Customize your JAMA Network experience by selecting one or more topics from the list below.
Keenum Z, McGwin G, Witherspoon CD, Haller JA, Clark ME, Owsley C. Patients’ Adherence to Recommended Follow-up Eye Care After Diabetic Retinopathy Screening in a Publicly Funded County Clinic and Factors Associated With Follow-up Eye Care Use. JAMA Ophthalmol. 2016;134(11):1221–1228. doi:10.1001/jamaophthalmol.2016.3081
To what extent do patients with diabetes in a diabetic retinopathy screening program in a publicly funded county clinic adhere to the timetable of recommended follow-up eye examinations?
In this prospective follow-up study, only 29.9% of patients adhered to recommendations to have an eye examination within indicated time frames, even though cost and accessibility were minimized as barriers. Two years later, 50.9% of patients still had not undergone an eye examination.
These data suggest that diabetic retinopathy screening programs unlikely meet their goals about adhering to eye care recommendations.
The public health success of diabetic retinopathy (DR) screening programs depends on patients’ adherence to the timetable of follow-up eye care recommended by the screening program. African Americans are among those at highest risk for DR and have one of the lowest rates of eye care use.
To assess the rate of adhering to recommended follow-up eye care in a DR screening program administered in a safety-net health care facility and to examine factors associated with follow-up eye care use.
Design, Setting, and Participants
Prospective follow-up study of persons with type 1 or type 2 diabetes. The setting was an internal medicine clinic of a publicly funded health system in Alabama, serving a population largely uninsured and African American, that had implemented a DR screening program using a nonmydriatic camera for ocular imaging and remote reading centers for evaluation of images. Patients with physician appointments between January 26 and July 24, 2012, were eligible for screening if they had a diagnosis of type 1 or type 2 diabetes and were 19 years or older. Data from the county health system’s administrative database were obtained from January 26, 2012 (date of first enrollee), through May 1, 2015, to establish participants’ eye care use in the ophthalmology clinic after screening.
Main Outcomes and Measures
Adherence to the recommended interval of follow-up eye appointments in the facility’s ophthalmology service as determined by administrative records, as well as factors associated with adherence.
Diabetic retinopathy screening was completed in 949 adults with diabetes, of whom 84.5% (802 of 949) were African American, 64.5% (612 of 949) were women, and 71.7% (680 of 949) lacked health insurance. Participants ranged in age from 21 to 95 years, and their mean (SD) age was 53.9 (10.4) years. The mean (SD) age at diabetes diagnosis was 44.3 (12.5) years, and the mean (SD) duration of diabetes was 9.6 (9.4) years. Across interval recommendation types, 29.9% (284 of 949) adhered to obtaining comprehensive follow-up eye care within the recommended time frame. Two years after a participant’s screening date, 50.9% (483 of 949) had no record of having received eye care. Factors associated with adhering to interval appointments were having an advanced age (odds ratio, 1.02; 95% CI, 1.01-1.04) and knowing one’s glycated hemoglobin level (odds ratio, 2.00; 95% CI, 1.34-2.97). Agreeing to assistance in making a follow-up eye care appointment was associated with nonadherence (odds ratio, 0.67; 95% CI, 0.45-0.99).
Conclusions and Relevance
After a DR screening program in a public clinic largely serving an African American population, only one-third of participants adhered to interval recommendations for follow-up eye appointments, even though cost and accessibility were minimized as barriers to care. Our findings suggest that DR screening programs are not likely to meet their public health goals without incorporation of eye health education initiatives successfully promoting adherence to recommended comprehensive eye care for preventing vision loss.
