About 382 million people suffer from diabetes. Over the last decades, it has become one of the most common diseases not only in the western world [
Therefore, regular screening intervals for diabetic complications are essential. Screening on a regular, usually yearly basis for diabetic neuro- and nephropathy is mostly performed by the general practitioner or diabetologist [
The International Diabetes Federation (IDF) guidelines recommend annual screenings for diabetic retinopathy (DR) and screening within 3 to 6 months in case of a deterioration of the eye fundus since the last examination [
In this study, we analyzed whether screening for DR in a specialized diabetes center might improve early diagnosis of DR. Therefore, a nonmydriatic fundus camera was implemented in the diabetes outpatient clinic at the University Hospital of Heidelberg. Nonmydriatic stereoscopic retinal imaging has been demonstrated to be a reliable, cost-effective, highly sensitive, and specific method for DR [
A total of 502 patients with type 1 or type 2 diabetes were included in this study over a time period of six months. Patients were recruited from the outpatient clinic of the Department of Endocrinology at the University Hospital of Heidelberg, Germany. Eligibility criteria included patients with established diabetes mellitus type 1 or type 2, age of 18 years or older, and the ability to cooperate. Monocular vision was an exclusion criterion. The following data were collected from each patient: age, duration of diabetes, type of treatment, associated systemic risk factors, and history of eye treatment or examinations in the past. Data of all screened patients were documented in an ad hoc generated database. This database was built up of utilizable evaluation scores, and preexisting medical reports from internal and external sources were used as source data. The study protocol was approved by the ethics committee of the University of Heidelberg. Informed consent was given by all study participants.
All 502 patients underwent single-field 45-degree (in smaller pupil diameter mode: 37°) retinal photography with a nonmydriatic auto fundus camera (Nidek/Oculus AFC-230/210, NIDEK Co., Ltd., Japan) linked to a high resolution digital SLR (single lens reflex) camera (21.8 megapixel full frame sensor, Canon EOS 5D Mark II, Canon Deutschland GmbH, Krefeld, Germany). The images were captured with central focus on the macula including the optic disk. They were stored and sorted through an incorporated data filing system, called NAVIS-Lite. Photography was performed without pupillary dilation. To allow for stable quality, all images were taken by the same trained technician throughout the study. Before commencing the study, the medical technician was instructed in using the camera and interpreting the retinal photographs, until he felt safe on carrying out the examination. Studies verify that this way of screening correlates with a high level of accuracy [
Diabetic retinopathy was defined according to the International Clinical Diabetic Retinopathy Disease Severity Scale [
Four different pathologies which can be detected in nonmydriatic funduscopy. (a) Mild nonproliferative retinopathy with hypertensive fundus. (b) Moderate nonproliferative retinopathy. (c) Proliferative retinopathy after laser photocoagulation. (d) Hypertensive retinopathy grade I.
All patients were screened at the department within the last 12 months for diabetic neuro- or nephropathy. The latest results were taken for the evaluation of this study. Diabetic nephropathy was defined as microalbuminuria of more than 20 mg/L in two of three samples of morning urine obtained within twelve consecutive months [
Absence or presence of neuropathy was assessed using neuropathy symptom score (NSS) and neuropathy disability score (NDS) [
Statistical analysis was performed using specific, predefined parameters. We used descriptive analyses for characterizing the study population. Comparisons of categorical baseline characteristics between the patients with DR and without DR were conducted by chi-square test analysis. Continuous baseline variables were compared using
Logistic regression models were used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for mean last eye screening within 12 months or longer than one year.
The normally distributed data for descriptive analysis was declared as mean ± standard deviation, unless stated otherwise. The percentage values have been rounded off to whole numbers. Since microvascular complications differ with respect to prevalence and incidence rates, the ratio of new to preexisting complications was built. It was assumed that the rates for DN and DP represent more or less the rate of natural progress, since both complications are screened on a regular basis in the study center.
