Anatomical structure of the eye conforms to expected average dimensions and this is likely to vary according to factors such as age, refractive status, and disease processes. For this reason, OCTs (and instruments such as visual fields analysers) incorporate normative values against which the data of any single patient may be compared and flagged up if falling outside this expected normalised value.
Each instrument manufacturer uses its own normative database and these are based on different numbers of previously assessed and published data sets. Typically, results falling outside the expected normalised values are flagged up in traffic light colour coding, with green representing normal limits, orange as suspect and red as outside expected limits. Useful as these are as a guide, problems may occur if they are the sole basis for a referral without other factors (such as risk factors for a particular disease, or congenital and aberrant anatomical differences).
With this in mind, the interactive exercise (C40968) was designed to gauge both yours and the opinion of the centre to which you might be referring regarding the use of colour coded normative indicators. Your responses reflect that in most areas there is established, though varied, protocols relating to OCT data referral.
To whom do you refer your suspect primary open angle glaucoma patients?
Just under 50 per cent of respondents refer suspect glaucoma patients (where OCT assessment has been included in the assessment) first to the GP. Of the remainder, most, such as the scheme for the Hillingdon hospital, have a direct referral scheme. We had some overseas respondents where direct ophthalmology liaison seems the norm. One Australian reader noted: ‘in Australia, referrals can be direct – much more streamlined than the NHS!’
Do they have an OCT? If so, which?
As might be expected, there was a range of instruments in use, the most numerous reported being either Topcon, Heidelberg or Zeiss. Those in primary care did not necessarily match those at the secondary care end.
What would be their policy regarding the referral of a patient with few apparent risk factors for open angle glaucoma (other than age) but have some OCT values flagged up as orange coded relating to the normative dataset?
There was a range of responses here. At one end of the scale, some reported that referral of any OCT data flagged up as outside the normative expectation was acceptable – for example, ‘the local glaucoma consultant would be happy to receive such referrals’. Another responded: ‘They would want us to contact them to discuss any patient flagged by OCT results and would make a decision on a patient-by-patient basis depending on the OCT results found.’ In these cases, the onus on further assessment fell upon the referral centre and the initial OCT assessment acted merely as a basic screening module.
More typical, however, were answers like this: ‘a diagnosis of glaucoma is still clinical and dependent on several risk factors so, as no OCT software platforms have 100 per cent sensitivity or specificity for glaucoma, they can wrongly overcall abnormalities in normal. So, in the absence of a field defect, high IOP or other grounds [we] would review or repeat scan in three to six months.’
Another wrote: ‘If the referral is based on the OCT alone with no other significant risk factor apart from age I doubt it will be accepted due to the high rate of false positives the OCT can produce. There are no protocols within the department for referrals as such yet.’
A further response stated: ‘The ophthalmologist I spoke to said it can still be subjective as, although the information states that the discs had a CD ratio of 0.45, she said this is borderline for treatment with the associated IOPs. However, if she had personally seen the discs and they looked OK she would recommend repeating all the tests in six months. If the discs looked slightly suspicious she would err on the side of caution and start treatment immediately.’
Finally, a few responded with answers such as: ‘There is no protocol to use OCT for glaucoma diagnosis.’ One respondent contacted six local GPs and found: ‘All said that more sophisticated technology in their optometry practice would be brilliant and would help them to make better referral decisions.’
Conclusion
Suspicion of primary open angle glaucoma needs to be based on several key elements, including risk factors (age, family history, ethnicity, myopia), and the presence of certain signs such as loss of neuroretinal tissues (as visualised by increasing cup proportion of the disc), elevated IOP for any given corneal thickness or a full threshold field defect. OCTs provide an accurate assessment of disc architecture and can very sensitively show change over time. Comparison with a normative database can only ever indicate ‘normality’ but as there is significant non-pathological variation, should only be used as an indication for further analysis as a whole.