Title: Diabetic retinopathy in a population based and a hospital based sample in a rural district Narail in Bangladesh 


Short title: Diabetic retinopathy in a rural district Narail in Bangladesh

	
Fakir M Amirul ISLAM1,2*, Christine Critchley1, Arzan Hosen2, Lima Asma2, Mohammad Wahab2,3, Robert P. Finger4
1.	Department of Statistics, Data Science and Epidemiology; Faculty of Health, Arts and Design; Swinburne University of Technology, Hawthorn VIC 3122, Australia
2.	Organisation for Rural Community Development (ORCD), Dariapur, Narail, Bangladesh 
3.	Narail Diabetes Hospital, Machinda, Narail 7500, Bangladesh 
4.	Department of Ophthalmology, University of Bonn, Bonn, Germany 



Abstract
Background: Bangladesh is a low-resource country with an increasing prevalence of type 2 diabetes mellitus (DM). Diabetic retinopathy (DR) is one of the most common complications of DM and data on DR in Bangladesh is very limited. To compare the age and sex adjusted prevalence of diabetic retinopathy (DR) in a population-based sample and a hospital based sample in rural Bangladesh.
Methods: Cluster random sampling was used to identify eligible participants.  A total of 173 participants (105 from Abdul Karim diabetes hospital and 68 from the Bangladesh population-based diabetes and eye study (BPDES)) aged 40 years or older with type 2 diabetes mellitus (DM) were recruited.  It was hypothesised that the prevalence and severity of DR would be higher in a tertiary care hospital compared to those in the general population.  Two-field retinal photographs were taken using a 45º non-mydriatic digital fundus camera. Retinal photographs were graded using two certified graders and adjudicated by a retinal specialist.    
 Results:  The mean age (range) of hospital-based participants was 51 (40-78) years and population-based participants was 52 (41-72) years.  The prevalence of any DR in the hospital based study was 19.4% including 11.2% mild to moderate non-proliferative DR (NPDR), 4.6% severe NPDR and 3.6% proliferative or vision threatening DR (VTDR) compared to 20% in the population-based study, including 12% mild to moderate NPDR, 6.7% severe NPDR and 1.3% VTDR. There was no significant difference in the prevalence of any DR by age group or sex between the two groups. There was no difference in prevalence of diabetic macular oedema (DME) between the two groups either (6.6% in the hospital based study and 6.8% in the BPDES). Prevalence of DME was significantly higher in males compared to females (11.6% vs. 3.9%, p=0.04) in the hospital-based study.  
Conclusions: The prevalence of DR at population level is as high as in a hospital based sample in this working age rural sample. Screening for early detection to implement appropriate intervention is needed in the community for working aged adults to save sight from avoidable vision loss due to DM in rural Bangladesh.