|Home | About | Journals | Submit | Contact Us | Français|
To assess optometrists' ability to correctly identify and manage patients with different ocular conditions seen in the accident and emergency (A&E) department of an eye hospital.
Randomly selected patients presenting to the A&E department were initially examined by one of two senior optometrists and a consultant ophthalmologist. A diagnosis and a management plan were made for each patient by the optometrist and by the consultant, who was blinded to the optometrists' plan. Agreement was assessed between optometrist and consultant ophthalmologist for primary and secondary diagnoses, management plan and whether patients could be seen by an optometrist only. Weighted kappa (κ) statistics was used to assess the level of agreement in management between the two groups.
A total of 150 patients were assessed. The agreement in primary diagnosis and management outcome between the two groups were 89.3% (95% CI 83.2% to 93.8%) and 79.3% (95% CI 84.8% to 94.8%), respectively. A high level of agreement in management outcome was found (κ=0.82). No sight‐threatening conditions were misdiagnosed by the optometrist.
There was good agreement in both the diagnosis and management plan between optometrists and the ophthalmologist. This study has shown that optometrists can potentially work safely in an A&E department of a busy eye hospital.
With the ageing population, the development of shared care between different ophthalmic professions can have a large‐scale effect on healthcare delivery in the National Health Service in the UK. Over the past few years, hospital‐based shared care involving optometrists working alongside ophthalmologists to co‐manage patients with ocular disease have become well established, and there are a number of shared‐care projects which have been developed involving both community and hospital optometrists.1,2,3,4,5
To the authors' knowledge, there are no published data on hospital optometrists' ability to clinically appraise new referrals from general practitioners or community optometrists to an accident and emergency (A&E) department of an eye hospital. The aim of this study was to evaluate optometrists' ability to correctly identify and manage patients with different ocular conditions seen in the A&E department of an eye hospital.
This was a prospective observational study carried out in the A&E department at Moorfields Eye Hospital, London, UK in January 2005 over a 6‐months period. Local research ethics committee approvals were obtained before starting the study. Two senior optometrists with a minimum of 3 years of extended‐role experience and one consultant ophthalmologist who acted as a gold standard for comparison took part in the study.
The inclusion criteria included all patients who attended the A&E department, and the exclusion criteria were those patients in whom an ophthalmic diagnosis had already been made from a previous visit, or those being managed by the nurse practitioner.
Before being seen by the optometrist, each patient was initially seen by a nurse who triaged them into urgent and non‐urgent cases, and then allocated their record into separate trays. To reduce the effect of bias, the optometrist obtained a hospital record alternately from the two trays and informed consent was obtained before the optometrist examined each patient. The optometrist formulated a provisional diagnosis and a referral management plan: to see the ophthalmologist on the same day, refer to another ophthalmological subspecialty within 1 month (<1 month), refer to another subspecialty after 1 month (>1 month) and to discharge. The ophthalmologist, blinded to the optometrist's findings, made an independent assessment of the patient's diagnosis and management plan. The clinicians also recorded whether they felt it would have been appropriate for the patient to see an optometrist only in the A&E setting.
The primary outcomes were the percentage agreement in primary diagnosis and referral management between optometrists and the ophthalmologist. Other outcomes include secondary diagnosis and whether the patients could have been seen by the optometrist only.
A sample size of 150 was determined for an estimation of a 90% agreement between the two groups with 95% CI at the 5% significance level.
Descriptive statistics including mean, standard deviation (SD) and range were calculated; weighted kappa (κ) statistics were used for estimating the level of agreement in management outcome: poor if κ0.20, fair if κ=0.21–0.40, moderate if κ=0.41–0.60, substantial if κ=0.61–0.80 and good if κ>0.80.6
There were 150 patients, 78 (52%) were female and 72 (48%) male. The mean (SD, range) age was 46.5 (16.6, 10.1–90.6) years. In all, 96 (64%) patients were self‐referred, 30(20%) from community optometrists, 22 (14.7%) from general practitioners and 2 (1.3%) were referred from other hospitals.