The prevalence of type 1 and type 2 diabetes in the United States is 9.3% and is expected to increase, along with complications associated with diabetes, including diabetic retinopathy (DR).1-3 Diabetic retinopathy is the leading cause of new cases of blindness among adults aged 20 to 74 years old in the United States4 and is estimated to affect more than one-fourth of patients with diabetes.1 Routine dilated comprehensive eye examination remains a cornerstone in the early detection of DR, creating the opportunity for early intervention and management.5 The American Diabetes Association5 and the American Academy of Ophthalmology Retina Panel6 recommend routine, annual dilated eye examination for persons with diabetes, beginning 5 years after diagnosis for type 1 diabetes and at the time of diagnosis and annually thereafter for type 2 diabetes. Estimates indicate that only two-thirds of individuals with self-reported diabetes receive an annual dilated eye examination.7 Annual use estimates are even lower among African American individuals with diabetes, with studies8,9 indicating that fewer than half of African American adults with diabetes receive an annual examination. Low rates of eye care use in this population have been attributed to reduced accessibility, scarcity of eye care professionals in some communities, and cost.10-13 African American individuals reportedly receive a lower quality of diabetes care.14-16 These data are concerning because the prevalence of vision-threatening DR is 190% higher among African American individuals than among non-Hispanic white individuals.17
A growing body of evidence indicates that implementation of DR screening programs leads to more persons with diabetes receiving retinal screening, lower rates of sight-threatening DR in the future, and reduced incidence of blindness.18-21 Screening programs that use nonmydriatic cameras and telemedicine reading centers have established efficacy in detecting DR compared with the criterion standard of dilated fundus examination22,23 and have the advantage of being brief, cost-effective, and less burdensome to patients and conveniently based within primary care.24-28 However, a key element of DR screening programs is whether participating patients actually adhere to the timetable of follow-up comprehensive eye care recommended by the screening program. Recently, our group implemented a DR screening program using a nonmydriatic camera in a publicly funded county health system that primarily serves African American patients.29 Herein, we report on the extent to which patients complied with the interval recommendation for follow-up eye care in the county health system’s eye clinic (Cooper Green Mercy Health Services’ ophthalmology clinic), as well as factors that were associated with eye care use.
The institutional review board of The University of Alabama at Birmingham approved the study protocol, which followed the tenets of the Declaration of Helsinki. Informed consent was orally provided by participants. The DR screening program was based in an internal medicine clinic (Cooper Green Mercy Health Services’ internal medicine clinic) located in a county-operated safety-net health care facility serving Jefferson County, Alabama. County residents have access to services at the clinic regardless of the ability to pay or insurance status. Patients with physician appointments between January 26 and July 24, 2012, were eligible for screening if they had a diagnosis of type 1 or type 2 diabetes and were 19 years or older. All patients meeting these criteria were screened if they agreed to participate in the screening program and there was sufficient time during their clinic visit for the screening. Potential enrollees were informed that this participation was a screening program and not a replacement for a comprehensive eye examination by an ophthalmologist. Before screening, participants completed an interviewer-administered questionnaire requesting contact information, demographics, age when first told by a physician that they had diabetes, smoking status, and health insurance status, as well as whether they knew their glycated hemoglobin level and when they had received their most recent dilated eye examination. Visual acuity for distance with habitual correction (if they wore one) was assessed for each eye separately using an instrument (Titmus Vision Screener V2; Honeywell Safety Products). It was noted whether the patient wore eyeglasses during visual acuity testing.
Methods for fundus imaging and remote grading of the images have been described in detail previously29 and are summarized herein. Ocular imaging was performed by a trained technician using a nonmydriatic camera with autofocus (model AFC-230; Nidek Inc). Three photographs were obtained per eye, including anterior segment, nasal fundus, and temporal fundus. Images were either graded by the Wills Eye Hospital Reading Center or The University of Alabama at Birmingham Callahan Eye Hospital Clinic, with ratings provided by fellowship-trained retina specialists (C.D.W. and J.A.H.). Raters evaluated images using the United Kingdom’s National Health Service’s DR grading classification system.30 Other ocular findings were also noted. Screening results were summarized in a report that was sent to the program’s clinical coordinator.
The coordinator sent a letter to participants describing the results and recommended follow-up eye care for a comprehensive eye examination, and the results were also provided to the participant’s physician in the internal medicine clinic. Recommendations were derived from the American Academy of Ophthalmology’s guidelines for DR follow-up31 and were based on the presence and degree of DR (Table 1). At the time we implemented this screening program, it was unknown whether sufficient funds would be available to sustain it for more than 7 months; therefore, returning for a second DR screening in 1 year was not a recommendation we could issue. Participants whose results were R0 (no DR) or R1 (background DR) received letters recommending that they seek an appointment for a dilated eye examination in 1 year. In the case of previous photocoagulation treatment (grade P), patients were also referred for annual examination. However, if participants with R0, R1, or P grades had other ocular findings (eg, glaucomatous optic nerve, cataract, or hypertensive retinopathy), they were recommended for follow-up care within the next 3 months. Participants with grades of R2 (preproliferative DR), R3 (proliferative DR), or M (maculopathy) received a prompt referral (defined as 1 week) to an eye care professional, and the coordinator telephoned the participant within 48 hours of receiving the report from the reading center and informed the participant of the recommendation. A letter was also sent to the participant. Patients with images deemed to be ungradable owing to poor quality were informed that their results could not be read because the photographs were unclear and were advised to obtain a comprehensive dilated eye examination within the next 3 months. If the results differed in the 2 eyes, the result with the most immediate time frame recommendation was the one provided in the letter. Recommendation letters (or telephone calls in the case of prompt referrals) offered patients assistance in making the follow-up eye appointment.
