Development of a Clinically Valid Protocol for the Assessment and Investigation of Optic Neuropathy

AIMS: 1. Protocol development. To develop a simple, clinically-valid protocol to aid ophthalmologists' initial assessment and management of patients with previously-undiagnosed optic nerve disease. 2. Clinical characteristics. To compare the clinical features and final diagnoses of two populations of new-onset optic nerve disease (optic neuropathy) patients in Birmingham, UK,and Brisbane, Australia; and to determine the efficacy of imaging and laboratory investigations in the diagnosis of these patients. 3. Common diagnostic errors. To assess the errors made by ophthalmologists when initially assessing patients presenting with optic nerve disease; to determine the prevalence of such errors; and to determine whether such mistakes impact on patients’ clinical outcomes. 4. ICON. To propose the term Investigation-negative Corticosteroid-responsive atypical Optic Neuropathy (ICON) as a unifying descriptor for optic neuropathy patients whose clinical features are atypical for optic neuritis, anterior ischaemic optic neuropathy or glaucoma, have negative investigations, and improve rapidly with an empiric trial of corticosteroid treatment; and to analyse these cases in detail. 5. Education package. To develop and test an education package incorporating the optic neuropathy management protocol; and to distribute this package to all practising and trainee ophthalmologists in Australia and New Zealand. PATIENTS: One hundred consecutive Birmingham patients and 172 consecutive Brisbane patients with new-onset previously-undiagnosed optic neuropathies were enrolled in two separate prospective case series. METHODS: 1. Protocol development. A practical protocol for the clinical assessment and initial management of optic neuropathy patients was developed. The protocol consisted of an initial clinical assessment protocol, followed by a standard investigation protocol for patients who could not be clinically diagnosed with certainty. 2. Clinical characteristics. The clinical courses of the optic neuropathy patients were analysed in detail, including clinical features, diagnosis, treatment, visual and systemic outcome, and the diagnostic yield of imaging and laboratory investigations. A novel neuro-ophthalmic computer database using multiple searchable criteria was developed and used to record data for the study populations. 3. Common diagnostic errors. The diagnoses and management by the initial examining practitioners were assessed. Errors made by these practitioners were analysed in detail, including their contribution to avoidable adverse patient outcomes. 4. ICON. Patients meeting the diagnostic criteria for ICON were analysed as a separate subset,including pre- and post-corticosteroid visual function, length of immunosuppressive treatment required and outcome. 5. Education package. The management protocol was incorporated into an education package aimed at the majority of ophthalmologists who have not undergone advanced neuro-ophthalmic training. The package was tested by ophthalmic trainees in Brisbane and Melbourne, using mock clinical scenarios to estimate its potential efficacy in improving patient outcomes by reducing serious initial management errors. RESULTS: 1. Protocol development. The protocol proved to be practical and effective in everyday clinical situations. Follow-up demonstrated that diagnoses made with the protocol were accurate and stable over time. 2. Clinical characteristics. The relative prevalence of the different types of optic neuropathy was similar in Birmingham and Brisbane. Approximately one third of patients had an optic neuropathy which could be diagnosed on purely clinical grounds, the other two thirds requiring investigation; a third of patients had a treatable optic neuropathy; half of the patients experienced some visual recovery (a third with treatment, two thirds spontaneously); one in six patients had a life-threatening but treatable disease, first presenting to an ophthalmologist as an optic neuropathy; and one in seven patients undergoing neuro-imaging for chronic optic neuropathy was found to have a previously undiagnosed causative tumour. 3. Common diagnostic errors. Thirty-two percent of the Birmingham optic neuropathy study patients and 25% of the Brisbane study patients had been misdiagnosed by their initial examining practitioner. “Optic neuritis” was over-diagnosed and often falsely diagnosed; two thirds of the patients initially diagnosed as having typical optic neuritis did not in fact have this disorder. More than two thirds of patients first presenting to their ophthalmologist with blurred vision due to a previously undiagnosed brain tumour were initially misdiagnosed or not appropriately investigated. Two thirds of the misdiagnosed patients suffered an avoidable adverse outcome. Twenty-three different major initial diagnostic process errors were identified. 4. ICON. Approximately 8% of the patients in each of the Birmingham and Brisbane optic neuropathy populations were diagnosed as having Investigation-negative Corticosteroid-responsive atypical Optic Neuropathy (ICON). Despite frequently poor initial visual acuity (6/60 or worse in half the patients), more than 90% of ICON patients achieved a good final visual outcome (VA 6/9 or better) following treatment. 5. Education package. An education package “Optic nerve disease: more than meets the eye” was developed, incorporating the optic neuropathy assessment protocol as well as a brief introduction explaining the need for the protocol and its suggested use. An initial test of the package with ophthalmic trainees using mock clinical scenarios yielded an improvement in diagnostic accuracy from 39% (without the package) to 90% (with the package), and an improvement in management scores from 42% (without) to 85% (with). The education package was forwarded to all practising and trainee Australian and New Zealand ophthalmologists by RANZCO. An accompanying self-test module was approved for continuing professional development (CPD) points. CONCLUSIONS: Patients with sight- or life-threatening diseases first presenting as optic nerve dysfunction are frequently misdiagnosed by the initial examining ophthalmologist, often leading to avoidable adverse outcomes. A practical protocol has been developed to assist ophthalmologists in the initial assessment and immediate management of patients presenting with new-onset optic neuropathy.