Statistical significance level was considered at a two-side probability level of 0.05 or less. Statistical analyses were performed using Excel 2003 and SPSS (PASW Statistics 18, IBM Deutschland GmbH, Ehningen, Germany).
The study was conducted as a cross-sectional, nonrandomized, noncontrolled, prospective study. Patients were recruited over a time period of six months.
Overall, 502 nonrelated Caucasian patients with type 1 (
Baseline parameters of patients with type 1 and type 2 diabetes.
Characteristics | DM type 1 |
DM type 2 |
---|---|---|
Diabetes duration, y | 28 ± 15 | 14 ± 10 |
|
12 (10.7%) | 166 (42.6%) |
11–19 years, |
24 (21.4%) | 133 (34.1%) |
|
76 (67.9%) | 91 (23.3%) |
Gender, m/w (% of total) | 56/56 (50.0%/50.0%) | 196/194 (50.3%/49.7%) |
Age, mean (SD), y | 52 ± 17 | 65 ± 12 |
Age, |
||
|
74 (66.1%) | 129 (33.1%) |
61–64 years | 5 (4.5%) | 60 (15.4%) |
65–69 years | 11 (9.8%) | 46 (11.8%) |
70–74 years | 14 (12.5%) | 58 (14.9%) |
75–79 years | 5 (4.5%) | 56 (14.4%) |
|
3 (2.7%) | 41 (10.5%) |
BMI, kg/m² | 25.7 ± 3.9 | 33.0 ± 8.2 |
Hypertension, |
59 (52.7%) | 347 (89.0%) |
Systolic blood pressure, mmHg | 135 ± 15 | 138 ± 19 |
Diastolic blood pressure, mmHg | 81 ± 10 | 81 ± 12 |
Nicotine use, |
8 (7.1%) | 41 (10.5%) |
History of cerebrovascular disease, |
4 (3.6%) | 43 (11.0%) |
History of cardiovascular disease, |
11 (9.8%) | 114 (29.2%) |
HbA1c (%) | 7.4 ± 1.1 | 7.3 ± 1.5 |
Triglyceride level (mg/dL) | 94 ± 55 | 213 ± 217 |
Total cholesterol level (mg/dL) | 180 ± 35 | 181 ± 40 |
LDL cholesterol level (mg/dL) | 91 ± 26 | 94 ± 32 |
HDL cholesterol level (mg/dL) | 70 ± 22 | 48 ± 14 |
Antidiabetic treatment, |
||
Lifestyle | 0 (0%) | 34 (8.7%) |
Oral agents | 0 (0%) | 159 (40.8%) |
Insulin treatment | 112 (100%) | 197 (50.5%) |
When all patients with diabetes were analyzed, prevalence of DR was 25.0% (
The incidence rate for both types of diabetes was 6.4% (
Of the 26 patients with type 2 diabetes and new DR, 96.1% (
Additionally, signs for associated hypertensive retinopathy were found in 43.4% (
Next, patients with newly detected DR were characterized: Patients with type 1 diabetes and new onset DR had significantly higher systolic and diastolic blood pressure, positive history of cerebrovascular disease, and significantly higher total and LDL cholesterol levels and were more likely to have microalbuminuria (Table
Baseline parameters of prevalence and severity of retinopathy in persons with type 1 diabetes (
Characteristics | DM 1 |
DR preexisting |
DR new |
|
DR all |
DR absent |
|
|
---|---|---|---|---|---|---|---|---|
Gender, |
||||||||
Male | 56 (50.0%) | 20 (58.8%) | 4 (66.7%) |
|
24 (60.0%) | 32 (44.4%) |
|
|
Female | 56 (50.0%) | 14 (41.2%) | 2 (33.3%) | 16 (40.0%) | 40 (55.6%) | |||
Age, y | 52 ± 17 | 64 ± 11 | 51 ± 11 |
|
62 ± 12 | 47 ± 17 |
