Diabetic retinopathy in African–Americans with end-stage kidney disease: a cross-sectional study on prevalence and impact on quality of life

Abstract

Introduction

The prevalence, severity, and quality of life (QoL) impact of diabetic retinopathy (DR) among African–Americans (AAs) with end-stage kidney disease (ESKD) undergoing dialysis are unknown.

Research design and methods

A cross-sectional study was conducted on 93 AA adults with diabetes and ESKD. The diagnosis of DR was based on a review of medical records and/or a positive photograph with a portable hand-held device reviewed by both artificial intelligence software and a retinal specialist. QoL, physical disability social determinants of health (SDoHs) were assessed by standardized questionnaires.

Results

The prevalence of DR was 75%, with 33% of participants having mild, 9.6% moderate and 57.4% severe DR. A total of 43% had normal visual acuity; 45% had moderate visual impairment; and 12% had severe visual impairment. We found a high burden of disease, multiple SDoH challenges, and low QoL and general health among patients with ESKD. The presence of DR had no significant impact on physical health and QoL compared with participants without DR.

Conclusions

DR is present in 75% of AA patients with diabetes and ESKD on haemodialysis. ESKD has a significant burden on general health and QoL; however, DR has a minor additional impact on the overall physical health and QoL in people with ESKD.

Correspondence to Professor Guillermo E Umpierrez; [email protected]

WHAT IS ALREADY KNOWN ON THIS TOPIC

• This study reports on the prevalence, severity, and quality of life (QoL) impact of diabetic retinopathy (DR) among African–Americans (AA) with end-stage kidney disease (ESKD) receiving dialysis are unknown. The diagnosis of DR was based on a review of medical records and/or a positive photograph with a portable hand-held device reviewed by both artificial intelligence software and a retinal specialist.

WHAT THIS STUDY ADDS

• The prevalence of DR was 75%, with 33% of subjects having mild, 9.6% moderate, and 57.4% severe DR. We found a high burden of disease, multiple social determinants of health challenges, and low QoL and general health among patients with ESKD. The presence of DR had no significant impact on physical health and QoL compared with subjects without DR.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• DR is common and reported in 75% of AA patients with diabetes and ESKD on haemodialysis. ESKD has a significant burden on general health and QoL; however, DR has a minor additional impact on the overall physical health and QoL in people with ESKD.

Introduction

Diabetes is the leading cause of legal blindness and end-stage kidney disease (ESKD) in working-age adults worldwide.^{1 2} Diabetic retinopathy (DR) affects 7.7 million Americans, and that number is projected to increase to more than 11 million people by 2030.³ Recent retrospective studies have reported that non-Hispanic black patients were 2.6 times more likely to have significant retinopathy and vision loss compared with non-Hispanic white patients.^{4 5} Similarly, strong racial and ethnic health disparities are reported in chronic kidney disease and the need for dialysis in the USA.⁶ African–Americans (AAs) are four times more likely than white Caucasians to develop chronic kidney disease⁷ and are more than three times more likely to require renal replacement therapy than non-Hispanic white patients.^{8 9}

Visual disturbance and DR are common in patients with ESKD, with a reported prevalence of 45%–73% in patients with type 2 diabetes on dialysis.¹⁰ Watanabe et al¹⁰ reported on the long-term outcomes of 268 patients with diabetic ESKD treated with haemodialysis and found that 73.1% of patients had some disturbance in visual acuity and 11.3% lost light perception. Whether dialysis influences the status and severity of DR and macular oedema remains controversial,¹¹ with long-term follow-up studies reporting that the degree of DR might be either improved or stabilized after initiation of renal replacement therapy.^{12 13} In addition, the impact of DR on quality of life (QoL) and the effects of social determinants of health (SDoHs) in non-Hispanic AAs with ESKD are not known. Accordingly, we aimed to fill this important knowledge gap on the prevalence of DR, as well as its impact on QoL in AA patients with ESKD on haemodialysis treatment.

Methods

This cross-sectional study was conducted at the Emory University dialysis centres in Atlanta, Georgia. We included adults aged 18–80 years old who self-identified as non-Hispanic AAs, had a diagnosis of type 1 or type 2 diabetes, and ESKD treated with haemodialysis for ≥6 months before enrolment. We excluded individuals with mental conditions rendering them unable to understand the nature, scope, and possible consequences of the study, and vulnerable individuals such as pregnant women and prisoners.

