Optometry in Practice 2014
Volume 15
Issue 3
87 – 100
The effect of low refractive corrections on rate of reading Claire I O’Leary1,2 BSc(Hons) PhD MCOptom, Bruce JW Evans1,2,3 BSc PhD FCOptom DipCLP DipOrth FAAO FBCLA and David F Edgar2 BSc FCOptom 1Cole
Martin Tregaskis Optometrists, Brentwood, Essex for Applied Vision Research, City University London 3Neville Chappell Research Clinic, Institute of Optometry, London 2Centre
EV-21872 C-37698 1 CET point for UK optometrists
Introduction The core function of optometrists is the prescribing of spectacles to alleviate symptoms and improve visual function, yet the criteria for prescribing spectacles have received little attention in the clinical and scientific literature. The decision whether to correct a small refractive error in particular can be difficult, for there is no clear cut-off point between normality and abnormality and there is therefore a need for prescribing guidelines to aid this decision. However, in a previous paper we found the current prescribing guidelines to be conflicting and rarely based on clinical research (O’Leary and Evans 2003). In addition, this survey of optometrists discovered that a wide variety of criteria were used when deciding whether to prescribe, particularly in the case of low refractive errors where visual acuity (VA) may have limitations as an indicator of benefit. Several other more recent publications have highlighted the lack of uniformity in prescribing criteria for hypermetropia (Cotter 2007; Farbrother 2008; Leat 2011; Reiter et al. 2007). Our previous research showed that in cases of low refractive errors optometrists are likely to rely on symptoms when deciding whether to prescribe (O’Leary and Evans 2003). However, although eliminating symptoms is obviously important, it has been shown that symptoms relating to low refractive errors can be vague and non-specific (Brookman 1996) and that the presence of symptoms may not correlate with the magnitude of the anomaly found (O’Leary and Evans 2003) or the final prescription given (Ball 1982b). Additionally, practitioners might wish to prevent symptoms occurring in the first place, and some patients, especially children, may not recognise symptoms until they are eliminated. In a previous paper we used a rate of reading test to investigate at what point it became beneficial to correct heterophoric anomalies (O’Leary and Evans 2006). In the research described here we have used the same dynamic reading test to investigate at what point correcting small degrees of hypermetropia, presbyopia and astigmatism becomes beneficial.
Optometric testing of visual performance The main method used by optometrists to measure visual performance is VA, typically evaluated using a test chart. In the UK, Snellen charts are most commonly used for checking distance acuities, and a near point chart used for reading,
despite both types of chart having significant drawbacks (Bailey and Lovie 1976; McGraw and Winn 1994; Wolffsohn and Cochrane 2000). For example, with most near point charts the words can easily be guessed from the context of the passage and the surrounding words and the lowest line is often larger than the limit of VA. Various near charts of improved design are available (Bailey and Lovie 1976; Evans and Wilkins 2000; Wolffsohn and Cochrane 2000); however, these tests still allow the participant to view the target for an unspecified time period and are essentially static tasks: the eyes are relatively steady and are not making the large number of saccades that are characteristic of reading. For distance vision, VA reliably detects myopia but not hypermetropia or astigmatism in children (O’Donoghue et al. 2012). In modern societies, perhaps the most commonplace demanding requirement for the visual system is reading text. Typically, people wish to read text as quickly and accurately as possible and they expect optometric interventions to help them with this. Rather than using acuity tasks that obtain a threshold based on angular resolution, we hypothesise that it may be clinically more relevant to use a task based on the rate of reading of detailed text. Wilkins et al. (1996) developed the Wilkins Rate of Reading Test (WRRT). The WRRT (Figure 1) employs small, crowded text including 10 of the most commonly occurring words in the English language (eg to, for, and, but, see, the). The test is therefore relatively independent of reading skill and does not assess linguistic or semantic factors. In visual terms, the test has been made particularly demanding by reducing the horizontal spacing between words and printing the text in a small typeface. The text consists of words printed in Times font at N9 size (typical letter body 1.6 × 1.5mm) and is set single-spaced with four-point (0.36mm) spacing between words. There are 10 lines with 15 words on each line, set as a paragraph 72.5mm wide and 33.4mm high with an interline space of 3.15mm. The words are arranged in random order and the participant is instructed to read them out loud as quickly as possible for maximum accuracy, with the score taken as the number of words correctly read in 1 minute. The WRRT results are repeatable (Wilkins 2002; Wilkins et al. 1996) and would appear to be very dependent on dynamic visual skills and to require sustained binocular single and clear vision.
