Abstract
Keywords
Introduction
Diabetic retinopathy is the most common microvascular complication of diabetes and the leading cause of new cases of blindness among working age adults in the United States. 1 In 2021, an estimated 9.6 million people with diabetes in the United States had diabetic retinopathy with 1.84 million people living with vision-threatening diabetic retinopathy.2,3 Given the high prevalence of type 2 diabetes in Alabama, the state has a significant burden of diabetic retinopathy with an estimated 3.06% of the population, approximately 150 000 people, affected.4,5 As with diabetes prevalence, racial and ethnic minorities are more likely to be affected; non-Hispanic Black and Hispanic individuals have higher prevalence of diabetic retinopathy than non-Hispanic White individuals. 2
Early identification and treatment of diabetic retinopathy can reduce the risk of progression to severe vision loss by over 90%. 1 The American Diabetes Association (ADA) Standards of Care recommend dilated and comprehensive eye exams by an ophthalmologist or optometrist for people with type 2 diabetes at the time of diagnosis and then at least every 1 to 2 years. 6 Although these evidence-based guidelines for screening exist, implementation is suboptimal and a large proportion of people with type 2 diabetes do not receive recommended comprehensive eye exams annually or biennially. 7 In 1 study of Medicare beneficiaries with diabetes, 54% of people with diabetes nationally had an annual eye exam contrasted to only 47% in Alabama and only 43% of Black adults in Alabama. 8 Many factors contribute to the research-to-practice gap, including educational deficiencies, time constraints for providers, lack of feedback mechanisms and decision support, limited access to specialty care, as well as cultural factors and organizational climate. Improved processes, tools, and resources for providers and patients are needed to increase rates of screening for diabetic eye disease.
Understanding contextual factors from the very beginning is necessary to tailor an intervention or implementation strategies to a specific setting. 9 In this study, our objective was to qualitatively assess barriers to documentation, screening, and referral to recommended eye screening for patients with diabetes among providers and practice staff through guided discussions in 2 clinical settings in Alabama.
Methods
Setting
We conducted this qualitative study at 2 sites in a densely populated county in Alabama, with a resident population exceeding 670 000. One site is a safety-net healthcare system offering a range of ambulatory care services such as primary care, a multi-disciplinary diabetes clinic, and ophthalmology. The second site is a specialty endocrinology clinic integrated within a large academic medical center. This qualitative study was part of a larger quality improvement initiative; we conducted discussion groups at each site.
Participants
We employed purposeful sampling to select participants for this study, targeting a diverse group of healthcare providers including residents, physicians, advanced practice providers, nurses, and administrators. Invitations were extended via email to the relevant providers and administrators involved in primary care and specialty care for patients with diabetes (eg, endocrinology, ophthalmology) at each site. Participants were offered a complimentary lunch as a token of appreciation for their involvement. Additionally, each participant received a gift card for their time and contribution. This study was approved by the Institutional Review Board, and all participants provided informed consent prior to the discussion.
Data Collection
To identify facilitators and barriers to diabetic eye exam referrals, appointment scheduling, and documentation and tracking, we developed a discussion guide with questions related to each of these processes. The guide also included questions about past or current quality improvement initiatives related to diabetic eye health and communication about risk of diabetic eye disease. The facilitator’s guide is provided in Table 1. Both discussion groups were conducted in August 2023. The discussion groups were conducted in-person, taking place in designated conference rooms at both study sites. Each session spanned a duration of 1 h. Facilitation of the discussion groups was carried out by program directors holding an MPH and PhD qualifications in clinical psychology, respectively. Their extensive experience in qualitative research provided valuable expertise in guiding the discussions. All discussion groups were audio-recorded, subsequently transcribed, and validated for data quality and accuracy. In addition to the recordings, detailed notes were taken during the sessions to supplement the transcript data.
Focus Group Discussion Guide.
Data Analysis
Transcripts from the discussion groups were coded by 2 members of the research team; a program manager with an MSPH and extensive training in conducting qualitative research, and a primary care physician-researcher with proficiency in qualitative research. A combined inductive and deductive approach was applied to perform qualitative content analysis. Each reviewer read both transcripts to gain an overarching understanding of the group discussions. Reviewers developed initial codes from facilitator’s guide as part of the deductive analysis. Reviewers identified meaningful units in the transcript text; codes were then created and assigned to each meaningful unit. Emerging codes were developed during the coding process. Reviewers discussed the codes to reach consensus about codebook. Reviewers independently applied the codebook to each transcript, then subsequently reviewed the codes together to ensure consensus. Overall, 68 comments were captured, organized into 20 subthemes and 6 overarching themes. The number of discussion groups was based on the number of sites enrolled in the overall quality improvement initiative. To reach saturation of themes, it would have been necessary to enroll additional sites, which was beyond the scope of the pre-implementation phase of this initiative.