The primary objective was to determine the prevalence of DR among AAs with diabetes and ESKD on dialysis. We collected demographic information (age, sex, body mass index (BMI), and diabetes type), duration of diabetes, the current treatment of diabetes, duration of ESKD and dialysis, metabolic control (most recent glycated hemoglobin (HbA1c)), comorbidities (hypertension, cardiovascular disease, heart failure, peripheral neuropathy, cerebrovascular disease, peripheral vascular disease, depression, and cognitive impairment). We assessed their visual acuity and the impact of visual disturbance on the restriction of participation in daily activities using the Impact of Vision Impairment questionnaire,¹⁴ as well as the presence of SDoH factors by validated questionnaires.⁶

The diagnosis of DR was confirmed by a documented ophthalmology report in the electronic health records and/or by the presence of DR with a portable hand-held camera (Optomed Aurora retinal camera), and results were reviewed by both artificial intelligence (AI) software (AEYE Health’s AI-based retinal screening system (AEYE-DS)) and by a retinal specialist at Emory University. AEYE Health’s AI-based retinal screening system is integrated into Optomed handheld fundus camera Aurora.^{15 16} Retinal photographs were taken by the study team before a dialysis session in a normally lighted room. If photographs could not be read by the software due to poor quality, new photographs were reattempted either by placing the patients in a darker room after their dialysis session or by retaking the photograph with the use of dilating eye drops. Both the OPTOMED handheld camera and AEYE Health’s AI-based retinal screening system are Food and Drug Administration (FDA)-approved devices.^{15 16} The images were saved in the camera memory card and graded as no apparent, mild non-proliferative, moderate non-proliferative, severe non-proliferative, and proliferative DR.

Visual acuity was assessed using the 3 m Snellen chart V.2,¹⁷ with letters/numbers with ‘tumbling Es’. Snellen charts were hand-held 36 cm from the participant’s eye level. The participants were asked to start at the top of the chart and read the smallest letter they can see. All acuities were recorded without corrected lenses. Visual acuity was classified into three groups: normal acuity (≥20/50), moderate visual impairment (≥20/200 and <20/50), and severe visual impairment (<20/200).¹⁷

The impact of vision impairment on a person’s functioning and independence was measured through the validated Vision Impact Score, a self-reported measure developed for individuals with low vision to measure the impact of visual impairment on a person’s functioning and independence.¹⁸ It includes 32 questions assessing how eyesight deficiency interferes with a different aspect of a person’s life. The survey results were divided into five subscales representing leisure and work, social and consumer interactions, household and personal care, mobility, and emotional reaction to vision loss.

Kidney Disease Quality of Life (KDQOL)-36 was assessed using a self-reported disease-specific measure developed for individuals with kidney disease and those on dialysis.¹⁹ It includes 36 kidney disease-targeted items, such as the effects of the disease on activities of daily living, work status, social interaction, physical health, and mental health. The KDQOL is divided into four subscales: physical/mental health, the burden of kidney disease (BKD), symptoms/problems, and effects of kidney disease (EKDs).^{20 21}

Standardized questionnaires were used to collect information on SDoH variables including health insurance, employment status, personal and household income, marital status, and housing. Depression screening was performed using the Patient Health Questionnaire (PHQ)-9, which scores each of the nine Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, criteria as ‘0’ (not at all) to ‘3’ (nearly every day).²² Participants completed all questionnaires themselves; if unable to visualize or independently write down responses, a clinical research coordinator collected information by verbally reading the questions and writing the answers.

Statistical analysis

All characteristics and study outcomes of interest were summarized using counts and/or percentages if they are categorical and means±SD if they are continuous. The primary endpoint of this study was to determine the prevalence of DR in AA patients with diabetes and end-stage kidney disease (ESKD). The prevalence of DR was computed as the proportion of patients with DR in this study cohort. The KDQOL survey was calculated according to the procedures set forth by the Rand Corporation and standardized by UCLA. The KDQOL survey was divided into categories of physical component summary, mental component summary and QoL related specifically to ESKD (BKD, symptoms and problems of kidney disease and EKDs. We compared clinical characteristics and outcomes such as demographics, duration of diabetes, and KDQOL survey outcomes between patients with and without DR using non-parametric Wilcoxon tests for continuous variables and χ² tests (or Fisher’s exact tests) for categorical variables. A two-sided p value less than 0.05 was considered significant. Statistical analysis was performed using SAS V.9.2.