Date of acceptance: 12 August 2014. Address for correspondence: The Institute of Optometry, London, SE1 6DS,
[email protected] © 2014 The College of Optometrists
87
C O'Leary et al.
the WRRT. In summary, the goal of the present research is to investigate whether the WRRT could identify an effect on visual performance of prescribing borderline refractive corrections.
Methods Subjects
Figure 1. A passage from the Wilkins Rate of Reading test. (Image kindly provided by Professor Arnold Wilkins.) The WRRT was originally developed to assess the effect of coloured filters on reading in children with reading difficulties (for reviews, see Allen et al. 2012; Wilkins 2002), where it has been shown that the coloured filter is likely to be beneficial if there is an improvement in the rate of reading of >5% (Kriss and Evans 2005). Children who show a significant initial improvement in the rate of reading with a coloured overlay are likely voluntarily to use an overlay for prolonged periods (Jeanes et al. 1997; Wilkins et al. 1996), indicating a benefit during everyday reading. Subsequently, the WRRT has also been successfully used to demonstrate the effects of coloured overlays on the reading ability of children with autism (Ludlow et al. 2006) and the effect of head tilt on reading speed in normal readers (Firth et al. 2007). In previous work with the WRRT the most commonly used cut-off criterion for a clinically significant improvement in reading speed is 5% (see Discussion). If an optometric correction for a refractive anomaly would be likely to help symptoms, in particular to make prolonged and/or detailed visual tasks more comfortable, then we hypothesise that the WRRT could be used to identify the benefit from the refractive intervention. Additionally, if an optometric intervention is likely to bring about an improvement in visual performance during everyday tasks (eg in office-based activities), then it seems likely that such a benefit could be identified as improved performance at
Subjects were selected from patients attending a community optometric practice for routine eye examination. This clinic is a busy independent practice with five optometrists in a town centre, with a cross-section of patients that is likely to be typical of primary eyecare practices in the UK. All eligible patients were given a full eye examination (Evans 2007), including tests of ocular health (ophthalmoscopy, pupil reactions, visual fields, and, in patients over the age of 35 years, tonometry), refraction (including monocular and binocular visual acuities at distance and near) and binocular coordination (at distance and near: cover test and horizontal and vertical Mallett aligning prism; at near: ocular motility, near point of convergence, amplitude of accommodation, Maddox wing test, horizontal fusional reserves, Randot 1 shapes and circles stereoacuity, and the Mallett foveal suppression test). All patients meeting the inclusion criteria were invited to return for a research appointment, unless they also met one of the exclusion criteria (Table 1). All five optometrists working at the practice use similar techniques for determining refractive errors. For example, all use the crossed cylinder technique for astigmatism and for near addition use the near working distance to estimate the initial presbyopic add, before modifying this result using plus and minus lenses until the best subjective vision is obtained. Subjects were only allocated to one category, and if a subject met the inclusion criteria for more than one of the groups being examined, then the individual was placed in the highest category in Table 1 for which s/he was eligible. This method was an attempt to maximise the numbers in each group. A total of 208 subjects participated: 32 hypermetropes with a mean age of 14 years (age normally distributed, range 6–35 years) and with a refractive error in the eye with
Table 1. Participant inclusion and exclusion criteria
88
Inclusion criteria
Exclusion criteria
Hypermetropic group • Subjective refraction reveals the least astigmatic eye to have astigmatism of ≤–0.50DC • Aged 40 years or under • Retinoscopy or subjective refraction in the least hypermetropic eye, of +0.75 to +1.75DS inclusive Presbyopic group • Subjective cylinder in the least astigmatic eye of ≤–0.50DC • Aged 35–60 years inclusive • Subjective reading addition of between +0.50 and +1.50DS Astigmatic group • Any age • Astigmatism found subjectively in the least astigmatic eye of between –0.50 and –1.75DC inclusive