Data from the county health system’s administrative database were obtained from January 26, 2012 (date of first enrollee), through May 1, 2015, to establish participants’ eye care use in the ophthalmology clinic after screening. Using the date when participants were contacted (via letter or telephone) and the recommended timing of follow-up care, each participant was classified as meeting or not meeting the recommended timing for the eye examination.
Using χ2 and t tests, patients who did and did not have a follow-up visit within the recommended period were compared with respect to age, sex, race/ethnicity, visual acuity, age at diabetes diagnosis, duration of diabetes, degree of DR found at screening, knowledge of personal glycated hemoglobin level, smoking status, time since the last dilated eye examination, health insurance status, and desire for help in scheduling a follow-up appointment. Logistic regression was used to evaluate the significant, independent associations between having these characteristics and receiving recommended screening. Two-sided P ≤ .05 was considered statistically significant. Statistical analyses were conducted using a software program (SAS, version 9.3; SAS Institute Inc).
A total of 949 adults with diabetes participated in DR screening. Most (84.5% [802 of 949]) were African American, and two-thirds (64.5% [612 of 949]) were women (Table 2). Participants ranged in age from 21 to 95 years, and their mean (SD) age was 53.9 (10.4) years. The mean (SD) age at diabetes diagnosis was 44.3 (12.5) years, and the mean (SD) duration of diabetes was 9.6 (9.4) years. Approximately one-fourth (28.6% [271 of 949]) of participants reported smoking or using other tobacco products. Most participants (71.7% [680 of 949]) did not have any type of health insurance. With respect to recent eye care use, 50.8% (482 of 949) of participants reported having a dilated eye examination within the past year and 65.5% (622 of 949) within the past 2 years. Once they knew their screening results, most participants (85.9% [815 of 949]) indicated that they wanted assistance with making an appointment for follow-up eye care in the clinic’s ophthalmology service. Better-eye visual acuity was worse than 20/40 for 8.6% (77 of 896) of participants, and worse-eye visual acuity was worse than 20/40 for 27.8% (249 of 896) of participants. A distance correction was used during acuity screening for 31.6% (300 of 949) of participants.
Table 3 summarizes the extent to which patients in the screening program adhered to the timetable of follow-up comprehensive eye care recommended by the screening program. Of the 112 patients who received a referral for prompt and urgent evaluation (within 1 week) by an ophthalmologist, 20 actually received the care. A total of 158 patients had other ocular findings yet no DR or background DR or a history of photocoagulation treatment. These patients were recommended to seek care within the next 3 months, and 22.2% (35 of 158) received it. For the 537 patients who had grades of R0, R1, or P but no other ocular findings, 35.2% (189 of 537) of them received follow-up eye care within the 1-year recommended interval. Images in 1 or both eyes were ungradable for 15% (142 of 949) of patients, and 28.2% (40 of 142) of these patients received follow-up care within the recommended 3 months. Across interval recommendation types, 29.9% (95% CI, 27.0%-32.8%) of the 949 patients adhered to comprehensive follow-up eye care within the recommended interval. Overall, 49.1% (466 of 949) of the patients received an examination within 2 years after their screening visit, and 50.9% (483 of 949) of the patients did not receive any eye care for the subsequent 2 years after screening.