Additionally, the Practical Robust Implementation and Sustainability Model (PRISM) was integrated into our thematic analysis to help us organize the major themes and subthemes.9,10 PRISM outlines multi-level contextual domains that interact with an evidence-based intervention program, including perspectives from patients and organizations, characteristics of patients and organizations, implementation and sustainability infrastructure, and the external environment. PRISM can be applied in pre-implementation or planning stages for a future intervention. In this study, we chose to apply PRISM because its domains were aligned with the themes and subthemes developed from the discussion groups.
The software application NVivo (version 12 Plus) was utilized for data management, coding, and organization. Preliminary findings from the thematic content analysis were presented to participants for member checking, enhancing the validity and trustworthiness of our results.
Results
Discussion Group Participant Characteristics
Table 2 summarizes the characteristics of the participants in the 2 discussion groups (N = 28). Physicians and faculty were the predominant group within the sessions (42.9%) followed by clinical managers and administrators (25%). Time within current roles ranged from less than 1 year to 27 years, with fellows and administrators, on average, spending less time in their current role as compared to providers and clinical managers/supervisors.
Characteristics of Discussion Group Participants.
Qualitative Themes and Subthemes
We organized our themes into 4 domains using the PRISM framework as a guide: organizational characteristics (eg, existing processes for referrals and scheduling for diabetic eye exams), patient characteristics (eg, competing demands, knowledge and beliefs), organizational perspectives (eg, determinants, capacity), and implementation and sustainability infrastructure (eg, information technology and support, policies, personnel, skills or training).9,10 We present themes, subthemes, and illustrative quotes from the 2 discussion groups in Table 3 and briefly summarize these below.
Summary of Major Themes and Subthemes by PRISM Domain and Site.
Organizational Characteristics
Participants described the existing processes for placing referrals for diabetic eye exams, scheduling eye exam appointments, and documenting or tracking completed eye exams. Individual providers described tracking whether a patient had an eye exam through chart review or documenting outside eye exam in their note during a clinic visit; standardized review or tracking processes were lacking at both sites. Providers placed a referral, which was then acted upon by referral management or administrative clinic staff to schedule the appointments or submit referral paperwork for eye exams, including contacting patients with details about appointment date and time. In terms of where patients were referred, most safety-net primary care providers described referring their patients to ophthalmology on-site, whereas the specialty clinic providers reported that most of their patients follow up with an eye health provider outside of their system.
Patient Characteristics
Providers at both sites described multiple patient-level determinants of completing diabetic eye exams including patient difficulty in attending an additional appointment, frequent missed appointments for eye exams, financial barriers to attending eye exam appointment, and limited motivation to attend eye exam appointment if the patient is not having vision symptoms. Providers reported regularly communicating with their patients about the risk of eye disease and importance of eye screening.
Organizational Perspectives
Clinical or system-related determinants identified included competing demands during patient appointments, difficulty obtaining outside records for documentation of eye exams, limited utility of clinical reminder functions in the electronic medical record (EMR), and limited mechanisms for follow up with patients for missed appointments. Additionally, providers identified a challenge of patients being unaware of referral appointment information because these were scheduled after their primary care or specialty appointment and patients often reported not receiving appointment information for their eye exam referral. Providers also described challenges related to the use of different EMRs by different sites of care, including specialty eye hospital, as well as difficulties in knowing the appropriate place of care for a patient experiencing visual symptoms.
Implementation and Sustainability Infrastructure
Our findings from the discussion groups provided information about the available implementation and sustainability infrastructure at each site. Providers outlined potential improved processes for scheduling diabetic eye exams and initiatives to increase referral access and completion, including providing patients the appointment date and time for their eye exam when they are present for their clinic visit and tracking of status of referrals with follow-up when needed. Additionally, providers described previous quality improvement initiatives related to diabetic eye exams, need for provider education and training in the use of retinal cameras for screening, and importance of relationships between referring providers and ophthalmologists. A commonly identified barrier to implementation and sustainability was the lack of common or interoperable EMR between sites and the specialty eye hospital. At both sites, providers described ambiguity regarding ownership of the task of referrals for eye exams. Specialists noted often deferring eye exam referrals or screening to the patient’s primary care provider, while primary care providers suggested that other members of the healthcare team would be better positioned to track and coordinate screening and referrals.