• Central visual field loss (within 25°) • Patients who either cannot read, or whose binocular visual acuity is not good enough to be able to read the text in the Wilkins Rate of Reading Test (which is N9 in size) • Nystagmus • Patients attending solely for contact lens aftercare • Patients with incomitant deviations (abnormal head position may confound results) • Patients with induced astigmatism (eg from contact lenses)
The effect of low refractive corrections on rate of reading
the lowest hypermetropic error between +0.75 and +1.75DS; 58 presbyopes (mean age 46 years; range 39–58 years) with a variety of distance refractive errors (within the constraints specified in Table 1) and a reading addition between +0.50 and +1.75DS; and 118 astigmats (mean age 46 years; range 7–79 years) with a cylindrical prescription in the eye with the lowest cylindrical power of between –0.50 and –1.75DC. All subjects gave informed consent to their participation in the study. The study was approved by the Research and Ethical Committee of the Division of Optometry and Visual Science at City University London, and it conformed to the tenets of the Declaration of Helsinki.
Procedure The orthoptic tests listed above were carried out as recommended in the clinical literature (Evans 2007). Near VA was measured using a Lighthouse Bailey–Lovie design logMAR near VA chart at 40cm, monocularly and binocularly. The procedure adopted was that described by Elliott (1997), with each letter read correctly contributing to the final score. The WRRT consists of four differing passages of text. To check that all subjects were able to read the text, and that the instructions were understood, they were initially asked to read the first line from the fourth passage whilst wearing their current spectacles at their usual reading distance, or without spectacles if none had previously been prescribed. All the subjects were then asked to read the WRRT twice with the prescribed lens and twice with the control lens (size lenses or SER lenses, as explained below), in an ABBA configuration (explained below), with the order of testing determined using computer-generated random numbers, allocated to subjects at their research appointment. For example, the randomisation meant that for approximately half the subjects the prescribed lens was condition A. The subjects were instructed to read the paragraph out loud as quickly as possible for maximum accuracy, with the text held at their normal reading position and with normal consulting room light levels (approximately 125cd/m2). The subjects were allowed to set their own reading distance so that they were tested at the distance at which they usually read. A different passage of text was used for each of the four measurements, with no interval between each run apart from the time taken to change the lenses in the trial frame. The results of the two runs with the refractive correction in place were averaged to give the subject’s rate of reading with the prescribed lens. The rate of reading with the control lens was calculated in the same way. For the astigmatic group and the presbyopic group, the prescribed lens used for the study was that found for reading based on the final subjective refraction during the subject’s routine eye examination. For the hypermetropic group, the prescribed lens used was taken as the lower power of either the non-cycloplegic retinoscopy or the non-cycloplegic subjective refraction. This procedure follows the common clinical practice of prescribing the weakest correction necessary to alleviate symptoms for pre-presbyopic hypermetropes. It has also been reported that over-plussing the distance correction is the
