Interviews
The findings from the interviews identified many differing opinions on the choice of screening test (s), consistent with the prior Delphi study. The qualitative data provided insights into the difficulties agreeing on the screening test(s) and provided insights into how a screening intervention could be tailored according to particular groups who might not access eye care services. Of particular interest, and relevant to the potential implementation of a future screening programme, was concerns about professional tensions between expert provider groups in eye care services, and these concerns were reflected in their respective discussions about the feasibility and acceptability of the component parts (test, site, target population) of the intervention.
Preferred site
Most interviewees agreed that screening should take place in a primary care setting. Some thought that capitalising on current organisation and infrastructure in primary care, either in high street optometry or general practice, made most sense, although this was caveated by concerns about the business ethos of high street optometry, and how this may present a barrier to recruiting high risk groups:
'That's (private eye care services) probably going to gradually increase in a way because the likes of ... [specific company names], all the cheap and nasty services... they model and they tell their optometrists that they are expected to convert 80% of the people they see into a pair of glasses which means there's no wasting time with repeating examinations and doing the things that ... they may wish to do, you know time is money, you are here to flog glasses mate, get on with it, and that's very much ... the growing Business model'. (Ophthalmologist)
The reality of the commercial influences defining high street optometry featured in optometrists' responses. Currently there is variation among UK devolved administrations as to remuneration procedures for 'high street' (community) optometrists providing eye care services, with optometrists in Scotland currently receiving considerably more than optometrists in other areas of the UK:
'Currently there's no real incentive for optometrists to do it, so it's not a pre-requisite to get involved in glaucoma management or enhanced screening.' (Optometrist, England)
Some interviewees identified the positive appeal that placing general practitioners as key players in the programme represented. However, it was recognised that this was mitigated in practice by general practitioners' retreat from primary eye health, as government policy, and indeed the public, has increasingly identified optometry as the primary provider site of eye services in the community:
"...so on the one hand there's this move and I think a number of patients are getting used to it and accepting it, that when they've got an eye problem they go to the optician, they don't come to the GP. And it would kind of seem counterintuitive if then you've got this new system aimed at eye conditions which has gone back in GP surgeries, do you see what I mean?" (Practicing and academic GP)
The use of a mobile van was commonly endorsed as was the suggestion that the diabetic retinopathy screening service could be used as a possible template for glaucoma screening or indeed, that glaucoma screening could be piggybacked onto the existing screening for retinopathy.
Target population
A minority of interviewees highlighted the need for any screening programme to be directed towards the hard-to-reach groups, such as African and Caribbean ethnic groups. However, a larger number of the interviewees disagreed with this targeted approach - their own experiences as practitioners left them wondering whether it would be fairer to screen the whole population to ensure that no one with glaucoma was missed out. The dilemma that arose, therefore, was that what appeared to be the most cost-effective option [the targeted approach] was not necessarily the fairest choice, in terms of access to services. Below is an example of a respondent calling for a targeted approach to screening:
'We need to focus particularly our attentions on populations of people from black ethnic minority groups who have a higher susceptibility to certain eye disease and also to people on low incomes, those folk are furthest from engagement with the eye care sector' (policymaker)
Whereas another interviewee acknowledged the limited evidence to make judgements on whom to screen illustrating a difference in opinion between policy makers and providers:
'I think that's [universal screening at age 50] as good as anything because we've got very little data about what will happen if you go out to the community screening glaucoma. You don't know what's happening' (ophthalmologist)
Preferred Test(s) and preferred operator
Interviewees found it difficult to differentiate between identifying test(s) relevant for a screening intervention (ie a test to initially identify those more likely to have glaucoma) and those for the full diagnosis of glaucoma. Clinicians generally stressed the necessity of using three tests - visual field examination, measurement of intraocular pressure and imaging of the optic nerve - which reflects the practice of a diagnostic strategy used by many of the interviewees. A number also pointed out that the clinical grounds for recognition of glaucoma has undergone a revision, from one based on raised intraocular pressure to one that includes optic neuropathy:
''Optometrists... the training has improved and....also the dependence on intraocular pressure has been the main criteria. That has sort of shifted more towards visual fields and optic discs. I say it in that order because they still aren't very good at looking at optic discs. You know, but they do visual fields and they attempt to have a look at the optic disc..... and that has improved the detection rate. In time I think it's increased the number of false positives being referred to secondary care quite dramatically...."(Ophthalmologist)
Discussions about preferred tests highlighted differences in opinions between the two main clinical provider groups, ophthalmologists and optometrists. While clinician groups generally agreed that optometrists should have a key role in the delivery of tests, the majority indicated that the role of optometrists was to 'moderate' between the trained technical operators and the expert ophthalmic diagnostic role.
Overwhelmingly, and primarily out of financial consideration, clinicians suggested that 'technicians' supported by optometrists trained in glaucoma detection, represented the optimum choice of test operator. There were concerns raised, however, about ensuring adequate training for both technicians and optometrists - training together with quality assurance were identified as resource intensive essentials in any future screening trial. Unfortunately it was difficult to explore this in more detail as we had difficulties recruiting 'technicians' working in ophthalmic services to the study. Of the eight technicians we approached to take part only three were willing to be interviewed. However, we did have informal telephone discussions with a small number of technicians (who declined to formally participate in the study) who indicated their concerns regarding the lack of career structure for technicians, and the already intense workload that they experienced for little remuneration in often poor working conditions. This they saw as working against the premise that, in practice, there would be adequate technicians available, and committed to, resourcing a massive explosion of their workload should a screening programme be initiated. Their insights echoed those concerns expressed by the formally recruited technician group of respondents.
