Barnaby Carr
Introduction
Roughly one in twenty-five emergency department attendances involves an ocular problem, and most of these patients are first assessed not by an ophthalmologist but by resident clinicians covering the acute take (1). Resident doctors are therefore well placed to detect serious eye disease, not because of specialist training but because of where they sit in the system. The argument developed here is that the limiting factor at that point of entry is rarely diagnostic knowledge. It is timescale, and whether the clinician who sees the patient first can identify the correct escalation route. The resident doctor does not need to be an ophthalmologist. They need to know which clock the patient is on, and to refer before it runs out.
Three presentations illustrate the range of those clocks. Transient monocular visual loss and central retinal artery occlusion run to minutes and hours, and the main risk is misdirection to an ophthalmology clinic when the presentation belongs on a stroke pathway. Giant cell arteritis (GCA) runs to hours and days, and is unusual in that the first clinician can start high-dose corticosteroids without waiting for specialist review. Malignant hypertension allows the longest interval but still requires same-day action, and here the failure is more often that the fundus was never examined at all (2). In none of these cases is the diagnosis obscure. The delay happens after recognition, or in place of it.
The Problem
No single specialty is solely responsible for any of these patients. Amaurosis fugax lies between emergency medicine and stroke services, GCA crosses emergency medicine, rheumatology and ophthalmology, and malignant hypertension sits between emergency and general medicine. Harm accumulates at these boundaries, where no individual clinician is accountable for the delay. Across 4,135 ophthalmology negligence claims opened between 2006/07 and 2024/25, the two most common causes were failure or delay in treatment and failure to arrange follow-up (3). Neither is a purely diagnostic failure. Both are, at least in part, failures of communication, escalation or pathway design. GCA litigation shows the same pattern, with visual loss in 91% of claims and 86% of those associated with delayed or failed diagnosis (4).
It is tempting to attribute this to inadequate ophthalmology training, and some evidence supports that. Undergraduate placements in the UK average 7.6 days, and some schools have no compulsory placement at all (5). There is no mandatory ophthalmology rotation within foundation programmes, and confidence in examining a dilated eye reportedly falls from 88% in fourth year to 43% by final year (6) These findings should nevertheless be interpreted cautiously, since a 2025 systematic review noted that much of this evidence is of low quality and relies on self-reported measures that assess confidence rather than demonstrated competence or patient outcomes (7)
The more consistent finding is not absent knowledge but absent opportunity to apply it, and absent knowledge of where the patient goes next. An audit of emergency management of suspected GCA identified poor awareness of local escalation pathways rather than failure to recognise the disease (8), and a retrospective audit found a mean 35-day delay between symptom onset and starting steroids, with up to one in five patients losing vision (9).
Capacity compounds this. In March 2025, more than 586,000 patients were waiting for an ophthalmology appointment, only 24% of eye units reported sufficient consultants, and around a quarter of consultants planned to leave within five years (3). Proposed remedies such as additional consultants and digital triage are reasonable, but they are downstream interventions that depend on a workforce already under pressure. If specialist capacity is the binding constraint, the quality of assessment at the point of entry matters more rather than less, since improving recognition upstream is also a way of allocating scarce specialist time.
Some failures never reach these datasets at all. Non-attendance at diabetic retinopathy screening is 27% higher in the most deprived areas, with younger age and minority ethnicity also independent predictors (10). These slower-moving failures rarely appear in litigation or emergency audit, though avoidable sight loss is formally recognised as a modifiable outcome within the Public Health Outcomes Framework (11).
Potential Solutions
The appropriate response is not to demand that resident doctors become ophthalmologists, but to reduce the number of decisions they need to make.
The simplest approach may be a lanyard card or QR-linked one-page guide listing four or five high-risk presentations, their timescales and a direct number for the on-call ophthalmology service, issued at foundation induction alongside other emergency contact information. Generic messaging is not sufficient on its own, since slogans such as “time is sight” already exist, yet the 35-day delay above shows that awareness alone does not change behaviour (9). FAST is the instructive comparison, since its effectiveness came not from the acronym but from linking recognition to a defined action and a pathway that accepted the patient.