Table 4 compares patients who did and did not receive eye care within the recommended interval with respect to many variables. Eye care use within the recommended follow-up period was unrelated to sex, race/ethnicity, age at diabetes diagnosis, duration of diabetes, last dilated eye examination, smoking or other tobacco product use, health insurance status, or visual acuity. Patients receiving follow-up eye examinations at recommended intervals were more likely to be older, have an older age at diabetes diagnosis, and know their glycated hemoglobin level and were less agreeable to assistance in making a follow-up appointment. In a multivariable model, having an advanced age (odds ratio, 1.02; 95% CI, 1.01-1.04) and knowing one’s glycated hemoglobin level (odds ratio, 2.00; 95% CI, 1.34-2.97) were associated with an increased odds of receiving follow-up eye care, while agreeing to assistance in making a follow-up eye care appointment (OR, 0.67; 95% CI, 0.45-0.99) was associated with a decreased odds.
A critical factor required for the success of DR screening programs from a public health perspective is whether patients adhere to recommendations for follow-up care, including complying with the interval of the recommendation. We report on a DR screening program based in a county-funded, safety-net internal medicine clinic that primarily serves uninsured African Americans in terms of whether patients with diabetes follow interval recommendations for eye examination. African American individuals are among those at highest risk for DR in the United States. Our results indicate that only 29.9% (95% CI, 27.0%-32.8%) of the 949 patients screened received eye care within the recommended interval. This low adherence rate to the interval recommendation for examinations is striking considering that several often-cited barriers to eye care10-13 were minimized. The county clinic had its own ophthalmology service at the same location as the internal medicine service where patients access care for diabetes; therefore, patients were familiar with the location. Examinations were low cost or free. At this county-operated clinic, no one seeking care is turned away if he or she is a registered patient at the facility, which was the case for all of our enrollees because they were already being seen in the internal medicine service. The screening coordinator offered assistance in making the appointment. Cost and accessibility have been cited as major barriers to eye care adherence by patients with diabetes in surveys,32 but these variables were greatly minimized in our setting. Even when the cost of the examination is covered, some aspects of cost, such as lost wages on the day of the follow-up appointment for those employed, would remain unaddressed. Younger adults are more likely to be employed and were also less likely to be adherent to interval recommendations for follow-up care.
Other factors should be considered as contributing to the high rate of nonadherence observed. Patients in our screening program may have been unaware of their high-risk status for vision loss and of the preventive importance of dilated eye care at regular intervals. If they lacked this knowledge, they may not have recognized the need for eye care at recommended intervals. Many persons with diabetes, particularly those from underserved and minority populations, have gaps in their knowledge about diabetic eye complications.32-36 Eye health education targeted at adults with diabetes can have a positive benefit. A program that provided eye health educational resources (booklet, video, or telephone counseling) in a clinic where eye care services were available doubled the eye examination rate among African American patients.37 The results of that study and the present study suggest that participating in a DR screening program alone is not a sufficient eye health educational experience to prompt high adherence rates to follow-up recommendations for comprehensive eye care. Therefore, it is probably unrealistic to expect DR screening initiatives on their own, without the inclusion of an educational initiative, to lead to high rates of adherence for follow-up care.
One hundred twelve patients had DR screening results that merited urgent attention (preproliferative or proliferative disease or macular edema), but only 17.9% (20 of 112) sought care within 7 days of receiving their screening results. Of these, 24.1% (27 of 112) still had no record of an eye care visit to the ophthalmology service 2 years after their screening visit. Patients with no DR or only background DR with no other ocular findings adhered to recommended intervals at the rate of 35.2% (189 of 537). The reasons underlying their greater compliance with recommended eye care compared with those with more severe DR remain unknown. Perhaps they were in better health with respect to their diabetes and more likely to attend medical appointments, although these data are unknown. Diabetes complications, such as DR, are less likely in persons with good glycemic control.38
It is important to consider factors related to adherence to recommended interval examinations because they could inform possible ways to enhance the effectiveness of DR screening programs. Knowledge of one’s glycated hemoglobin level has been associated with adherence to physician recommendations,39 which also could underlie the greater likelihood of attending recommended eye examinations. Agreeing to accept the screening coordinator’s assistance in making the recommended follow-up appointment decreased the likelihood that a patient would attend the appointment. This finding may be counterintuitive because best practices in DR screening programs incorporate providing the patient with assistance in making follow-up appointments. Perhaps those not wanting assistance in making an appointment are individuals who have taken charge of diabetes self-management and view themselves as not needing help to navigate the appointment system. Younger patients were less likely to comply with obtaining interval examinations, consistent with other studies8,40,41 showing lower eye care use rates among younger adults with diabetes. This result suggests that educational programs on eye health in diabetes focusing on the younger adult population may be particularly needed.8