Discussion
In this study, we determined organizational characteristics and perspectives, patient characteristics, and implementation and sustainability infrastructure relevant to completion of guideline-recommended eye exams in patients with diabetes. Common themes emerged related to existing processes for referrals, scheduling, and tracking including lack of standardizing processes for referrals; patient-level determinants to receiving recommended eye screening; clinic or system-related factors, and considerations for process improvement for eye exam referrals and appointments.
Our findings include clinic and system-level determinants of diabetic eye screening including challenges in service delivery and resource availability. A systematic review of qualitative studies related to eye screening in diabetes involving patients and providers demonstrated common barriers including cost and competing demands, limited knowledge, and challenges in service delivery. 11 Similar to another study in primary care, providers reported challenges identifying who was due for eye screening, relying on individualized practices to determine if a patient was up to date rather than standardized review processes. 12 The clinic and system-level determinants identified in this study point towards the need for more robust processes for connecting patients to needed eye care. Models that have been successful in achieving this include team-based care, which has been shown to improve receipt of recommended preventive care for patients with diabetes. 13 Additionally, care coordination with outreach to patients with diabetes increased referrals and rates of completed appointments for diabetic eye screening. 14
In our discussions, providers described patient-reported barriers to attending eye appointments including financial, transportation, and scheduling related to work or childcare. Our findings are consistent with prior studies that demonstrate patient-level barriers to access including costs and competing demands or concerns.11,15 A recent study found the adults with an unmet social need, such as housing or food insecurity, were less likely to have had an eye appointment within the previous year. 16 These findings support the efforts to consider a patient’s social determinants of health as part of their care. 17
Tele-ophthalmology, including digital retinal imaging, is a potential avenue to expand access to screening for diabetic eye disease and increase screening rates. 18 Successful implementation of teleophthalmology programs has been demonstrated; for example, England deployed a nationwide program and achieved annual screening rates of greater than 80% of people with diabetes. 19 Coordinating appropriate follow-up specialty care for patients who have an abnormal screen is a key part of successful teleophthalmology programs. A prior study in a safety-net healthcare system demonstrated low levels of adherence to recommended follow-up care after tele-retinal screening with less than one-third of patients completing recommended appointments. 20 This underscores the importance of ensuring adequate systems are in place to track and connect patients to specialty eye care when needed. Additionally, artificial intelligence-driven methodologies are being developed and tested in diabetic retinopathy across the spectrum of care from diagnosis and screening to predicting disease progression.21,22 These technologies have the potential improve access to screening, and it will be important to understand how the relevant context is influencing implementation across different treatment settings.
This study does have the following limitations. Our study design captures provider and practice staff experiences at 1 point in time. Although these findings represent provider experiences in both a safety-net healthcare system and specialty clinic within a large academic medical center, they may not be applicable to other settings such as privately owned practices and community clinics. Our findings were generated from a small sample size of 2 discussion groups, representing the 2 sites of the QI initiative. Although we sought out patient perspectives in another aspect of the QI project, this paper lacks the direct patient perspective.
Conclusion
In this study, providers and practice staff provided valuable insight into the contextual factors influencing diabetic eye screening in the pre-implementation phase of a larger quality improvement initiative. We plan to apply these findings to inform the strategies deployed to improve rates of recommended routine, comprehensive eye exams for patients living with type 2 diabetes. Additional work is needed to close the research-to-practice gap to determine the most effective strategies to improve the health system infrastructure to increase eye exam referral and screening rates for people with diabetes.
Footnotes
Acknowledgements
The authors would like to thank April Agne, MPH and Trudi Horton, PhD for lending their time and expertise to facilitating the discussion groups. The authors would also like to acknowledge the time and contributions of discussion group participants.
Author Contributions
All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by CP, MC, and DH. The first draft of the manuscript was written by CP and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was funded through a partnership with the American Diabetes Association (ADA) and Genentech, a member of the Roche Group, as part of the ADA’s Health Equity Now work. Dr. Presley also reports funding support from National Center for Complementary and Integrative Health (K23AT011375).
