Cardiff University | Prifysgol Caerdydd ORCA
Online Research @ Cardiff 
WelshClear Cookie - decide language by browser settings

Epidemiology of diabetic retinopathy and an assessment of screening intervals

Thomas, Rebecca Louise 2015. Epidemiology of diabetic retinopathy and an assessment of screening intervals. PhD Thesis, Cardiff University.
Item availability restricted.

PDF - Accepted Post-Print Version
Download (6MB) | Preview
[img] PDF - Supplemental Material
Restricted to Repository staff only

Download (388kB)


This thesis was an observational study to retrospectively examine the prevalence and incidence of DR and its associated risk factors from two screening services (Wales and Johannesburg, South Africa) in order to determine safe screening intervals in persons without evidence of DR at initial screening based on digital photography. Between 2005 and 2009 a total of 135,152 persons with diabetes over 12 years of age in Wales were screened. However, a total of 43,759 persons were excluded from analysis as they did not have their type of diabetes recorded (29,807) or where it was recorded it was outside of the pre-specified age at diagnosis of diabetes range of ≥30 years for type 2 DM and <30 years for type 1 DM (13,952). In the Centre for Diabetes and Endocrinology, Johannesburg, South Africa a smaller population of 5,565 were screened between 2001-2010. A total of 50 persons were excluded from this analysis as they had a type of diabetes recorded other than type 1 DM or type 2 DM. Therefore, data from 91,393 (86,390 T2DM, 5,003 T1DM) persons from the Wales screening service and 5,515 (3,978 T2DM, 1,537 T1DM from South Africa, were analysed. In Wales, the prevalence of any DR was 31.0%, background DR (BDR) 26.6% and referable DR (RDR) 4.4% in T2DM at baseline. The prevalence was higher in T1DM at 56.2%, 39.8% and 16.4% respectively. Increased duration of diabetes was independently associated with increased prevalence and incidence of any DR, BDR and RDR for T2DM and T1DM as well as treatment modality in T2DM. The four year cumulative incidence of RDR was 1.6% for T2DM and 5.6% for T1DM. At the Centre for Diabetes and Endocrinology Johannesburg the prevalence of any DR was 21.6%, BDR 14.8% and RDR 6.7% in T2DM at first screening and higher at 36.9%, 27.2% and 9.7% respectively for T1DM. Glycaemic control (HbA1c) and duration of diabetes were significantly associated with the prevalence and incidence of any DR, BDR and RDR in both T2DM and T1DM. Ethnicity and hypertension were also risk factors. The seven year cumulative incidence of RDR was 4.7% for T2DM and 5.0% for T1DM. Risk factor analysis indicated that the screening intervals could be extended beyond annual based on type of diabetes, duration of diabetes, HbA1c, ethnicity, hypertension and treatment modality in T2DM. This analysis adds to an increasing evidence base for considering extending the screening intervals beyond annually for those at low risk with no DR at initial screening. Limitations of this study included the need to exclude a high number of persons in order to ensure the quality of the data; the numbers lost to follow up (meaning all results from 3 years should be interpreted with caution) and the fact that the visual impairment and blindness within the programmes could not necessarily all be attributed to DR as recordings of other lesions were not included in the datasets.

Item Type: Thesis (PhD)
Status: Unpublished
Schools: Medicine
Subjects: R Medicine > R Medicine (General)
Funders: RCUK
Date of First Compliant Deposit: 30 March 2016
Last Modified: 28 Jun 2019 03:29

Actions (repository staff only)

Edit Item Edit Item


Downloads per month over past year

View more statistics