|
|
Age, |
||||||||
|
74 (66.1%) | 13 (38.2%) | 5 (83.3%) |
|
18 (45.0%) | 56 (77.8%) |
|
|
|
3 (2.7%) | 3 (8.8%) | 0 |
|
3 (7.5%) | 0 |
|
1.00 |
Diabetes duration, y | 28 ± 15 | 40 ± 12 | 29 ± 8 |
|
38 ± 12 | 23 ± 13 |
|
|
|
12 (10.7%) | 0 | 0 | 1.00 | 0 | 12 (16.7%) |
|
|
11–19 years, |
24 (21.4%) | 2 (5.9%) | 1 (16.7%) |
|
3 (7.5%) | 21 (29.2%) |
|
|
|
76 (67.9%) | 32 (94.1%) | 5 (83.3%) |
|
37 (92.5%) | 39 (54.2%) |
|
|
Median time to last eye screening (months) | 6.5 ± 7.6 | 4.2 ± 4.0 | 7.0 ± 3.1 | 0.12 |
4.7 ± 4.0 | 7.6 ± 8.9 |
|
|
Hypertension, |
59 (52.7%) | 30 (88.2%) | 4 (66.7%) |
|
34 (85.0%) | 25 (34.7%) |
|
|
Systolic blood pressure, mmHg | 135 ± 15 | 137 ± 16 | 146 ± 13 |
|
139 ± 16 | 133 ± 14 |
|
|
Diastolic blood pressure, mmHg | 81 ± 10 | 78 ± 10 | 91 ± 8 |
|
80 ± 11 | 81 ± 10 |
|
|
Statin therapy | 51 (45.5%) | 25 (73.5%) | 2 (33.3%) |
|
27 (67.5%) | 24 (33.3%) |
|
|
Biochemical characteristics | ||||||||
Total cholesterol level (mg/dL) | 180 ± 35 | 169 ± 33 | 212 ± 42 |
|
175 ± 37 | 182 ± 33 |
|
|
HbA1c (%) | 7.4 ± 1.1 | 7.1 ± 1.2 | 8.0 ± 1.2 |
|
7.2 ± 1.2 | 7.4 ± 1.1 |
|
|
DR = diabetic retinopathy, HDL = high density lipoprotein, and LDL = low density lipoprotein.
Baseline parameters of prevalence and severity of retinopathy in persons with type 2 diabetes (
Characteristics | DM 2 |
DR preexisting |
DR new |
|
DR all |
DR absent |
|
|
---|---|---|---|---|---|---|---|---|
Gender, |
||||||||
Male | 196 (50.3%) | 29 (48.3%) | 14 (53.8%) |
|
43 (50%) | 153 (50.3%) |
|
|
Female | 194 (49.7%) | 31 (51.7%) | 12 (46.2%) | 43 (50%) | 151 (49.7%) | |||
Age, y | 65 ± 12 | 71 ± 9 | 66 ± 11 |
|
69 ± 9 | 64 ± 12 |
|
|
Age, |
||||||||
|
129 (33.1%) | 9 (15.0%) | 10 (38.5%) |
|
19 (12.8%) | 110 (74.3%) |
|
|
|
41 (10.5%) | 11 (18.3%) | 3 (11.5%) |
|
14 (25.5%) | 27 (49.1%) |
|
|
Diabetes duration, y | 14 ± 10 | 24 ± 10 | 18 ± 11 |
|
22 ± 10 | 11 ± 8 |
|
|
|
166 (42.6%) | 4 (6.7%) | 5 (19.2%) |
|
9 (10.5%) | 157 (51.6%) |
|
|
11–19 years, |
133 (34.1%) | 16 (26.7%) | 13 (50.0%) |
|
29 (33.7%) | 104 (34.2%) |
|
|
|
91 (23.3%) | 40 (66.7%) | 8 (30.8%) |
|
48 (55.8%) | 43 (14.1%) |
|
|
Median time to last eye screening (months) | 9 ± 13 | 5 ± 12 | 11 ± 11 |
|
7 ± 12 | 10 ± 13 |
|
|
Hypertension, |
347 (89.0%) | 59 (98.3%) | 24 (92.3%) |
|
83 (96.5%) | 264 (86.8%) |
|
|
Antidiabetic treatment, |
||||||||
Lifestyle | 34 (8.7%) | 1 (1.7%) | 2 (7.7%) |
|
3 (3.5%) | 31 (10.2%) |
|
|
Oral agents | 261 (66.9%) | 28 (46.7%) | 19 (73.1%) |
|
47 (54.7%) | 214 (70.4%) |
|
|
Insulin treatment | 197 (50.5%) | 52 (86.7%) | 19 (73.1%) |
|
71 (82.6%) | 126 (41.4%) |
|
|
Statin therapy | 237 (60.8%) | 41 (68.3%) | 19 (73.1%) |
|
60 (69.8%) | 177 (58.2%) |
|
0.14 |
HbA1c (%) | 7.3 ± 1.5 | 7.8 ± 1.5 | 8.4 ± 2.0 |
|
8.0 ± 1.7 | 7.1 ± 1.5 |
|
|
DR = diabetic retinopathy.