Results

From August 2021 to April 2022, we approached 146 eligible participants at four dialysis centres affiliated with Emory University (figure 1). A total of 100 participants consented to participate; of them, 7 had no information on retinopathy either by electronic medical records or readable fundus imaging and were excluded from further analysis. Among the 93 included participants with retinopathy information, 70 patients (75%) had a positive diagnosis of DR by previous ophthalmology report or by positive retinal photograph with AEYE AI report. Review of ophthalmology reports in medical records revealed that 59 patients had a positive diagnosis of DR; 14 had no evidence of DR; and 20 had no previous report in Electronic medical records (EMR) at either Emory Healthcare or Grady Health System. Among the 93 patients examined by retinal photographs with handheld camera and AEYE Health’s AI system, 40 patients had evidence of DR; 23 patients had no evidence of DR; and 30 patients had poor-quality photographs. Over 90% of participants reported having had an ophthalmology appointment (61% within 1 year, 20% within 2 years, and 14% more than 2 years) without difference between groups. Among patients with a diagnosis of DR, 32 (46%) individuals were aware or recalled having a diagnosis of DR (figure 2A,B).

Figure 1. Study flow. AI, artificial intelligence; DR, diabetic retinopathy.

Figure 2. (A) Visual acuity in patients with end-stage kidney disease with and without DR. (B) DR and impact on kidney quality of life. DR, diabetic retinopathy.

The clinical and sociodemographic characteristics of patients with and without DR are shown in table 1. There were 9 (10%) patients with type 1 and 84 (90%) with type 2 diabetes with a median duration of diabetes of 16.0 (IQR 10.0–25.0) years and haemodialysis of 3.0 (2.0–6.0) years. Patients with positive DR were younger (57.8±11 vs 62.9±14 years, p=0.04) and had a longer duration of diabetes (median 19.0 (10.0–25.0) vs 10.0 (6.0–20.0) years, p=0.013) and were more likely to be treated with insulin and/or oral antidiabetic agents (p=0.023) compared with those without DR. Otherwise, there were no differences between groups on sex, BMI, duration of dialysis, mean HbA1c, hospitalizations during the prior 1 year or cardiovascular complications. Eighty-seven per cent had a history of hypertension; 12% had a prior history of lower extremity amputations; 33% had coronary heart disease; and 32% had a history of heart failure (table 1).

Table 1

Clinical characteristics

Variable	(−) DR (n=23)	(+) DR (n=70)	P value
Age	62.9±14	57.8±11	0.043
Female/male sex, n (%)	11 (48)/12 (52)	35 (50)/35 (50)	0.86
BMI, kg/m2	32.5±10	31.6±8	0.92
Type of diabetes, n (%)			0.44
Type 1	1 (4.3)	8 (11)
Type 2	22 (96)	62 (89)
Duration dialysis (years), median (Q1–Q3)	10.0 (6.0–20.0)	19.0 (10.0–25.0)	0.013
HbA1c (%)	5.92±1.05	6.51±1.15	0.10
Diabetes treatment, n (%)			0.023
Insulin	6 (26)	36 (51)
Oral antidiabetic agents	0 (0.0)	6 (8.6)
Insulin and oral agents	1 (4.3)	4 (5.7)
No pharmacological treatment	16 (70)	24 (34)
Hospitalizations during the past year, n (%)	1.55±1.37	1.46±2.10	0.34
Selected comorbidities, n (%)
Coronary heart disease	6 (26)	25 (36)	0.45
Heart failure	9 (41)	20 (29)	0.31
Hypertension	21 (91)	60 (86)	0.72
Lower extremity amputation	1 (4.3)	10 (14)	0.28
DR diagnosis, n (%)
EMR or retinal photograph		70 (100)
EMR		59 (84)
Retinal photograph		43 (61)
DR severity, n (%)
Mild	0 (0)	20 (28.5)
Moderate	0 (0)	6 (8.5)
Severe and proliferative	0 (0)	35 (50)
DR treatment, n (%)
Laser		16 (23)
Intraocular injection		14 (20)
Laser and injection		8 (11)
Last ophthalmology visit (years), n (%)			0.14
<1	14 (61)	43 (61)
1–2	1 (4.3)	14 (20)
≥2	7 (30)	10 (14)
Visual acuity, n (%)			0.34
Normal	13 (57)	26 (38)
Moderate impairment	8 (35)	33 (49)
Severe–profound impairment	2 (8.7)	9 (13)

BMI, body mass index; DR, diabetic retinopathy; HbA1c, glycated hemoglobin.

