Point of Care Retinopathy Screening Study (PARROTS)
Loading...
Date
2020-06-05
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Kamuzu University of Health Sciences
Abstract
Type of study
A pilot of point of care cross-sectional (POC) screening for diabetic retinopathy (DR)
delivered by ophthalmic clinical officers (OCOs) and optometrists will be undertaken at
Queen Elizabeth Central Hospital and Chiradzulu District Hospital. This study will investigate
both the accuracy of the screening test and acceptability of the system of POC screening to
patients and health care workers.
The Problem
Malawi has an adult population prevalence of diabetes of approximately 5% and a high
prevalence of sight threatening diabetic retinopathy (STDR) in persons attending diabetes
clinics. Visual loss from diabetes is preventable and treatable. Early detection is key but
European style screening programmes would not be cost-effective or sustainable in Malawi.
Low-cost fundus cameras used together with novel automated grading software could
provide the means to provide effective point-of-care screening by health workers. Recent
work in 2018/2019 has been undertaken to assess the quality of retinal photographs
produced by 2 portable fundus cameras and to investigate the accuracy of an automated
grading algorithm used together with a hand-held fundus camera for detection of DR.
Automated grading addresses the problems of recruiting, training and cost for human
grading of retinal photographs. A POC result overcomes the problem of communicating
results at a distance. The planned next stage of this project is described below.
Objectives
Broad Objectives
1. To determine whether new and low cost portable screening cameras alongside a
grading algorithm for diabetic retinopathy will benefit patients with diabetes
Specific objectives
1. To investigate the accuracy of an automated grading algorithm used together with a
low-cost fundus camera for detection DR which imminently threatens vision in 2
settings: a Central Hospital and a District Hospital.
2. To investigate the acceptability to patients of a system of POC DR screening using an
automated grading algorithm used together with a low-cost fundus camera.
3. To investigate the acceptability to health care workers of a system of POC DR
screening using an automated grading algorithm used together with a low-cost
fundus camera
This is a prospective pilot study of a screening intervention based at 2 sites. People living
with diabetes (PLWD) attending a diabetes clinic at each site (n=246) will be recruited. DR
screening with POC automated grading will be delivered by OCOs and optometrists. Digital
fundus photography of 2 standard fields (disc-centred and macula-centred) will be
performed through dilated pupils using 2 hand-held fundus cameras (Volk Pictor Plus and
Volk InView).
Photographs will be analysed using Eye Star automated grading software. Screen positive
cases (DR which imminently threatens vision: LDES level 50+ and exudates within 500um of
the central fovea) will be referred to a tertiary referral centre: Lions Sight First Eye Unit at
Queen Elizabeth Central Hospital, Blantyre for assessment and, if necessary, treatment. All
screen positive subjects will be examined by an ophthalmologist (PK) to determine the
positive predictive value. A proportion of screen negative subjects will be examined by an
ophthalmologist to determine the negative predictive value.
The acceptability of the intervention will be assessed by qualitative methods. All patients
and health care workers will be invited to complete a questionnaire. A sub-group of patients
(n=28) and health care workers (n=10) will be interviewed. A separate sub-group of patients
will be invited to participate in focus groups (n=20). The impact of distance, gender, and risk
perception on acceptability of the POC DR screening to patients will be assessed. The study
will be undertaken in a sequential manner: phase 1 at Queen Elizabeth Central Hospital;
phase 2 at Chiradzulu District Hospital. After phase 1 of the study, a planned interim analysis
will be undertaken, and any necessary changes made to the screening model. Expected results
The results are expected to demonstrate the effectiveness of screening for sight threatening
diabetic retinopathy using handheld cameras in a LMIC hospital setting. Results will also
establish acceptability for patients and health care staff. Data from this pilot study will
inform future large studies of this intervention. If the expected results are seen,
implementing diabetic retinopathy screening at a district level would be sought. Approval
would be sought for deployment of the fundus cameras and grading system at a national
level, from the appropriate governing body.
Dissemination of results
The findings of the study will be disseminated through presentation at national and
international meetings and through peer reviewed publication. The Committee of Medical
Research and Ethics Committee (COMREC) will be notified of the findings in the first instance.
Findings from the study will inform a phased introduction of diabetic retinopathy screening
throughout Malawi.