HDL = high density lipoprotein, LDL = low density lipoprotein.
Prevalence of nephropathy was 32.5% (
In total, 17.8% (
The prevalence of any DP in both types of diabetes combined was 63.5% with 37.5% (
As described above, 32 DRs were newly detected by using a nonmydriatic fundus camera with onsite screening. Since prevalence of all three microvascular complications is different due to underlying pathogenesis, progression rates were calculated (Table
Progression rate of DR, DN, and DP for all patients with diabetes was calculated as ratio of new detected complications to preexistent complications.
New detected ( |
Preexistent ( |
Progression rate (%) | |
---|---|---|---|
Diabetic retinopathy | 32 | 94 | 34.0 |
Diabetic neuropathy | 53 | 266 | 19.9 |
Diabetic nephropathy | 29 | 134 | 21.6 |
Finally factors contributing to the diagnosis of new DR were analyzed. Since the number of patients with type 1 diabetes was rather small and characteristics for patients with type 1 and type 2 diabetes were different, only patients with type 2 diabetes were included for following analysis. Interestingly, younger age was found to be a risk factor for nonattendance of regular retinopathy screening intervals (
Odds ratio (95% CI) of failure to attend eye screening within one year.
Characteristics | All ( |
|
Type 2 ( |
|
---|---|---|---|---|
Gender | ||||
Male | 0.91 (0.54–1.52) | 0.72 | 0.935 (0.54–1.62) | 0.81 |
Female | ||||
Age | ||||
|
2.00 (1.19–3.36) |
|
2.33 (1.33–4.07) |
|
61–64 | 1.57 (0.79–3.13) | 0.19 | 1.47 (0.73–3.00) | 0.28 |
65–69 | 0.46 (0.16–1.31) | 0.14 | 0.49 (0.17–1.42) | 0.18 |
70–74 | 0.08 (0.01–0.57) |
|
0.08 (0.01–0.60) | 0.002 |
75–79 | 1.14 (0.54–2.44) | 0.73 | 1.06 (0.49–2.30) | 0.88 |
|
0.63 (0.22–1.81) | 0.39 | 0.57 (0.19–1.65) | 0.29 |
Diabetes duration | ||||
|
4.30 (2.50–7.40) |
|
3.54 (1.97–6.37) |
|
11–19 years | 0.65 (0.36–1.19) | 0.16 | 0.48 (0.25–0.93) | 0.027 |
|
0.20 (0.09–0.45) |
|
0.32 (0.13–0.77) |
|
Diabetic retinopathy | 0.42 (0.20–0.88) |
|
0.50 (0.23–1.09) | 0.08 |
Preexistent | 0.25 (0.08–0.83) |
|
0.25 (0.08–0.83) |
|
New diagnosed | 1.22 (0.45–3.28) | 0.70 | 1.34 (0.48–3.70) | 0.75 |
The results of the study show that onsite screening for DR with a nonmydriatic, digital fundus camera can contribute to early diagnosis of diabetic retinopathy in a diabetes outpatient clinic.
