. For the 3 months before project implementation (May to July 2022), the average DR screening rate was 45.3%, and this improved to 52.1% during the 5-month implementation period (August to December 2022) (Figure 2). A Fisher exact test showed that this change represented a statistically significant increase (P = 0.018) and a 15% relative improvement in average DR screening rates, indicating that the first project aim was achieved.

Of the 46 patients who were screened, five were found to be positive for DR, and four of those patients had no prior diagnosis of DR. Three were determined to have moderate nonproliferative DR (n = 2 without macular edema and n = 1 with macular edema), and two were determined to have mild DR (n = 1 with mild nonproliferative DR with macular edema and n = 1 with mild proliferative DR without macular edema). One patient was unable to complete DR screening because eye dilation was insufficient.

The second aim of this project was to increase participants’ knowledge and awareness of the importance of DR screening. Knowledge scores were computed as a percentage of correct responses ranging from 0 to 100% for each participant pre- and post-education. A paired t test was conducted to compare the mean knowledge scores and showed improvement from pre-screening (mean 81.5 ± 24.9%) to post-screening (mean 88.6 ± 19.5%) (t[45] = −2.37, P = 0.022).

Satisfaction was computed as a percentage marking “true” regarding a preference to complete DR screening exams at the primary care clinic. The overall satisfaction rate was 83% of patients (across the sample size of 46 patients) who answered “true” that completing DR screening in their PCP’s office was convenient and preferred.

The purposes of this QI project were to increase DR screening rates and improve knowledge and awareness of DR in adult patients with type 2 diabetes. This project was built on a foundation of well-established ADA guidelines for DR screening in people with diabetes (9), and the results demonstrate the importance of addressing DR screening barriers and gaps in knowledge and awareness. Additional fundamental goals of this project were patient-centered, including increasing access to care, reducing financial burden by eliminating specialty copayments, and saving time by enabling patients to receive DR screening during a diabetes follow-up visit with their PCP (7,12–14).

The concept of implementing DR screening within a primary care setting created hesitancy among PCPs and clinic staff, who feared that it would lead to increased workload and additional responsibility. This issue was anticipated because existing literature revealed similar reasons for reticence among providers (7,13,18). During project implementation, the PCPs and staff realized that the intervention did not increase their daily workload. This fact, combined with the improvement in DR quality metrics and positive reactions from patients, began to change PCP and staff attitudes. As PCPs began noticing the favorable outcomes, they began inquiring about future planned DR screening dates. These inquiries were encouraging because they demonstrate willingness to step outside of traditional practice to meet the needs of patients and close gaps in care.

Participating patients were enthusiastic about completing their annual DR screening exam in their PCP’s clinic because this was a novel idea from their perspective. Anecdotal qualitative comments from the patients included remarks about the convenience and ease of the examinations, the comforting familiarity of their PCP’s clinic, and the efficient process of the screening exam, which are also common themes in the literature.

Partnering with THN was instrumental for project development, implementation, and outcomes because there was some apprehension from the clinic’s PCPs and administration. The partnership with THN also provided additional support because THN was partnered with an ophthalmology group for remote eye exam interpretation and also aided the clinic in contacting and scheduling patients for DR screening exams. Change through innovation cannot occur without collaboration and support, especially when implementing a novel concept. It was through partnership with community resources that the project gained momentum and resulted in a powerful change.

Since its completion, this QI project has drawn attention from health care system administrators and the PCP patients associated with the participating clinic. Discussions of sustainability were held throughout the QI project, and after completion of a few intervention days, it became apparent that planning for future screening dates was warranted. The project champion and THN agreed to continue the partnership for 2023 and have already secured five additional DR screening dates for the clinic. Additional dates are anticipated to be scheduled throughout the year. Future goals include implementing DR screening exams in other local primary care clinics associated with the health care system.

Several limitations to this project should be acknowledged. The most impactful limitation was the unexpected and unavoidable temporary closure of the participating clinic, which significantly affected patient participation. The clinic was temporarily closed to complete urgent building repairs and was divided and relocated between two other local primary care clinics within the health system. The project champion and team were relocated to a clinic 22 miles from the original participating clinic. Two of the eight scheduled intervention dates were held at the original clinic before this temporary closure. Participation before the clinic closure attracted 13–15 patients per screening day, whereas participation after the temporary relocation ranged from 1 to 6 patients per screening day.

Additionally, the anticipated total sample size for the project was 100 patients, but the final total sample size was 46 patients. Primary reasons for the lower participation rate included patients’ unwillingness to drive 22 miles from their local clinic to the temporary location, and intimidation because direct access to the relocation site primarily consisted of a major interstate. Scheduled patients who did not show for their screening were contacted by the THN representatives or project champion and offered another screening date. Those who arrived late for their appointment but on the scheduled day were worked in that day.

The inclusion criteria also created several limitations. The project was open to established patients of the practice and therefore was not available to members of the community at large. Qualified established patients were those who were insured and whose insurance provider was contracted with THN. Although this criterion captured a large proportion of patients at the participating clinic, a large proportion of patients within the clinic were unable to participate because they were either uninsured or underinsured and without the THN benefit. A future goal is to begin exploratory discussion regarding opening access to DR screening exams for established patients with and without insurance regardless of the existence of a THN contract. The project time constraint of 5 months was another limitation in that it hindered data collection and reduced the expected sample size.

DR quality metric scores at project completion showed a 15% relative increase compared with screening rates 3 months before project implementation. This increase confirms that the project was impactful and successfully demonstrated proof of the concept. The project team presumes that DR screening quality metric scores would have been much higher had the participating clinic not been required to relocate during the implementation period. Patient knowledge and awareness of DR and the importance of routine screening increased after education and examination. Since project completion, the intervention has resumed in the original clinic site, and the team has noted increased DR screening participation.

Despite several limitations, the overall outcome of this QI project was positive and resulted in a change in health care delivery within the participating clinic. The results are valuable in supporting the project’s aims, and they are pivotal in reinforcing the ADA’s eye exam recommendations. With a better understanding of DR, effortless access to screening, and a patient-centered process, patients will be more willing to participate in routine DR screening, which may prevent irreversible DR outcomes and improve quality of life.

The authors thank the members of the THN team for their valuable partnership and contribution to the project and LeBauer Primary Care PCPs and staff for their support and participation.

No potential conflicts of interest relevant to this article were reported.

K.K.C. completed a review of the literature; developed, organized, and implemented the QI project; and wrote the manuscript. J.G., J.R.C., and B.I.P. contributed to discussion and project planning, and reviewed and edited the manuscript. K.K.C. is the guarantor of this work and, as such, has full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.