There was an agreement in 134 (89.3%; 95% CI 83.2% to 93.8%) primary diagnoses between the optometrists and ophthalmologist. The severity of each condition was subdivided into acute and non‐acute diagnosis; there were 104 (77.7%) acute cases deemed to be potentially sight threatening (table 11).
There were four cases where the optometrists detected abnormality, but the ophthalmologist recorded nothing abnormal detected (NAD). Of the 31 (20.7%) patients with secondary diagnoses, there were 26 (83.9%) cases of agreement and 5 (16.1%) cases of disagreement (table 22).
There were 136 (90.7%; 95% CI 84.8% to 94.8%) cases of agreement in management outcome between optometrists and ophthalmologist; the weighted κ statistics indicated good agreement between the two groups (table 33).
There were 12 cases where there was an agreement in primary diagnoses but disagreement in management outcome and this is shown in table 44.. There were only 2 (1.3%) cases of disagreement in both primary diagnosis and management outcome (patients 128 and 129 in table 22).
The ophthalmologist considered that 61 (45.5%) of the 134 agreed primary diagnoses cases were suitable to be seen by the optometrist only, whereas the optometrists considered 80 (59.7%) patients to be suitable (table 11).
To our knowledge, this is the first study to evaluate the diagnostic accuracy and referral management of hospital optometrists in an ophthalmic A&E department in the UK. The results of this study showed good agreement in terms of primary diagnoses (89.3%) and referral management (90.7%) between the optometrists and consultant ophthalmologist.
There were 22.3% of cases categorised as non‐acute conditions, and this included 9 (6.7%) patients with NAD. This indicates that not all patients who present to an ophthalmic A&E department have any detectable abnormalities, and this is in agreement with other published studies.7,8,9
The ophthalmologist recorded NAD in four cases, but the optometrists made a diagnosis for each of these patients (table 22).). The reason might be because the findings from the ophthalmologist were deemed insignificant or the optometrists were not confident in diagnosing NAD in an A&E setting.
Some ocular conditions that the optometrists failed to diagnose were intermediate uveitis, marginal keratitis, multiple sclerosis‐related eye problem, Herpes simplex virus keratitis and a case of idiopathic chronic orbital inflammation (table 22).). There were two cases where both groups disagreed in primary and secondary diagnoses (patient 50 and 150), but the conditions misdiagnosed by the optometrists were not sight threatening. The prevalence of these conditions is much lower, which may reflect the lack of exposure the optometrists have with these rarer conditions and indicates that further training or support from ophthalmologists was required.
Weighted κ value for the referral management outcome indicated good agreement (table 33).). This shows that the optometrists were competent in making referral management decisions. None of the 12 conditions where there was an agreement in primary diagnoses but a difference in referral management were sight threatening (table 44).). The four cases where the optometrist indicated a same day referral was because of their limited therapeutic prescribing power, as a topical steroid was indicated for three cases. With the two cases (patient 128 and 129) of disagreement in both diagnosis and management, the optometrist was cautious in referring patients when he or she could not identify a cause for the patient's symptoms. This highlights the need of establishing an A&E protocol for conditions that can be managed by optometrists.
There were more cases that the optometrists considered they could manage by themselves if they were able to prescribe certain topical medications such as steroids and antiviral drugs. However, the conditions that both groups agreed on are mainly non‐sight‐threatening disorders—for example, dry eyes and posterior vitreous detachment, and this is in tandem with the recent introduction of legislations by the Department of Health in the UK to grant suitably trained optometrists the rights to extend their therapeutic prescribing power.10,11 Future studies in assessing whether therapeutically trained optometrists are able to prescribe safely and competently in an A&E department are warranted.
In conclusion, this study has shown that appropriately trained optometrists show strong agreement with a consultant ophthalmologist when assessing patients in an ophthalmic A&E department.
We thank Dr C Bunce and Dr M Crossland for their advice and comments on this study.
A&E - accident and emergency
NAD - nothing abnormal detected
Competing interests: None.