most common reason for spectacle non-tolerance (Hrynchak 2006). Cycloplegic refractions are recommended if there are suspicious symptoms or family history, an eso-deviation, low accommodation, reduced VA or stereoacuity, unstable or abnormal non-cycloplegic refraction or spasm of the near triad (Evans 2007). Viner (2004) suggested that cycloplegia is of little benefit in a symptom-free cooperative child with no abnormal findings. As the purpose of our research was to provide guidance for the possible prescribing of borderline refractive corrections encountered in everyday optometric practice, we felt it would be more useful to concentrate on the non-cycloplegic refractive findings. In view of the age range of our subjects, a cycloplegic refraction was not generally necessary for clinical reasons. Four of the subjects in the hypermetropic group had previously had a cycloplegic refraction (although not as part of the present research), all of which were undertaken at least a year prior to their research appointment (mean age of these subjects at research appointment 8.5 years, mean age at cycloplegic refraction 6.25 years). The cycloplegic refractions were approximately +1.00D higher than the non-cycloplegic findings. As most practitioners would not prescribe the full cycloplegic refraction for patients with this clinical profile, but would reduce the correction by between 1.00DS and 1.50DS (Viner 2004), the final prescription given would be similar to the non-cycloplegic findings. The issue of cycloplegic refraction is returned to in the Discussion. For the hypermetropia and presbyopia groups we used plano-size lenses as a control, since plus-power lenses prescribed for hypermetropia and presbyopia create spectacle magnification. The size lens chosen for each subject as the control condition provided zero refractive power but gave approximately the same spectacle magnification as the prescribed lens. For astigmatism, the spherical equivalent refraction (SER), calculated as the spherical power plus half the cylindrical power, was used as the control lens. This lens should place the circle of least confusion on the retina. If the subject was also presbyopic, the reading addition found at the individual’s last eye examination was added to the SER. Many studies on astigmatism have used the SER in their analysis (Abrahamsson et al. 1988; Fulton et al. 1980). In order to minimise bias, one practitioner placed the appropriate lenses in a trial frame, and another practitioner, who was unaware of which lenses were being used (prescription lens or control size lens), conducted the WRRT. Subjects were also unaware of which lenses were being used, and the study was therefore double-masked. The subjects were instructed on how to perform the WRRT and were told that the purpose of the study was to determine at what point it would be beneficial to correct borderline refractive anomalies. They were told that they would be tested with two sets of lenses, but were not told that one of these sets was a control. The term ‘prescribed lens’ is used in this paper to indicate that this lens was individually prescribed for the present research based on refractive findings, in contrast with the control lens, which was chosen as described in the next section. It is not implied that all the subjects would have been prescribed lenses for use outside of the research (indeed, some were not). 89
C O'Leary et al.
To give an example, a presbyope with a distance refractive error of +2.00DS and a near addition of +1.50DS would have been tested under two conditions: (1) with a prescribed lens of +3.50DS; and (2) with a control lens, as described below, placed in the trial frame over the +2.00DS distance correction. In the Results and Discussion, this participant would be classified on the basis of the reading addition of +1.50DS.
(a)
When the participants’ refractive errors were determined, their optometrist would have followed their usual practice, with the result that in some cases the refractive findings would have been modified (eg a partial correction would have been prescribed in some cases); see Discussion. (b)
Control lens The size lenses used as a control for the hypermetropic and presbyopic groups had no refractive or prismatic power, but gave the same magnification (O’Leary 2009) as the prescribed lens being tested (Table 2). To test whether the size lenses were an adequate control, and that the increase in magnification alone would not significantly influence the WRRT results, a control group of 40 subjects undertook the WRRT twice with the size lenses over the top of their own prescription (taken as the subjective refraction from their most recent eye examination) and twice without. For half of the control group a size lens producing 1% magnification was used, and for the other half 2%. The results of the control group are described elsewhere (O’Leary and Evans 2006) and demonstrated that the size lenses are an appropriate control and did not generate a treatment effect.