Visual acuity was obtained in 91 out of the 93 patients and in 68 of 70 with a positive diagnosis of DR (figure 1). A total of 39 (43%) participants had normal visual acuity, 30 (33%) had moderate visual impairment and 22 (24%) had severe to profound visual impairment (table 1). Individuals with positive DR had higher rates of severe to profound visual impairment (29% vs 8.7%, p=0.085) and lower rates of normal acuity (38% vs 57%, p=0.34) compared with those without DR. In addition, patients with DR more frequently reported poor eyesight compared with patients without DR (54% and 30%, p=0.06).

Self-reported perceptions about the effect and burden of ESKD and DR on general health and QoL are shown in table 2. Patients with ESKD reported a significant burden on general health, in particular, kidney disease interference with social and daily activities, frustration with kidney disease and the time spent dealing with kidney disease. In contrast to kidney disease, the presence of DR had a minor impact on overall health with less than 20% of patients reporting eyesight interference with life or daily activities.

Table 2

Burden of disease on general health and quality of life

	All	(−) DR	(+) DR	P value
General health, n (%)				0.60
Excellent	1 (1.1)	1 (4.3)	0 (0)
Very good	7 (7.5)	2 (8.7)	5 (7.1)
Good	36 (39)	8 (35)	28 (40)
Fair	37 (40)	9 (39)	28 (40)
Poor	12 (13)	3 (13)	9 (13)
Physical health interference with social activities, n (%)				0.43
All the time	16 (17)	3 (13)	13 (19)
Most of the time	13 (14)	1 (4.3)	12 (17)
Some of the time	14 (15)	4 (17)	10 (14)
Little of the time	10 (11)	4 (17)	6 (8.6)
None of the time	40 (43)	11 (48)	29 (41)
Kidney disease interference with life, n (%)				0.22
Definitively true	25 (27)	4 (17)	21 (30)
Mostly true	30 (32)	5 (22)	25 (36)
Don’t know	4 (4.3)	1 (4.3)	3 (4.3)
Mostly false	13 (14)	5 (22)	8 (11)
Definitively false	21 (23)	8 (35)	13 (19)
Too much time dealing with kidney disease, n (%)				0.41
Definitively true	23 (25)	5 (22)	18 (26)
Mostly true	26 (28)	4 (17)	22 (31)
Don’t know	3 (3.2)	0 (0)	3 (4.3)
Mostly false	18 (19)	6 (26)	12 (17)
Definitively false	23 (25)	8 (35)	15 921)
Feels frustrated with kidney disease, n (%)				0.71
Definitively true	14 (15)	5 (22)	18 (26)
Mostly true	32 (34)	6 (26)	26 (37)
Don’t know	2 (2.2)	0 (0)	2 (2.9)
Mostly false	17 (18)	5 (22)	12 (17)
Definitively false	28 (30)	7 (30)	21 (30)
Feels like burden on family, n (%)				0.73
Definitively true	12 (13)	2 (8.7)	10 (14)
Mostly true	23 (25)	6 (26)	17 (24)
Don’t know	2 (2.2)	1 (4.3)	1 91.4)
Mostly false	9 (9.7)	3 913)	6 (8.6)
Definitively false	47 (51)	11 (48)	36 (51)
Self-reported eyesight, n (%)				0.26
Excellent		1 (4.3)	3 (4)
Good		15 (65)	29 (42)
Poor		4 (17)	27 (39)
Bad		3 (13)	8 (12)
Blind		0 (0)	2 (3%)
Eyesight interference with life, n (%)
Not at all	46 (49)	15 (65)	31(44)
Rarely	12 (13)	0 (0)	12 (17)
Little	10 (11)	4 (17)	6 (8.6)
Fair amount	8 (8.6)	2 (8.7)	6 (8.6)
A lot	5 (5.4)	1 (4.3)	4 (5.7)
All the time	12 (13)	1 4.3)	11 (16)
Depressed mood, n (%)				0.61
Minimal		9 (41)	31 (44)
Mild		8 (36)	25 (36)
Moderate		5 (23)	8 (11)
Moderately severe		0 (0)	4 (5.7)
Severe		0 (0)	2 92.9)