A second approach is point-of-care fundus photography. In the FOTO-ED study, nurse practitioners using a non-mydriatic camera obtained diagnostic-quality images in 83% of 350 emergency department patients, in a median of 1.9 minutes (12). Emergency physicians in the same cohort had examined only 14% by direct ophthalmoscopy and missed every relevant finding; given the photographs but untrained in reading them, they identified 46% (13). That shortfall is the case for routing images to a reader rather than interpreting them locally. The required skill shifts from performing a difficult examination to recognising when it is indicated.
The third is an escalation route that is actually defined. Recognising that a presentation requires same-day specialist assessment is only useful if the clinician knows exactly how to obtain it, and poor awareness of local hospital systems is precisely what the GCA audit identified (8).
Conclusion
The clinician who first sees these patients is often the most junior in the team, and the harm that follows tends to be concentrated at the handover rather than at the bedside. That burden will not fall evenly, particularly where differences in access and delayed presentation already exist. The solution does not depend on expanding a constrained ophthalmology workforce, but on making better use of the clinicians already at the interface between emergency and specialist care. That does not require resident doctors to become ophthalmologists. It requires them to know when the eye is signalling an emergency, how quickly they need to act, and who needs to see the patient next.
References
1. Rawlings A, Hobby AE, Ryan B, et al. The burden of acute eye conditions on different healthcare providers: a retrospective population-based study. British Journal of General Practice. 2024;74:e264–e274. doi:10.3399/BJGP.2022.0616
2. Nijskens CM, et al. Funduscopy: yes or no? Hypertensive emergencies and retinopathy in the emergency care setting; a retrospective cohort study. Journal of Clinical Hypertension. 2021;23(1). doi:10.1111/jch.14064
3. The Royal College of Ophthalmologists. Written evidence submitted to the Public Accounts Committee inquiry into the costs of clinical negligence (CCN0016). November 2025.
4. Getting It Right First Time programme. Giant cell arteritis-related harm: a retrospective analysis of litigation claims data from England. Rheumatology. 2026;65(Suppl 2).
5. Baylis O, Murray PI, Dayan M. Undergraduate ophthalmology education – a survey of UK medical schools. Medical Teacher. 2011;33(6):468–471. doi:10.3109/0142159X.2010.540594
6. Schulz C, Hodgkins P. Factors associated with confidence in fundoscopy. Clinical Teacher. 2014;11(6):431–435. doi:10.1111/tct.12171
7. Hussain S, Aziz R, Cheloni R, Mohyudin M. A systematic review of the adequacy of undergraduate ophthalmology education in the United Kingdom. Cureus. 2025;17(10):e94186. doi:10.7759/cureus.94186
8. Barriers to immediate corticosteroid treatment in suspected giant cell arteritis with visual symptoms: insights from a tertiary eye care unit. BMC Ophthalmology. 2025. doi:10.1186/s12886-025-04608-5
9. Ezeonyeji AN, Borg FA, Dasgupta B. Delays in recognition and management of giant cell arteritis: results from a retrospective audit. Clinical Rheumatology. 2011;30:259–262. doi:10.1007/s10067-010-1616-y
10. Lawrenson JG, et al. Trends in diabetic retinopathy screening attendance and associations with vision impairment attributable to diabetes in a large nationwide cohort. Diabetic Medicine. 2021;38(4):e14425. doi:10.1111/dme.14425
11. Office for Health Improvement and Disparities. Public Health Outcomes Framework: commentary, February 2025. London: GOV.UK; 2025.
12. Bruce BB, Lamirel C, Biousse V, et al. Feasibility of nonmydriatic ocular fundus photography in the emergency department: phase I of the FOTO-ED study. Academic Emergency Medicine. 2011;18(9):928–933. doi:10.1111/j.1553-2712.2011.01147.x
13. Bruce BB, Thulasi P, Fraser CL, et al. Diagnostic accuracy and use of nonmydriatic ocular fundus photography by emergency physicians: phase II of the FOTO-ED study. Annals of Emergency Medicine. 2013;62(1):28–33. doi:10.1016/j.annemergmed.2013.01.010