A review paper published in this month’s Ophthalmic and Physiological Optics (Optician 09.09.16) has drawn the conclusion ‘that the use of coloured lenses or overlays to ameliorate reading difficulties cannot be endorsed and that any benefits reported by individuals in clinical settings are likely to be the result of placebo, practice or Hawthorne effects [behaviour changes caused by being observed]’.
Perceptual difficulties when reading have been cited for many years. These include reports of print blurring, doubling or moving, glare from textual patterns, and asthenopic-like symptoms of head and eye ache.
Claims that coloured filters could alleviate these were reported by a New Zealand teacher, Olive Meares, using coloured overlays and a US psychologist, Helen Irlen, using selectively coloured spectacle lenses.
Their work gave rise to the term ‘Meares-Irlen syndrome’ to describe the symptoms, a term still widely used, including, for example, in the College of Optometrists final assessment for qualification, but one still questioned by many as to its validity as a distinct syndrome.
The term ‘visual stress’ was introduced by UK scientist Professor Arnold Wilkins to describe such symptoms which, once all refractive and binocular status influences have been managed, might be alleviated with coloured overlays and lenses specified by his Intuitive Colorimeter – a system which some readers may themselves deploy in practice. Other systems exist, such as the ReadEZ computerised filter selection system from Thomson Software Solutions or the coloured contact lens known as ChromaGen, but all rely on an assumption that the ‘visual stress’ symptoms may be managed by some selective colour filtering of the incident light to the eye.
Reports of success with such systems abound, and the use of colour overlays or lenses is now an accepted management approach for many eye care practitioners, teachers, educational psychologists and NHS departments. Indeed, endorsements by the popular media and some charities abound. But is the evidence for such an intervention robust enough to validate this approach as a treatment for a set of symptoms so recently defined and, in the case of ‘visual stress’, yet to be recognised by the World Health Organisation or the American Psychiatric Association? The paper by Griffiths et al claims to be such a review of the published evidence for the effect of coloured overlays and lenses on reading.
Literature review
Previous reviews in this area have concluded that the published evidence is of too poor quality to afford meaningful conclusion, have been guilty of poor design or have derived conflicting views.
However, the Griffiths et al paper used three research search engines (Medline, PsychInfo and Embase) and applied search filter terms based on colour (tint, overlay and so on), reading (text, print and so on) or reading difficulty/visual stress associated terminology. Of the subsequently listed 244 papers, 195 were rejected after indepen-dent review as having no evidence of a randomised control trial (RCT) structure or were anecdotal in nature or had no original data within. A further four studies were identified from the references of the remaining 49 and, another four were suggested by the reviewers. The final 51 studies for review were arrived at after four were excluded for a lack of a control group, one used no formal reading measure and one did not use overlays or lenses.
Each paper was then evaluated using the Cochrane Collaboration for assessing bias. This includes consideration of selection, performance, detection, attrition and reporting bias.
The Griffiths et al paper firstly argues the results of the studies based on the intuitive system need to be ‘viewed with caution’ with a ‘high risk of generating false positives’. The three masked RCTs were ‘each prone to bias’. These included bias in recruitment from non-generalised populations or by advertising to an interested group, lack of a placebo group, or a high attrition bias, such as non-reporting of a significant percentage of data.
Regarding Irlen studies, the use of such coloured lenses or overlays ‘cannot be endorsed’. The two trials at the lowest risk of bias showed no improvement in the outcome. Indeed, the use of Irlen techniques has led to a perceived over-diagnosis of readers as having Irlen syndrome. No direct conclusion as to who might benefit from this is stated.
Too few studies used the ChromaGen/Harris system and these did not have a placebo-lens or ‘no lens’ control so provide only ‘weak’ information. The disparate ‘other’ studies were analysed individually as relate to different systems that are not easily grouped. The overall conclusion was, however, that ‘none of these studies contain strong evidence that the use of overlays or lenses leads to benefits in the measures of reading in individuals with reading difficulties and/or visual stress’.
Can we trust this paper?
Griffiths et al rely heavily upon the Cochrane Risk of Bias tool. Under double-masked conditions, some recent reports suggest agreement between judges when using the ‘Risk of Bias’ tool is poor and so assessments of this kind may not be valid.
Furthermore, researchers with experience in this area have called into question the conclusions of the authors. Professor Bruce Evans has recently written ‘Griffiths and co-authors are well known for their sceptical views on the existence and treatment of visual stress’. Their methodology has been called into question by some.
As Prof Evans has pointed out, there appears to have been some selection in the results cited from studies included in the review. For example, of the 22 (sic) studies involving intuitive systems, ‘eight found statistically significant improvements in reading continuous or discontinuous text with coloured filters. Of the four studies that did not find such an improvement, two found a significant reduction in symptoms and a third found a significant improvement with one reading test (WRRT) but not another (Neale)… to completely ignore the key findings obtained by the original researchers is to conceal an important feature of the literature.’
Whether using such systems or not, I suggest it is important to know there is a paucity of randomised control trial data published to support any beneficial effect of coloured lenses or overlays. A full meta-analysis is not possible, according to Griffiths et al, due to the high bias risk endemic throughout the studies that have been published. However, it does assume reliance on the bias assessments.
Prof Evans responded: ‘An alternative systematic review, which I would argue has more appropriate selection criteria, has recently been published in the refereed and open access Journal of Optometry. By reading both reviews I think that interested practitioners and researchers will gain a balanced insight into this fascinating topic. The findings of our systematic review support the existence of VS and its treatment with coloured filters.’
The need for further studies is something that cannot be denied. This can be stated with confidence based on the current approach to validating scientific interventions aimed at changing a perceived state or concern.
So what now?
How to use this information is now where the real debate begins and also, to my mind, where some of the real questions begin. Should I deny children who present to me with clearly reported symptoms any management options that experience tells me have been successful for some. I have attended events and, indeed, spoken with patients who will profess to the ‘life-changing’ benefit of their tinted lenses or overlays. Should future patients be denied this opportunity while the evidence base accrues or stagnates?
Conversely, can I justify charging for offering a confidently described solution which offers a hope or belief to someone looking for an answer to a problem they perceive in themselves or others. Without the evidence base, are we no more than preachers offering a readily consumed ‘answer’?
Just as many low vision patients will report positive effects with some interventions, the perception of a benefit, where no options have previously been offered, always needs to be balanced with the thought that this is a vulnerable group that by definition will gravitate to anything that may address how they view their current perceptual status. This is especially important in children where such perceptions are influenced by a very wide range of social, education and political influences.
Optometric practice is, in many areas, a balance between the subjective and the objective. Though we should not dismiss anecdotal evidence of improved perceptions, once a system is viewed as an acceptable course of therapy, it is not unreasonable to assume there is a good solid evidence base underlying the intervention. This is perhaps more important where funding is being diverted towards such interventions, or indeed where the intervention is addressing a target where other external influences upon performance (perhaps some more insidious, such as societal expectation) may be a factor and so an assumed syndrome and treatment offers an all-too-easy blanket approach to what is in fact a much wider and variable concern.