There were no differences in SDoH variables among patients with and without DR regarding health insurance, employment, personal or household income, marital status, and housing (table 3). Most participants had public health insurance and were disabled or retired. Both groups reported low personal and household incomes, with 42% and 37% of patients with and without DR meeting the poverty threshold, respectively (p=0.78). About one-third of the participants owned a house, were married and had education below the high school level. Overall, the mean depression PHQ-9 score fell in the mild depression category without differences among patients with and without DR.

Table 3

Social determinants of health variables

	(-) DR (n=23)	(+) DR (n=70)	P value
Health insurance
Private, n (%)	8 (35)	23 (33)	1.00
Public, n (%)	19 (83)	63 (90)	0.46
Employment status, n (%)			0.44
Employed, n (%)	1 (4.5)	13 (18)
Disabled, n (%)	14 (64)	36 (51)
Retired, n (%)	7 (32)	16 (23)
Annual personal income, USD ($)			0.55
<20 000, n (%)	10 (43)	36 (51)
20 000–50,000, n (%)	9 (39)	16 (23)
>50 000, n (%)	0 (0)	6 (8.5)
Annual household income			0.72
<20 000, n (%)	6 (27)	24 (36)
20 000–50,000, n (%)	7 (32)	15 (23)
>50 000, n (%)	2 (8)	8 (11)
Highest education degree			0.039
< High school, n (%)	12 (52)	28 (38)
High school/College, n (%)	8 (35)	36 (51)
Graduate School, n (%)	2 (8)	6 (8)
Marital status			1.00
Single, n (%)	8 (36)	23 (33)
Married, n (%)	8 (36)	23 (33)
Divorced/separated, n (%)	6 (2.6)	24 (34)
Housing
Own, n (%)	8 (35)	21 (30)
Rent, n (%)	14 (61)	45 (64)
Homeless, n (%)	1 (4.3)	2 (2.9)
Other, n (%)	0 (0)	2 (2.9)
Food Security			1.00
High	17 (74)	53 (76)
Low	4 (17)	11 (16)
Very Low	2 (8.7)	6 (8.6)
Substance Abuse Score	5.74±2.70	5.72±2.83	0.81
Vision Impact Score	34.65±41.62	50.77±47.56	0.20
Depression Score	5.82±4.26	6.01±5.55	0.80
Discrimination?

Discussion

This cross-sectional study aimed to determine the prevalence and the impact of DR on general health and QoL among AA patients with ESKD undergoing dialysis treatment. Our results indicate an overall prevalence of DR in 75% of AA with ESKD, with about one-third of them having mild, moderate and severe DR. Assessment of visual acuity indicated that over half of patients with ESKD had significant visual impairment, with positive DR associated with greater visual impairment. We found that ESKD had a significant burden on general health and QoL. In contrast to kidney disease, the presence of DR had no additional impact on overall health with few patients reporting eyesight interference with life or daily activities.

The epidemiology and burden of DR have been determined mainly in populations lacking robust minority groups. The reported rates of DR and vision loss are higher among non-Hispanic black patients and Hispanic patients compared with Caucasian patients.^{4 5} According to a National Eye Institute report, more than 800 000 AAs have DR, and this number is projected to increase to approximately 1.2 million people by 2030.²³ In a retrospective study, Varma et al⁴ reported that non-Hispanic black patients were more likely to have significant retinopathy than non-Hispanic white patients (OR 2.64, 95% CI 1.19 to 5.84). Similarly, the Veterans Administration Diabetes Trial²⁴ reported that the odds of having clinically significant DR and macular oedema was 2.30 (95% CI 1.33 to 4.00) for AA patients compared with non-Hispanic white patients. In our community-based sample of AA adults with ESKD on dialysis, the DR prevalence was higher than previously reported rates of 45%–50% in general dialysis populations.¹⁸ Our study used a new technology, the Optomed Aurora portable camera, which had a reported 91.9% to 96.9% sensitivity and 93.6% to 94.8% specificity in recognizing DR in non-dialysis individuals.^{16 25} We used AEYE Health’s solution that takes the retinal image with results analyzed within 1 min using AI with 93% sensitivity and 91.4% specificity, which recently received FDA clearance for diagnostic screening system for DR.²⁶ Due to the lighting in the dialysis suite, many patients had insufficient-quality images that required dilation with phenylephrine eye drops. The system runs on a secured Amazon Web Services–Health Insurance Portability and Accountability Act-compliant cloud compatible with most EHRs.