A number of optometrists (especially those who had combined primary and secondary care careers) raised concerns about the structural factors mitigating against high-street optometrists being able to access courses on glaucoma detection, notably that they received no appropriate reimbursement for the cost of the courses and time out of their commercial enterprises. Additionally, a number pointed to Scotland as the country in the UK where the interpretation of the optometrists' contract potentially provided more opportunity for specialising in glaucoma.
Summary of the results of the interviews
The provider and policy perspectives regarding the configuration of an intervention for a national glaucoma screening trial indicated support for screening in primary care, using 'technicians' supported by optometrists. Potential barriers to implementation were identified. In particular the interview findings highlighted many of the complexities around a glaucoma screening programme such as the ethical issues of adopting a targeted approach to those in higher risk groups versus one of universal population provision, or on strategic and/or organisational grounds (priority setting in the NHS, adequate resourcing) and the associated constraints of a testing strategy. The study also flagged difficulties in being able to definitively identify preferred test(s), preferred operator, location and target population 'in principle and in isolation'. Rather, the choice of each of these configurations was seen to be dependent on a range of contextual factors. In particular, interviewees' opinions were influenced by their experiences of the current context of professional and organisational tensions in eye health services.
In summary:
-
Preferred screening site was a primary care setting
-
Preferred screening operator was a trained technician supported by a specialist optometrist
-
Most interviewees thought that ultimately an ophthalmic consultant must make the diagnosis of glaucoma
-
The preferred age to initiative screening was 50 years for the general population
-
Most interviewees identified the 'at risk' groups as those with a primary family history, and those of African and Caribbean ethnic origin. Respondents indicated that the age at which screening were to be initiated should be under 50 for these risk groups.
-
Potential intervention should have a low false positive rate
-
Interviewees were aware of the finite resources of the NHS and the limitations this posed on service development, but nonetheless believed that a screening trial should occur because it could prevent people going blind
-
The commercial culture of high street optometry was seen by many respondents as a barrier to capturing at risk groups. Current professional tensions between ophthalmologists and optometrists and changes in the policy and organisational determinants of eye care services that give optometrists increasing professional autonomy back, influenced respondents' preferred screening intervention properties and their broader discussions.
-
Many interviewees, while advocating a screening trial, wanted the current infrastructure of case detection, referral refinement and shared care to be capitalised upon.
-
Most ophthalmologists and optometrists in clinical practice were adamant that the screening tests must be based on current best clinical practice and that simplifying this protocol for a screening context was not acceptable practice
Economic assessment of likely cost-effectiveness
All the potential screening components (i.e. nominated test(s), location(s), test operator(s), target population and diagnostic pathway(s)) put forward by the interviewees were summarised in tabular form for consideration in the economic model [
20]. A total of 189 screening configurations were developed from interview data, further illustrating the difficulty in identifying the clinical components of a screening intervention. The multidisciplinary project management group (PMG) agreed there were no additional relevant options that could be developed from combination of components mentioned by different interviewees (consideration a. - see methods section). From this comprehensive list, a subset was immediately rejected as candidates for a screening intervention on the basis of known (from the previous model) lack of cost-effectiveness; 57 because screening required three or more tests (complex testing) or equipment costing more than £25 per testing based on a cost-effectiveness threshold of £20-30,000 per Quality Adjusted Life Year gained and 40 on the basis that the screening be undertaken by a highly trained health professional. The original economic modelling evaluation suggested that screening for glaucoma was only likely to be cost-effective if the screening is targeted at forty year olds with additional risk factors rather than general population screening [
20]; on this basis, a further 62 configurations which specified screening a general population would have been ruled out. However, it was apparent from the qualitative interviews that an equity argument was voiced in favour of general population screening and considerations should be given to screening in remote rural locations. Both cost-effectiveness and equity arguments are likely to be important to decision-makers and taking into account the findings from the economic model and the interviews the PMG agreed that the 30 remaining configurations were realistic (consideration b. - see methods section). Seven of the remaining configurations considered a mobile unit as the setting where screening would be conducted. We considered this a variation of a primary care setting particularly relevant for remote areas. Key features of the remaining 23 configurations were variations on screening location including 'high street' optometry or in general medical practice. The tests were optic nerve photography or perimetry (a measure of visual field sensitivity) with or without tonometry (a device to measure intraocular pressure), operated by paramedical staff (e.g. nurse, technician or self-assessment, the latter being possible for perimetry available on Personal Computers and self assessment tonometers.
Integrating the findings from the qualitative interviews with the modelling thus identified the components of the screening test intervention that could be implemented in a trial from a service perspective. These are general population screening, for a cohort at age forty (based on findings of the qualitative interviews of the need to balance feasibility and equity with cost-effectiveness criteria) in a primary health care setting. Screening would be conducted by ophthalmic trained technical assistants, undertaking optic nerve photography or screening mode perimetry (a measure of visual field sensitivity) with or without tonometry. The precise location of screening a community setting would need to be tailored to local circumstances illustrated by the suggestion of using widely available eye care services e.g. the 'high street' optometrist or mobile units for remote areas.