The effect of COVID-19 on the vitreoretinal service of a tertiary referral centre: Real-world experience from the Royal Hallamshire Hospital

To the Editor,

We read with interest Jasani et al.'s manuscript demonstrating significant reduction in the presentation of rhegmatogenous retinal detachments (RRDs), and increase in the proportion of macula-off RRDs and proliferative vitreoretinopathy cases at three tertiary Centres in North-West of England during the Coronavirus disease 2019 (COVID-19) lockdown when compared to a pre-lockdown cohort. ¹

We conducted a similar, retrospective analysis of all vitreoretinal procedures undertaken at Royal Hallamshire Hospital, Sheffield, between March 21, 2020 (beginning of the lock-down period in the UK) and May 16, 2020, and compared this to a comparative cohort between March 23, 2019 and May 19, 2019.

Royal Hallamshire Hospital provides a tertiary referral service for patients with vitreoretinal conditions and emergencies, including those with complex medical needs in Sheffield and South Yorkshire region, serving a combined population of 2.5 million patients.

The baseline characteristics of our cohorts are presented in Table 1. The overall decrease in vitreoretinal workload compared to the correspondent period in 2019 was of 48.6%, and 34.1% in the presentation of RRDs, similarly to previous reports.^1,2 We demonstrate a reduced number of vitreoretinal lists (−39.4%) and of procedures performed in each list (−15.5%), an increased duration of each theatre session (p = 0.00001), increased number of days between onset of symptoms and presentation to Vitreoretinal Unit (p = 0.025), increased proportion of macula-off RRDs (+20%) and increased use of local anaesthetic during the pandemic (+9.5%) (Figure 1). Notably, the duration of vitreoretinal procedures and the number of days between onset of symptoms to vitreoretinal surgery did not change significantly (p = 0.426 and p = 0.456 respectively), in spite of the increased time spent in wearing personal protective equipment and in patients’ turnover in theatre, and of reduced theatre sessions available (Figure 1). Interestingly, although we recorded a difference in presenting mean visual acuities of patients with macula-OFF RRDs during the 2019 (1.54 ± 1.37 logMar) and 2020 (1.77 ± 1.36 logMar), this was not significant (p = 0.24 – Table 1), contrary to previous reports.²⁻⁴

OPEN IN VIEWER

Figure 1.

Comparison between 2019 and 2020 cohorts.

OPEN IN VIEWER

Table 1.

Cohort characteristics.

2019			2020
Characteristics	Measure	Number (%)	Characteristics	Measure	Number (%)	Percentage of reduction compared to pre-COVID era
Total number of cases		191 (100%)	Total number of cases		98 (100%)	−48.6%
Eye	Right	106 (55%)	Eye	Right	49 (50%)
	Left	85 (45%)		Left	49 (40%)
Priority level	Urgent	110 (58%)	Priority level	Urgent	75 (76.5%)	−31.8%
	Male	72 (65%)		Male	50 (66.6%)
	Female	38 (35%)		Female	25 (33.3%)
	Number of retinal detachments	85 (77%)		Number of retinal detachments	56 (74.7%)	−34.1%
	Macula on	38 (45%)		Macula on	19 (34%)	−50%
	- Preoperative best corrected visual acuity	0.51 ± 1.11 logMar (20/65 Snellen equivalent)		- Preoperative best corrected visual acuity	0.19 ± 0.29 logMar (20/31 Snellen equivalent)	t-test p = 0.132
	Macula off	47 (55%)		Macula off	37 (66%)	−21.3%
	- Preoperative best corrected visual acuity	1.54 ± 1.37 logMar (20/693 Snellen equivalent)		- Preoperative best corrected visual acuity	1.77 ± 1.36 logMar (20/1178 Snellen equivalent)	t-test p = 0.240
	Number of days between symptoms onset and vitreoretinal assessment	0 – 28 (mean 5.6 ± 6.4)		Number of days between symptoms onset and vitreoretinal assessment	0–60 (mean 8.3 ± 10.7)	t-test p = 0.025
	Number of days between symptoms onset and surgery	1–80 (mean 9.6 ± 12.2)		Number of days between symptoms onset and surgery	0–63 (mean 9.8 ± 12.7)	t-test p = 0.456
	General anesthesia	25 (23%)		General anesthesia	12 (16%)
	Local anesthesia	85 (77%)		Local/topical/subconjunctival anesthesia	63 (84%)
	Duration of procedure	10–208 min (mean 51.1 ± 31.1).		Duration of procedure	10 – 172 min (mean 52.17 ± 33.13).	t-test p = 0.426
	Procedures in patients with high priority or with excessive waiting time	81 (42%)		Procedures in patients with high priority or with excessive waiting time	23 (23.5%)	−71.6%
	Male	37		Male	12 (52%)
	Female	44		Female	11 (48%)
	Number of months in waiting list	0.3–10 (mean 2.78 ± 1.98)		Number of months in waiting list	0.03– 11.3 (mean 2.94 ± 2.79)
	General anesthesia	23 (28%)		General anesthesia	3 (13%)
	Local anesthesia	58 (72%)		Local anesthesia	20 (87%)
	Duration of procedure	13–130 min (mean 55.1 ± 26.4)		Duration of procedure	10–167 min (mean 51.0 ± 32.7)
Number of theatre sessions		89	Number of theatre sessions		54	−39.4%
Mean number of cases per session		2.15	Mean number of cases per session		1.81	−15.5%
Duration of procedures		10–208 min (mean 52.88 ± 29.05)	Duration of procedures		10 – 172 min (mean 52.17 ± 33.13)	−1.35% (p = 0.426)
Duration of theatre session		63–299 min (mean 172.34 ± 58.04).	Duration of theatre session		177 – 433 min (mean 273.32 ± 38.87)	+59% (p = 0.00001)

Shown are the cohort characteristics for the COVID-19 cohort and comparative 2019 cohort.

Our results pose important implications also on the vitreoretinal surgical training, having the surgical case volume decreased by almost 49% compared to pre-COVID period and being the overall efficiency of operating lists reduced. This is in line with previous studies which reported the significant effects of COVID-19 on general ophthalmology and subspecialty vitreoretinal training.⁵

We therefore suggest significant changes to the vitreoretinal practice of a busy tertiary referral centre in response to COVID-19 restrictions. Changes in management include reduced surgical case volume and efficiency of operating lists, prioritisation of urgent activities, longer turnover time in between cases, judicious use of general anaesthetic and reduced surgical teaching opportunities.