Chronic kidney disease and ESKD represent significant examples of racial and ethnic disparities in health. AA patients are three to five times more likely to require renal replacement therapy compared with non-Hispanic white patients.^{4 24 27} In agreement with previous studies in dialysis populations, we found that ESKD was associated with a significant burden on general health, with most patients reporting that kidney disease had significant interference with personal and family life activities and that they felt frustrated about the time spent dealing with kidney disease treatment.

CKD-related disparities are linked to a variety of clinical, socioeconomic, and cultural, as well as SDoH factors.^{8 28} Nationally, 20% of AAs are below the federal poverty level compared with only 9% of white individuals, only 76% of AAs graduated high school vs 88% of white individuals, and the uninsured rate for AAs in 2016 was 10.5% compared with 6.3% for white individuals.²⁹ In agreement with these reports, most AAs with ESKD in our study had public health insurance, were unemployed or disabled, and were below the personal and family poverty line.

Previous publications have shown a relationship between certain SDoH factors and the prevalence of DR. Some studies have revealed an increased risk of DR for those belonging to minority racial and ethnic groups,^{30 31} for those of lower economic status,^{30 32} and for those who have less than a high school education.^{30 32} A recent cross-sectional study using data from the 2018 Behavioral Risk Factor Surveillance System reported the association between SDoH and DR in patients with type 2 diabetes. Exposure variables included home ownership, marital status, income, healthcare coverage, completed level of education, and urban and rural environments. Using logistic regression analysis calculated ORs and 95% CIs found that Alaskan Native/Native American (OR 2.11, 95% CI 1.14 to 3.90), out of work (OR 2.82, 95% CI 1.62 to 4.92), unable to work (OR 2.14, 95% CI 1.57 to 2.91), did not graduate from high school (OR 1.91, 95% CI 1.30 to 2.79), or only attended college or technical school without graduating (OR 1.42, 95% CI 1.09 to 1.86). To our knowledge, ours is the first study to assess DR and SDoH in AA patients with ESKD. We found no differences in SDoH variables among patients with ESKD with and without DR regarding health insurance, employment, personal or household income, lower education levels and increase DR development, consistent with a previous cross-sectional study¹¹ and inconsistent with others.^{10 12 25}

Our study has several limitations including the relatively small number of participants on haemodialysis, excluding patients on peritoneal dialysis. This cross-sectional single-centre study was conducted in AA patients; thus, the results cannot be extrapolated to Caucasians and other racial and ethnic groups. In this pilot study, we used a new technology, the Optomed Aurora portable camera and AEYE Health’s AI, both with a reported sensitivity and specificity greater than 90%^{16 25 26} in recognizing DR in non-dialysis subjects; however, further studies are needed to establish the accuracy of this technology in the chronic kidney disease population. In addition, larger studies are needed to assess the prevalence and impact of DR on visual impairment and QoL in mixed populations. In addition, prospective longitudinal studies are needed to assess the long-term impact of renal replacement therapy on the progression of DR and the impact of QoL.

In summary, the results of this prospective cross-sectional study indicate a prevalence of DR in 75% of AA with ESKD and dialysis treatment, with one-third of them having severe DR, which was associated with significant visual impairment. Individuals with ESKD on dialysis have a significant burden on general health and QoL, which interfered with social and daily activities. In contrast to kidney disease, we found that the presence of DR had no additional impact on overall health, with few participants reporting eyesight interference with life or daily activities.

We thank the Optomed Aurora Company for the training and use of their retinal camera. We also thank AEYE Health for their online diagnostic platform. These companies had no role in study design, data collection, data interpretation, or manuscript writing.

