City Research Online

Using large-scale visual field data to gain insights into management of patients with glaucoma

Kelly, S. (2019). Using large-scale visual field data to gain insights into management of patients with glaucoma. (Unpublished Doctoral thesis, City, University of London)


Electronic medical records (EMRs) are increasingly being used for patient management in eye clinics in the UK. Glaucoma clinics in particular have an increasing capacity to store and link patient data from a variety of sources, including perimetric visual field (VF) records, intraocular pressure records, medication and surgical records. The studies detailed in this thesis aim to show, for the first time, the versatility of these data through a range of proof-of concept examples. The studies were designed to show the different levels of scope that these data can be used for, from an individual patient level (patient performance), to a clinic level (auditing clinic performance) to disease level (retrospective cohort studies). The studies took advantage of two data extractions from EMRs from regionally different National Health Service (NHS) Hospital Trust glaucoma clinics in England. The first (2012) and second (2015) extraction yielded records from 88954 and 73994 people respectively. Central to these studies was the use of more than 1.2 million individual VF examinations. The first study tested the hypothesis that perimetric fatigue effects are greater in the eye examined second. The results showed that the second eye tested (usually the left eye) showed slightly reduced sensitivity (0.13 dB, p < 0.001) in mean deviation (MD) compared to the first eye. There was also a small but significant relative increase (3%) in the longitudinal variability of the second eye. This study provides evidence for statistically significant fatigue effects on the second eye tested in routine clinics. Although the average effect was small, there was a large variation among patients and as such, to keep any order effect consistent, eye testing should be carried out in the same order for each visit. The second study took routinely collected clinic data and audited aspects of service delivery of glaucoma. Metrics predicting loss of sight years and reliability of examinations varied between the centres (p < 0.001) indicating that some clinics may be performing better than others. This study illustrates the feasibility of assessing aspects of health service delivery in glaucoma clinics through analysis of VF databases and the proposed metrics could be useful for blindness prevention from glaucoma in secondary care centres. In the third study, the real-world conversion rates from ocular hypertension (OHT) to primary open angle glaucoma (POAG) were estimated. Within the study population (n = 3163), 17.5% (95% CI: 15.4% - 19.6%) were predicted to convert within 5 years. The use of intraocular pressure (IOP) lowering treatment was associated with a lower change of conversion with a Hazard Ratio of 0.45 (95% CI: 0.35 – 0.57, p < 0.001). Previously, it had only been in the context of clinical trials that OHT conversion rates and efficacy of IOP-lowering treatment have been reported, however this study shows the feasibility of using EMR data as an alternative. In the fourth study, the outcomes of patients with a combination of uveitis and glaucoma were compared with that of POAG only patients. The uveitis and glaucoma eyes were almost twice more likely to have rapidly declining VF function than the POAG only eyes (age-adjusted risk ratio of 1.9 [95% CI: 1.8 – 2.0]). It can be difficult to study the outcomes of a combination of diseases such as this due to the low incidence rate. However, when patient visits are recorded in EMRs in such large numbers it becomes feasible to carry out these retrospective cohort studies. In the fifth study, both previous datasets were assessed to investigate how socioeconomic factors affect likelihood of glaucoma detection and progression. People in the most deprived regions presented to the clinic with worse VFs than those in the more affluent regions. Once patients were diagnosed and treated under the management of the hospital eye services these dipartites dissipated. For example there was no evidence of differences in the rates of progression (worsening) of VF based on socioeconomic factors. In conclusion, the results from these studies show the breadth of research that is possible if EMR data from glaucoma clinics were to be fully utilised. Insights into individual patient care, the performance of clinics, and the prognosis of diseases could all be gained through the analysis of routinely collected data in eye clinics as well as an insight into the public health aspect of glaucoma

Publication Type: Thesis (Doctoral)
Subjects: R Medicine > RE Ophthalmology
Departments: Doctoral Theses
School of Health & Psychological Sciences > School of Health & Psychological Sciences Doctoral Theses
School of Health & Psychological Sciences > Optometry & Visual Sciences
[thumbnail of Stephen_Kelly_Thesis_2021.pdf]
Text - Accepted Version
Download (2MB) | Preview


Add to AnyAdd to TwitterAdd to FacebookAdd to LinkedinAdd to PinterestAdd to Email


Downloads per month over past year

View more statistics

Actions (login required)

Admin Login Admin Login