Effects of COVID-19 pandemic on glaucoma appointment scheduling in a tertiary hospital in London, UK

Abstract

Purpose To investigate the impact of the delay in patient appointments caused by the COVID-19 pandemic and the triage system on the glaucomatous disease of patients in a London tertiary hospital. Methods Observational retrospective study that randomly selected 200 glaucoma patients with more than 3 months of unintended delay for their post-COVID visit and other inclusion and exclusion criteria. Demographic information, clinical data, number of drugs, best-corrected visual acuity (BCVA), intraocular pressure (IOP), visual field (VF) mean deviation (MD), and global peripapillary retinal nerve fibre layer (pRNFL) thickness were obtained from the pre- and post-COVID visit. At the post-COVID visit, the clinical outcomes subjective clinical concern and change of treatment or need for surgery were also annotated. The variables were stratified by glaucoma severity (according to the MD into early, moderate and advanced) and by delay time (more and less than 12 months) and analysed using SPSS. Results We included 121 eyes (from 71 patients). The median patient age was 74 years (interquartile range -IQR- 15), 54% were males and 52% Caucasians. Different glaucoma types and all glaucoma severities were included. When data was stratified for glaucoma severity, at the pre-COVID visit, significant differences in BCVA, CCT and IOP were observed and there were significantly higher values in the early glaucoma group. The median follow-up delay was 11 months (IQR 8), did not differ between the glaucoma severity groups and did not correlate to the glaucoma severity. At the post-COVID visit, significant differences in BCVA, IOP, and Global pRNFL thickness were observed between the glaucoma severity groups, as lower BCVA and higher IOP and pRNFL thickness were observed in the early glaucoma group. At the post-COVID visit there was cause for concern in 40 eyes: 5 were followed more closely, 22 had a change of treatment and 13 were booked for surgery (3 for cataract and 10 for glaucoma surgery). However, the number of eyes with causes for concern were similar between the glaucoma severity groups and there was no correlation between these clinical outcomes and the delay of the post-COVID visit. The number of topical hypotensive medications increased significantly after the post-COVID visit, higher number of medications were observed in the advanced glaucoma group. When differences of IOP, MD and pRNFL thickness between the pre and post-COVID visit, only the MD difference was significantly different between the glaucoma severity groups because it was higher in the severe group. When data was stratified for delay longer or shorter than 12 months, no differences were observed between the groups except at the pre-COVID visit, when the numbers of patients with MD deviation >−6 dB had longer delay time. When differences in IOP, MD and RNFL thickness were calculated, only the pRNFL thickness showed significant differences between the delay groups, because it was higher in the longer delay group. Finally, when paired analysis of the variables at the pre- and post-COVID visits, stratified by glaucoma severity and delay were conducted, although there were no significant differences in IOP in any group, the BCVA decreased significantly in the overall group and in the longer delay groups, the number of hypotensive drugs increased significantly overall and in the moderate and advanced glaucoma, the MD of the VF worsened significantly in the overall group and in the early glaucoma and longer delay groups and the pRNFL thickness decreased significantly in all groups. Conclusions We document that delayed care impacts negatively on the glaucomatous disease of our patients because at the post-COVID visit there were reasons for clinical concern in a third of eyes that resulted in change of treatment or surgery. However, these clinical consequences were not related to IOP, glaucoma severity or delay time and reflect that the triage methods implemented worked adequately. The most sensitive parameter to indicate progression in our sample was the pRNFL thickness.