Data availability statement

Data are available upon reasonable request.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

This study involves human participants and the proposal was reviewed and approved by the Emory Institutional Review Board before study initiation (IRBSTUDY00002826). The participants gave informed consent to participate in the study before taking part.

Footnote

Contributors RLC and GEU wrote the initial draft of the manuscript. ME, RLC, EM, ZEZ, and BM served either as study coordinators screening and randomizing research candidates, collecting data or as clinical providers managing patients daily, including weekends, ensuring successful completion of the study protocol. RJG, PV, TI, GDO'K, reviewed and edited the study proposal and manuscript. LP analyzed the data. GEU wrote the initial research proposal, is the guarantor of this work and, as such, had 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.

Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

Competing interests ME, RLC, EM, ZEZ, BM, GDO'K, JN, RJG, PV, LP, and GEU report no conflicts of interest relevant to this article. PV and GEU are editorial board members of BMJ Open Diabetes Research & Care.

Provenance and peer review Not commissioned; externally peer reviewed.

References

¹ Fong DS, Aiello L, Gardner TW, et al. Retinopathy in diabetes. Diabetes Care 2004; 27 Suppl 1: S84–7. doi:10.2337/diacare.27.2007.s84

² Sabanayagam C, Banu R, Chee ML, et al. Incidence and progression of diabetic retinopathy: a systematic review. Lancet Diabetes Endocrinol 2019; 7: 140–9. doi:10.1016/S2213-8587(18)30128-1

³ 2/5/2021 Cfdcapvaehssvhwcgvvriha;

⁴ Varma R, Bressler NM, Doan QV, et al. Prevalence of and risk factors for diabetic macular edema in the United States. JAMA Ophthalmol 2014; 132: 1334. doi:10.1001/jamaophthalmol.2014.2854

⁵ Rabb MF, Gagliano DA, Sweeney HE. Diabetic retinopathy in blacks. Diabetes Care 1990; 13: 1202–6. doi:10.2337/diacare.13.11.1202

⁶ Silverberg EL, Sterling TW, Williams TH, et al. The association between social determinants of health and self-reported diabetic retinopathy: an exploratory analysis. Int J Environ Res Public Health 2021; 18: 792. doi:10.3390/ijerph18020792

⁷ Brown JS. Addressing the issue: health disparities and social determinants of health among blacks/African Americans in comparison to non-minorities with chronic kidney disease. NNJ 2021; 48: 489. doi:10.37526/1526-744X.2021.48.5.489

⁸ Laster M, Shen JI, Norris KC. Kidney disease among African Americans: A population perspective. Am J Kidney Dis 2018; 72 (5 Suppl 1): S3–7. doi:10.1053/j.ajkd.2018.06.021

⁹ Brown JS, Elliott RW. Social determinants of health: understanding the basics and their impact on chronic kidney disease. NNJ 2021; 48: 131. doi:10.37526/1526-744X.2021.48.2.131

¹⁰ Watanabe Y, Yuzawa Y, Mizumoto D, et al. Long-term follow-up study of 268 diabetic patients undergoing Haemodialysis, with special attention to visual acuity and heterogeneity. Nephrol Dial Transplant 1993; 8: 725–34. doi:10.1093/ndt/8.8.725

¹¹ Takamura Y, Matsumura T, Ohkoshi K, et al. Functional and anatomical changes in diabetic macular edema after Hemodialysis initiation: one-year follow-up multicenter study. Sci Rep 2020; 10. doi:10.1038/s41598-020-64798-4

¹² Sophie R, Lu N, Campochiaro PA. Predictors of functional and anatomic outcomes in patients with diabetic macular edema treated with Ranibizumab. Ophthalmology 2015; 122: 1395–401. doi:10.1016/j.ophtha.2015.02.036

¹³ Cai S, Bressler NM. Aflibercept, Bevacizumab or Ranibizumab for diabetic macular oedema: recent clinically relevant findings from DRCR.Net protocol T. Curr Opin Ophthalmol 2017; 28: 636–43. doi:10.1097/ICU.0000000000000424

¹⁴ Lamoureux EL, Hassell JB, Keeffe JE. The impact of diabetic retinopathy on participation in daily living. Arch Ophthalmol 2004; 122: 84–8. doi:10.1001/archopht.122.1.84