Our study has strengths and limitations. Strengths include the study setting, which focused on an African American subpopulation at high risk for DR, most of whom lacked health insurance. Because the health facility provided services at no or low cost and an ophthalmology clinic existed physically alongside an internal medicine service within the same building, we were able to examine follow-up appointment adherence rates largely devoid of cost and accessibility, which are 2 of the largest barriers to eye care. Eye care use data were obtained from administrative records (an objective source) and were not provided by patient self-report. A weakness of the study is that in assessing eye care use we defined use as an appointment in the ophthalmology clinic at the county health facility, and it is possible that participants may have sought eye care at private clinics elsewhere. However, persons who make use of this county health care facility tend to rely on its clinics for all of their health care needs. They are registered patients at the facility, services are free or low cost, and the clinic is familiar. Eye care at a private clinic would be substantially more expensive and require documentation of health insurance (impossible for two-thirds of the participants herein). Some patients (15.0% [142 of 949]) had ungradable images in 1 or both eyes. However, all of these patients were advised to see an eye physician for an eye examination within the next few months, and we offered to make appointments for them. With regard to the effect of poor visual acuity on adherence, given that only 8.6% (77 of 896) of our study group had better-eye visual acuity worse than 20/40, further study of a possible association between adherence to treatment recommendations and visual acuity is merited. We were not able to sustain our DR screening beyond the demonstration program described herein due to unavailable funding. However, lessons learned from this program and its evaluation as described in the present study and elsewhere29 serve as a strong motivation to identify funding sources to support a continuation of the program.
In summary, we implemented a DR screening program in a publicly funded internal medicine clinic largely serving African American patients, most without health insurance. Diabetic retinopathy screening programs are increasingly being implemented in the United States because they may be performed with noninvasive cameras, validly reveal diabetes-related retinal changes, and use off-site reading centers for grading images, not requiring on-site ophthalmological expertise. However, the public health success of DR screening programs depends on patients’ adherence to recommended follow-up eye care. In our setting, only one-third of patients screened were adherent to receiving eye care in the recommended interval, and one-half still had not received care 2 years after they were screened, even though barriers to eye care (namely, cost and accessibility) were minimized. Previous work has highlighted the inadequate understanding many patients with diabetes have about the importance of being adherent to routine eye care as a strategy for preventing serious vision loss. Integration of eye health education within the formal structure of DR screening programs, including those that rely on telemedicine, could lead to better adherence to follow-up appointments, an issue worthy of further investigation.
Corresponding Author: Cynthia Owsley, PhD, Department of Ophthalmology, The University of Alabama at Birmingham, 700 S 18th St, Ste 609, Birmingham, AL 35294 (email@example.com).
Accepted for Publication: July 10, 2016.
Published Online: September 15, 2016. doi:10.1001/jamaophthalmol.2016.3081
Author Contributions: Drs McGwin and Owsley had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Keenum, Owsley.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Keenum, McGwin, Witherspoon, Clark, Owsley.
Critical revision of the manuscript for important intellectual content: Keenum, McGwin, Haller, Owsley.
Statistical analysis: McGwin.
Obtaining funding: McGwin, Owsley.
Administrative, technical, or material support: Witherspoon, Haller, Clark, Owsley.
Study supervision: McGwin, Clark, Owsley.
No additional contributions: Keenum.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Haller reported relevant financial activities outside of the submitted work, including serving on the advisory board for the Johnson & Johnson data and safety monitoring board, Neurotech data and safety monitoring board, KalVista, Merck, and Regeneron and serving on the board of directors for Calgene. Dr Owsley reported being the principal investigator on a cooperative agreement from the Centers for Disease Control and Prevention to her institution (The University of Alabama at Birmingham) and reported receiving research funding through her institution from The EyeSight Foundation of Alabama and Research to Prevent Blindness. No other disclosures were reported.
Funding/Support: This research was funded by cooperative agreement 5U58DP002651 from the Centers for Disease Control and Prevention. Supplemental funding was provided by Research to Prevent Blindness, The EyeSight Foundation of Alabama, and the Buck Trust.
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Disclaimer: The findings and conclusions in this article are those of the authors and do not necessarily reflect the official position of the Centers for Disease Control and Prevention.
Create a personal account or sign in to: