![SEVERE LACK OF CAPACITY
Screening all diabetes patients in OC would use 99% of their 19,566
visits. Removing DR screening from OC (1,506 single visit & no major eye
diagnosis) will barely accommodate treatment only if the eye disease
rate among the unscreened is as low as 5% (688 X 2.1 visits = 1,445).
David C. Ziemer1, Pranay K. Neema2, Alex Mojonnier2, Lauren L. Asp1, Jiani Hu1, Luxi Liang1, Raymond L. Smith2, Allison Pall1, Catherine S.
Barnes1, Jane M. Caudle1
1Emory University School of Medicine, Division of Endocrinology, Metabolism and Lipids & 2Grady Health System, Clinical Decision Support – Atlanta, Georgia
ABSTRACT (617-P)
Early treatment of diabetic retinopathy (DR) reduces vision loss. Nationally, DR screening rates are
inadequate. Meeting DR screening guidelines is a high priority challenge in a public hospital system where
Ophthalmology Clinic (OC) is over-capacity.
To plan for better DR screening, we analyzed 2014 administrative data in the Grady Health System where
several clinics manage diabetes. DR screening was ruled complete for patients with OC, optometry or retinal
photography visits. We surveyed a convenience sample of diabetes patients at a clinic visit.
Among 19,361 persons with diabetes overall screening was low (29%) with wide variation by care site (5 - 66%);
5,000 had no diabetes continuity care visit. In a multivariable model, DR screening increased in diabetes &
primary care clinics and with age (OR: 2.8, 2.1, & 1.03/yr respectively); it decreased with [NON-]Hispanic ethnicity
and mental health diagnoses (OR 0.7 and 0.8) (all p<0.001).
Having in-clinic eye screening doubled DR screening (48% vs 22%) and decreased the number of screens done
in OC (45% vs 96%).
On the patient survey ~70% had eye screens in the prior year – 50% by OC, 25% by Diabetes Clinic and 13%
outside the system. Half of the unscreened said a referral was made. No show rate for screening appointments is
over 40%.
Many screening barriers must be addressed: screening capacity, lack of regular diabetes care, scheduling
processes, referral rates and needs of mental health and Hispanic patients. Better processes are needed to
recruit patients into and maintain regular diabetes care, and to improve appointment keeping and follow-up for
those missing appointments. Moreover, repeated annual screening will require a tracking system. Planned
expansion of teleretinal screening and optometry will increase capacity but also awareness of treatment-
requiring disease. Thus, planning for downstream treatment capacity is needed.
DR screening is a complex problem requiring not just infrastructure investments but also major care
management interventions.
INTRODUCTION
Diabetes is common, costly, and deadly in the US and the Grady
Health System (GHS). Diabetic retinopathy (DR) is a major cause of
vision loss in the US.1 In 2011, the age-adjusted percentage of adults
with diagnosed diabetes reporting visual impairment was 17.6%.2
This is a pressing issue as the number of Americans with DR is expected
to double from 7.7 million in 2010 to 15.6 million in 2050.3 This may have
a greater impact on non-Hispanic Black and Mexican Americans, who
have a higher prevalence of DR than non-Hispanic whites.4
Historically, diabetes has been present in 13% of all GHS patients, 30-
35% of all Primary Care visits, and 24% of GHS charges/cost.
METHODS
To evaluate DR screening, GHS administrative data on 19,361 diabetes
patients in 2014 were analyzed. Patients were eligible based on any
one of the following:
• At least two outpatient diagnoses on separate days in 2014
• One inpatient diagnosis in 2014
• Diabetes medication in 2012-2015
• At least two A1Cs ≥6.5 in 2014
DR screening was considered complete if an Ophthalmology Clinic
(OC), optometry, or retinal photograph visit was attended.
In addition, a convenience sample of 80 adult GHS diabetes patients
were asked about their DR screening in the prior year.
SUMMARY
• Most diabetes patients, 71%, do not receive DR screening
• Health system screening capacity is the major limitation
• In-clinic eye support increases DR screening dramatically
• DR screening is decreased by:
– Lack of continuity care
– Failure to keep appointments
– Failure to refer
– Mental health diagnoses other than depression
CONCLUSIONS
Comprehensive DR screening is a complex, multi-aspect challenge.
The system needs to dramatically increase:
• Capacity for screening and treatment
• The number of diabetes patients with continuity care
It also needs to develop:
• Process for tracking and scheduling those who need DR
screening, including those missing appointments
• Methods to Improve patient and provider engagement,
especially for those with mental health problems
• Consideration for alternative screening methods
ACKNOWLEDGEMENTS
This work was supported by the American Diabetes Association (7-12-HYPO-11).
1.Centers for Disease Control and Prevention (CDC), Division of Diabetes Translation, National Center for Chronic Disease Prevention and Health Promotion. 2015. Why is vision loss a public health problem? Accessed May 25, 2016: http://www.cdc.gov/visionhealth/basic_information/vision_loss.htm
2.CDC, National Center for Health Statistics, Division of Health Interview Statistics, data from the National Health Interview Survey. 2015. Crude and age-adjusted percentage of adults 18 years or older with diagnosed diabetes reporting visual impairment, United States, 1997-2011. Accessed May 13,
2016: www.cdc.gov/diabetes/statistics/visual/fig2.htm
All
(n=19,361)
Screened
(n=5,595)
Not Screened
(n=13,766)
p-value
Mean Diabetes Visits (SD) 1.6 (2.9) 3.1 (3.8) 1.0 (2.0) <0.0005
Mean Eye Visits (SD) 0.6 (1.5) 2.1 (2.2) 0.0 (0.0) <0.0005
Mean A1C (SD)* 7.80 (2.31) 7.75 (2.08) 7.83 (2.42) 0.043
Age in Years (SD) 56.8 (13.7) 60.4 (12.2) 55.4 (14.0) <0.0005
Female 59.4% 62.7% 58.1% <0.0005
African American/Black 83.2% 85.7% 82.2% <0.0005
Non-Hispanic 91.4% 91.2% 91.5% 0.727
Eye Disease Diagnosis 7.2% 17.2% 3.2% <0.0005
Cataract 1.7% 3.8% 0.8% <0.0005
Glaucoma 0.5% 1.3% 0.2% <0.0005
Macular Edema 2.8% 1.3% 0.2% <0.0005
Proliferative Retinopathy 4.1% 10.2% 1.6% <0.0005
Other Mental Health Diagnosis** 46.4% 44.2% 47.6% <0.0005
*values calculated with total n=14,034
**values calculated with total n=19,265
DEMOGRAPHICS
Population was mostly non-Hispanic, African American & female, with
mean age 57 years. Many (26% had no diabetes continuity care visit.
The unscreened had mean of only 1 clinic visit for diabetes care vs 3.1
for those screened.
48%
26%
7% 6%
13%
0%
20%
40%
60%
80%
Ophthalmology
Clinic
Diabetes
Center
"Grady" Neighborhood
Health Center
Outside GHS
Screening Location for 54 Diabetes Patients
Screened in Past Year
B
Values do not sum to total as patients may have visited multiple sites
*Likely includes treatment visits
64%
21% 19%
0%
0%
20%
40%
60%
80%
Ophthalmology
Clinic*
Diabetes Center Neighborhood
Health Centers
Infectious Disease
Center
Screening Location for 5,595 Diabetes Patients
Screened in 2014
SCREENING RATES SUB-OPTIMAL
In administrative data, only 29% of diabetes patients completed a DR
screening in 2014 (A). Patient, in-clinic survey screening rate was higher
than administrative data suggested at 68% (B).
This implies that those in regular care are more likely to be screened.
33%
68%
0%
20%
40%
60%
80%
Not Screened Screened in Past Year
Screening Rates for 80 Adult
Diabetes Patients in Past Year
B
71%
29%
0%
20%
40%
60%
80%
Not Screened Eye Screen in 2014
Screening Rates for 19,361
Diabetes Patients in 2014
A
A
MOST “SCREENING” IN OPHTHALMOLOGY CLINIC
In administrative
data, OC screened
64% of diabetes
patients with DR
screening(A). This
is similar to the
patient-reported
48% (B).
On survey 50% of
those not screened
reported they
received a referral.
Over 40% fail to
show for eye
appointments.
34% 27%
20%
9%
66% 64%
54%
48%
39% 39%
14%
0%
20%
40%
60%
80%
PCC +
NHC
no opt
PCC NHC
no opt
No
Care
Site
NHC +
DC
PCC +
DC
IDP +
PCC +
NHC
DC PCC +
NHC
with
opt
NHC
with
opt
IDP
Percent of Diabetes Patients Screened by Primary
Care Site
No Eye Support Local Eye Support
Abbreviations: PCC = Primary Care Center, NHC = Neighborhood Health Centers, opt = optometrist, DC = Diabetes Center, IDP = Infectious Disease Program
SCREENING RATES VARY BY CARE SITE
Among 19,361
diabetes patients
in 2014, screening
rates varied widely.
Local eye support
nearly tripled DR
screening rates
(48% vs 17%,
p<0.0001).
Values do not sum to total as patients may have visited multiple sites
*Likely includes treatment visits
96%
1% 4% 0%
45%
34%
28%
1%
0%
20%
40%
60%
80%
100%
Ophthalmology
Clinic*
Diabetes Center Neighborhood
Health Centers
Infectious Disease
Center
Screening Location for 5,595 Diabetes Patients in
2014 With and Without Local Eye Support
No Local Eye Support Local Eye Support
LOCAL EYE SUPPORT VITAL
Local eye support
decreased the
number of screens
done in OC (45% vs.
96%, p<0.0001).
MULTIVARIABLE DR SCREENING MODEL
Improving Diabetic Retinopathy Screening is a Complex Challenge
Contact: dziemer@emory.edu
Factor Odds Ratio p-value
Age 1.03/yr <0.0005
Depression 1.3 <0.0005
Neighborhood Center ― optometrist 3.1 <0.0005
Diabetes Center ― retinal camera 6.6 <0.0005
Schizophrenia 0.7 <0.0005
Other mental health diagnoses 0.8 <0.0005
Commercial insurance 0.7 <0.0005
Sex, race, other insurance were NS. Non-Hispanic OR=0.8, continuity 1.1
3. U.S. Department of Health and Human Services, The National Institutes of Health, National Eye Institute. Diabetic Retinopathy. Accessed May 13, 2016. Available from: https://nei.nih.gov/eyedata/diabetic
4. Zhang X, Saaddine JB, Chou C, Cotch MF, Cheng YJ, Geiss LS, Gregg EW, Albright AL, Klein BEK, Klein R. 2010. Prevalence of diabetic retinopathy in the United States, 2005-2008. JAMA; 304(6):649-656.](https://image.slidesharecdn.com/c1a62b8f-0fd1-4306-a9a3-d8f3c7542a62-160727165211/85/ADA-617-P_Improving-DR-screening-is-a-complex-challenge-1-320.jpg)
ADA 617-P_Improving DR screening is a complex challenge
- 1. SEVERE LACK OF CAPACITY Screening all diabetes patients in OC would use 99% of their 19,566 visits. Removing DR screening from OC (1,506 single visit & no major eye diagnosis) will barely accommodate treatment only if the eye disease rate among the unscreened is as low as 5% (688 X 2.1 visits = 1,445). David C. Ziemer1, Pranay K. Neema2, Alex Mojonnier2, Lauren L. Asp1, Jiani Hu1, Luxi Liang1, Raymond L. Smith2, Allison Pall1, Catherine S. Barnes1, Jane M. Caudle1 1Emory University School of Medicine, Division of Endocrinology, Metabolism and Lipids & 2Grady Health System, Clinical Decision Support – Atlanta, Georgia ABSTRACT (617-P) Early treatment of diabetic retinopathy (DR) reduces vision loss. Nationally, DR screening rates are inadequate. Meeting DR screening guidelines is a high priority challenge in a public hospital system where Ophthalmology Clinic (OC) is over-capacity. To plan for better DR screening, we analyzed 2014 administrative data in the Grady Health System where several clinics manage diabetes. DR screening was ruled complete for patients with OC, optometry or retinal photography visits. We surveyed a convenience sample of diabetes patients at a clinic visit. Among 19,361 persons with diabetes overall screening was low (29%) with wide variation by care site (5 - 66%); 5,000 had no diabetes continuity care visit. In a multivariable model, DR screening increased in diabetes & primary care clinics and with age (OR: 2.8, 2.1, & 1.03/yr respectively); it decreased with [NON-]Hispanic ethnicity and mental health diagnoses (OR 0.7 and 0.8) (all p<0.001). Having in-clinic eye screening doubled DR screening (48% vs 22%) and decreased the number of screens done in OC (45% vs 96%). On the patient survey ~70% had eye screens in the prior year – 50% by OC, 25% by Diabetes Clinic and 13% outside the system. Half of the unscreened said a referral was made. No show rate for screening appointments is over 40%. Many screening barriers must be addressed: screening capacity, lack of regular diabetes care, scheduling processes, referral rates and needs of mental health and Hispanic patients. Better processes are needed to recruit patients into and maintain regular diabetes care, and to improve appointment keeping and follow-up for those missing appointments. Moreover, repeated annual screening will require a tracking system. Planned expansion of teleretinal screening and optometry will increase capacity but also awareness of treatment- requiring disease. Thus, planning for downstream treatment capacity is needed. DR screening is a complex problem requiring not just infrastructure investments but also major care management interventions. INTRODUCTION Diabetes is common, costly, and deadly in the US and the Grady Health System (GHS). Diabetic retinopathy (DR) is a major cause of vision loss in the US.1 In 2011, the age-adjusted percentage of adults with diagnosed diabetes reporting visual impairment was 17.6%.2 This is a pressing issue as the number of Americans with DR is expected to double from 7.7 million in 2010 to 15.6 million in 2050.3 This may have a greater impact on non-Hispanic Black and Mexican Americans, who have a higher prevalence of DR than non-Hispanic whites.4 Historically, diabetes has been present in 13% of all GHS patients, 30- 35% of all Primary Care visits, and 24% of GHS charges/cost. METHODS To evaluate DR screening, GHS administrative data on 19,361 diabetes patients in 2014 were analyzed. Patients were eligible based on any one of the following: • At least two outpatient diagnoses on separate days in 2014 • One inpatient diagnosis in 2014 • Diabetes medication in 2012-2015 • At least two A1Cs ≥6.5 in 2014 DR screening was considered complete if an Ophthalmology Clinic (OC), optometry, or retinal photograph visit was attended. In addition, a convenience sample of 80 adult GHS diabetes patients were asked about their DR screening in the prior year. SUMMARY • Most diabetes patients, 71%, do not receive DR screening • Health system screening capacity is the major limitation • In-clinic eye support increases DR screening dramatically • DR screening is decreased by: – Lack of continuity care – Failure to keep appointments – Failure to refer – Mental health diagnoses other than depression CONCLUSIONS Comprehensive DR screening is a complex, multi-aspect challenge. The system needs to dramatically increase: • Capacity for screening and treatment • The number of diabetes patients with continuity care It also needs to develop: • Process for tracking and scheduling those who need DR screening, including those missing appointments • Methods to Improve patient and provider engagement, especially for those with mental health problems • Consideration for alternative screening methods ACKNOWLEDGEMENTS This work was supported by the American Diabetes Association (7-12-HYPO-11). 1.Centers for Disease Control and Prevention (CDC), Division of Diabetes Translation, National Center for Chronic Disease Prevention and Health Promotion. 2015. Why is vision loss a public health problem? Accessed May 25, 2016: http://www.cdc.gov/visionhealth/basic_information/vision_loss.htm 2.CDC, National Center for Health Statistics, Division of Health Interview Statistics, data from the National Health Interview Survey. 2015. Crude and age-adjusted percentage of adults 18 years or older with diagnosed diabetes reporting visual impairment, United States, 1997-2011. Accessed May 13, 2016: www.cdc.gov/diabetes/statistics/visual/fig2.htm All (n=19,361) Screened (n=5,595) Not Screened (n=13,766) p-value Mean Diabetes Visits (SD) 1.6 (2.9) 3.1 (3.8) 1.0 (2.0) <0.0005 Mean Eye Visits (SD) 0.6 (1.5) 2.1 (2.2) 0.0 (0.0) <0.0005 Mean A1C (SD)* 7.80 (2.31) 7.75 (2.08) 7.83 (2.42) 0.043 Age in Years (SD) 56.8 (13.7) 60.4 (12.2) 55.4 (14.0) <0.0005 Female 59.4% 62.7% 58.1% <0.0005 African American/Black 83.2% 85.7% 82.2% <0.0005 Non-Hispanic 91.4% 91.2% 91.5% 0.727 Eye Disease Diagnosis 7.2% 17.2% 3.2% <0.0005 Cataract 1.7% 3.8% 0.8% <0.0005 Glaucoma 0.5% 1.3% 0.2% <0.0005 Macular Edema 2.8% 1.3% 0.2% <0.0005 Proliferative Retinopathy 4.1% 10.2% 1.6% <0.0005 Other Mental Health Diagnosis** 46.4% 44.2% 47.6% <0.0005 *values calculated with total n=14,034 **values calculated with total n=19,265 DEMOGRAPHICS Population was mostly non-Hispanic, African American & female, with mean age 57 years. Many (26% had no diabetes continuity care visit. The unscreened had mean of only 1 clinic visit for diabetes care vs 3.1 for those screened. 48% 26% 7% 6% 13% 0% 20% 40% 60% 80% Ophthalmology Clinic Diabetes Center "Grady" Neighborhood Health Center Outside GHS Screening Location for 54 Diabetes Patients Screened in Past Year B Values do not sum to total as patients may have visited multiple sites *Likely includes treatment visits 64% 21% 19% 0% 0% 20% 40% 60% 80% Ophthalmology Clinic* Diabetes Center Neighborhood Health Centers Infectious Disease Center Screening Location for 5,595 Diabetes Patients Screened in 2014 SCREENING RATES SUB-OPTIMAL In administrative data, only 29% of diabetes patients completed a DR screening in 2014 (A). Patient, in-clinic survey screening rate was higher than administrative data suggested at 68% (B). This implies that those in regular care are more likely to be screened. 33% 68% 0% 20% 40% 60% 80% Not Screened Screened in Past Year Screening Rates for 80 Adult Diabetes Patients in Past Year B 71% 29% 0% 20% 40% 60% 80% Not Screened Eye Screen in 2014 Screening Rates for 19,361 Diabetes Patients in 2014 A A MOST “SCREENING” IN OPHTHALMOLOGY CLINIC In administrative data, OC screened 64% of diabetes patients with DR screening(A). This is similar to the patient-reported 48% (B). On survey 50% of those not screened reported they received a referral. Over 40% fail to show for eye appointments. 34% 27% 20% 9% 66% 64% 54% 48% 39% 39% 14% 0% 20% 40% 60% 80% PCC + NHC no opt PCC NHC no opt No Care Site NHC + DC PCC + DC IDP + PCC + NHC DC PCC + NHC with opt NHC with opt IDP Percent of Diabetes Patients Screened by Primary Care Site No Eye Support Local Eye Support Abbreviations: PCC = Primary Care Center, NHC = Neighborhood Health Centers, opt = optometrist, DC = Diabetes Center, IDP = Infectious Disease Program SCREENING RATES VARY BY CARE SITE Among 19,361 diabetes patients in 2014, screening rates varied widely. Local eye support nearly tripled DR screening rates (48% vs 17%, p<0.0001). Values do not sum to total as patients may have visited multiple sites *Likely includes treatment visits 96% 1% 4% 0% 45% 34% 28% 1% 0% 20% 40% 60% 80% 100% Ophthalmology Clinic* Diabetes Center Neighborhood Health Centers Infectious Disease Center Screening Location for 5,595 Diabetes Patients in 2014 With and Without Local Eye Support No Local Eye Support Local Eye Support LOCAL EYE SUPPORT VITAL Local eye support decreased the number of screens done in OC (45% vs. 96%, p<0.0001). MULTIVARIABLE DR SCREENING MODEL Improving Diabetic Retinopathy Screening is a Complex Challenge Contact: dziemer@emory.edu Factor Odds Ratio p-value Age 1.03/yr <0.0005 Depression 1.3 <0.0005 Neighborhood Center ― optometrist 3.1 <0.0005 Diabetes Center ― retinal camera 6.6 <0.0005 Schizophrenia 0.7 <0.0005 Other mental health diagnoses 0.8 <0.0005 Commercial insurance 0.7 <0.0005 Sex, race, other insurance were NS. Non-Hispanic OR=0.8, continuity 1.1 3. U.S. Department of Health and Human Services, The National Institutes of Health, National Eye Institute. Diabetic Retinopathy. Accessed May 13, 2016. Available from: https://nei.nih.gov/eyedata/diabetic 4. Zhang X, Saaddine JB, Chou C, Cotch MF, Cheng YJ, Geiss LS, Gregg EW, Albright AL, Klein BEK, Klein R. 2010. Prevalence of diabetic retinopathy in the United States, 2005-2008. JAMA; 304(6):649-656.