In 32 of the analyzed patients, diabetic changes of the retina were newly described, which represents a proportion of 25% of all retinopathies detected in this study. One part of those can be seen as natural progress since the last screening visit. Nevertheless, the other proportion represents so far missed complications, which have been detected due to the changed screening process with onsite fundus screening. In order to separate these two proportions, progression rates for DR, DN, and DP were calculated. Since DN and DP were screened on a regular basis at the study center, one can assume that all newly detected DNs and DPs represent the natural progress of the disease. Both progression rates were close to 20%. Therefore, also a natural progression rate of 20% for DR was assumed. Since 94 DRs were preexistent before the study one would expect 19 newly detected DRs due to the natural progress. Since 32 DRs were described which were unknown before, at least 13 are likely to be detected due to the new onsite retinopathy screening procedure. This number is rather underestimated due to the fact that underlying progression rate derives from DN and DP which have been shown to progress faster than DR does [
The prevalence of DR presented here is in line with current literature [
One might argue that for diabetic retinopathy screening an ophthalmologist is essential and digital fundus imaging alone might not be sufficient. However, recent studies comparing nonophthalmologists with ophthalmologists for the diagnosis of DR have shown similar accuracy in the detection of changes of the fundus even with paramedical staff performing the screening procedure [
When analysing factors contributing to the detection of new DR, younger patients’ age and shorter diabetes duration were associated with nonattendance of regular screenings and with higher incidence of DR. This is in line with previously reported data showing lower screening attendance rates in younger patients with diabetes [
However, there are also some limitations to this study. First, since only outpatients were included at a university hospital clinic, participants might not reflect the general diabetes population, which might differ in age and diabetes duration. Therefore, the rate of missed DR might differ, making generalization of the results difficult.
Additionally the study was performed in an area with a higher density of ophthalmologists than the mean of the country [
A further issue is high rate of arterial hypertension in the analysed population. The hypertension rates were about 90% of the total patients. As hypertension is a major systemic risk factor for diabetic retinopathy a certain influence on our results cannot be fully excluded and the results might not be so prominent in populations with a lower rate of hypertension. However, rates for arterial hypertension found in our study are comparable to those found in other studies from different regions of the world with hypertension rates up to 95% in patients with diabetes [
Furthermore, as data on screening intervals was collected via patient’s self-reports and medical data records, a bias of social desirability is possible.
Little is known concerning the effect of onsite screening with a nonmydriatic fundus camera in specialized diabetes outpatient clinics on early diagnosis of DR. The data presented imply that on the one hand this might result in an earlier diagnosis of complications. On the other hand one might speculate that it vice versa might also affect treatment quality and therapy goals, which have been shown to lower the progression of disease [
In conclusion, the results of this study suggest that onsite screening for diabetic retinopathy with a nonmydriatic digital fundus camera in a diabetes outpatient clinic detects missed diabetic retinopathies in a higher degree than by progression of the disease alone. Due to the epidemic burden of diabetes early identification of patients at risk might help to save time and resources and channel patients with a strong need for specialized eye care especially in younger patients who might profit from time saving diabetes care.
Diabetic nephropathy
Diabetic polyneuropathy
Diabetic retinopathy.
All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008 (5).
Informed consent was obtained from all patients for being included in the study.
The authors declare that they have no conflict of interests.
Gottfried Rudofsky and Hannes Kalscheuer reviewed/edited the paper and researched data. Daniel Lehnhoff, Roman Klein, and Jan B. Groener researched data. Peter P. Nawroth contributed to the discussion. Florian Schuett contributed to image evaluation and to discussion. Pia Roser wrote the paper and researched data.
Further, Gottfried Rudofsky is supported by a grant of the EASD.