¹⁵ Ruan S, Liu Y, Hu W-T, et al. A new Handheld fundus camera combined with visual artificial intelligence facilitates diabetic retinopathy screening. Int J Ophthalmol 2022; 15: 620–7. doi:10.18240/ijo.2022.04.16

¹⁶ Midena E, Zennaro L, Lapo C, et al. Handheld fundus camera for diabetic retinopathy screening: A comparison study with table-top fundus camera in real-life setting. J Clin Med 2022; 11: 2352. doi:10.3390/jcm11092352

¹⁷ O’Neill S, McAndrew DJ. The validity of visual acuity assessment using mobile technology devices in the primary care setting. Aust Fam Physician 2016; 45: 212–5.

¹⁸ Lamoureux EL, Tai ES, Thumboo J, et al. Impact of diabetic retinopathy on vision-specific function. Ophthalmology 2010; 117: 757–65. doi:10.1016/j.ophtha.2009.09.035

¹⁹ Peipert JD, Bentler PM, Klicko K, et al. Psychometric properties of the kidney disease quality of life 36-item short-form survey (KDQOL-36) in the United States. Am J Kidney Dis 2018; 71: 461–8. doi:10.1053/j.ajkd.2017.07.020

²⁰ Peipert JD, Nair D, Klicko K, et al. Kidney disease quality of life 36-item short form survey (KDQOL-36) normative values for the United States dialysis population and new single summary score. JASN 2019; 30: 654–63. doi:10.1681/ASN.2018100994

²¹ Mapes DL, Bragg-Gresham JL, Bommer J, et al. Health-related quality of life in the dialysis outcomes and practice patterns study (DOPPS). Am J Kidney Dis 2004; 44 (5 Suppl 2): 54–60. doi:10.1053/j.ajkd.2004.08.012

²² Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med 2001; 16: 606–13. doi:10.1046/j.1525-1497.2001.016009606.x

²³ NEIDRAAwDCPVL. Available: https://www.nei.nih.gov/sites/default/files/2019-06/african-americans-diabetes-can-prevent-vision-loss.pdf [Accessed 2 May 2021 ].

²⁴ Emanuele N, Moritz T, Klein R, et al. Ethnicity, race, and clinically significant macular edema in the veterans affairs diabetes trial (VADT). Diabetes Res Clin Pract 2009; 86: 104–10. doi:10.1016/j.diabres.2009.08.001

²⁵ Kubin AM, Wirkkala J, Keskitalo A, et al. Handheld fundus camera performance, image quality and outcomes of diabetic retinopathy grading in a pilot screening study. Acta Ophthalmol 2021; 99: e1415–20. doi:10.1111/aos.14850

²⁶ Available: https://www.fdanews.com/articles/210235-aeye-healths-diagnostic-screening-cleared-for-diabetic-retinopathy

²⁷ Roy MS. Diabetic retinopathy in African Americans with type 1 diabetes: the New Jersey 725: II. Arch Ophthalmol 2000; 118: 105. doi:10.1001/archopht.118.1.105

²⁸ National Center for Health Statistics. Health US, 2015: With Special Feature on Racial and Ethnic Health Disparities. Washington DC: US Government Printing Office, 2016.

²⁹ National Center for Education Statistics. Public high school graduation rates 2018. n.d. Available: https://nces.ed.gov/programs/coe/indicator_coi.asp>

³⁰ Zhang X, Cotch MF, Ryskulova A, et al. Vision health disparities in the United States by race/Ethnicity, education, and economic status: findings from two nationally representative surveys. Am J Ophthalmol 2012; 154 (6 Suppl): S53–62. doi:10.1016/j.ajo.2011.08.045

³¹ Luo H, Bell RA, Garg S, et al. Trends and racial/ethnic disparities in diabetic retinopathy among adults with diagnosed diabetes in North Carolina. N C Med J 2019; 80: 76–82. doi:10.18043/ncm.80.2.76

³² Funakoshi M, Azami Y, Matsumoto H, et al. Socioeconomic status and type 2 diabetes complications among young adult patients in Japan. PLoS One 2017; 12: e0176087. doi:10.1371/journal.pone.0176087