(c)
Table 2. Choice of size lenses for control condition in hypermetropic and presbyopic groups Spherical lens power (DS)
Spectacle magnification
Spectacle magnification (%)
Closest size lens (%)
+0.50
1.0078
0.8
1
+0.75
1.0118
1.2
1
+1.00
1.0159
1.6
1.5
+1.25
1.0198
2.0
2
+1.50
1.0239
2.4
2
+1.75
1.0280
2.8
3
Results The data from the hypermetropia group and the presbyopia group were significantly different from a normal distribution (Kolmogorov–Smirmov test, P = 0.039 and 0.037 respectively). Therefore, non-parametric statistics are used for comparative analyses. An initial inspection of scattergrams revealed that there was no detectable correlation between the difference (reading speed with intervention – speed with control) and the reading speed with the control (O’Leary 2009). This confirmed that it was appropriate to use the difference in the analyses below.
Figure 2. For each participant, the difference between the rate of reading (words per minute: wpm) with the prescribed lens and rate of reading with the control lens was calculated. The graphs show the difference in wpm ( y axis) for each individual for each refractive error (x axis) for each group: (a) hypermetropia (in dioptre sphere); (b) presbyopia (reading addition in dioptre sphere); (c) astigmatism (cylindrical power in dioptre cylinder). Positive values represent better performance with the prescribed lens than with the control. Individuals with an improvement in the Wilkins Rate of Reading Test of >5% are represented by a black circle, and those with a decrease of >5% by a black line.
Hypermetropia group To obtain an overview of the data a scattergram was plotted showing the difference in rate of reading (words per minute: wpm) between the prescribed lens and control lens for each
90
The effect of low refractive corrections on rate of reading
subject (Figure 2a). This scattergram revealed that 18 of the 32 subjects (56%) with hypermetropia between +0.50 and +1.75 inclusive had an improvement on the WRRT. Seven of the 32 subjects (22%) had an improvement in the WRRT of >5% (represented by the black circles). There were also seven subjects (22%) who had a decrease in the rate of reading of more than 5% (represented by the black lines). Although one subject with 1.00D of hypermetropia had a very small decrease in the rate of reading with the prescribed lens (2 wpm slower), this subject had a very slow general reading speed (31.5 wpm with the intervention and 33.5 wpm with the control), and therefore this equated to a 5.97% difference in wpm. Those subjects with at least a 5% decrease in the rate of reading were distributed fairly evenly across the range of hypermetropic powers. The subjects were divided into two subgroups according to spherical refractive error: +0.50 to +1.00 (n = 24) and +1.25 to +1.75 (n = 8). These subgroups were chosen in order to divide the range of refractive errors tested into two categories of equal dioptric range (+0.50, +0.75, +1.00 in one category and +1.25, +1.50, +1.75 in the other), whilst maintaining a useful number of subjects in each group. In both subgroups, the mean and median WRRT results were faster with the prescribed lens than with the control. However, the rate of reading was not significantly different with the prescribed lens or with the control lens in either of the subgroups (Wilcoxon signed ranks test).
Presbyopia group The scattergram of individual data (Figure 2b) showed that 40 out of the 58 tested (69%) had an improvement on the WRRT. Twenty-three subjects out of the 58 tested (40%) had an improvement on the WRRT of >5%, and the difference in the rate of reading increased as the reading addition increased (Figure 2b). There were only seven subjects (12%) who had a decrease in their rate of reading of >5%, and these subjects were evenly distributed between 0.50D and 1.25D reading additions. The data were then divided into subgroups relating to the power of the reading addition: +0.50 to +0.75DS (n = 24), +1.00 to +1.25DS (n = 26) and +1.50 to +1.75DS (n = 8). As there were more subjects in the presbyopia group compared to the hypermetropia group, the presbyopia group was divided into three subgroups, each containing two prescribing step sizes (eg +0.50 and +0.75). In each subgroup, the mean and median WRRT results were faster with the prescribed lens than with the control. A Wilcoxon signed ranks test was performed in thethree subgroups on the WRRT data to determine if the subjects performed significantly better with the prescribed lens or with the control lens. The rate of reading was not statistically significantly different (P = 0.33) between the conditions in the 0.50–0.75D subgroup, but the performance was marginally better with the prescribed lens than the control in the 1.00–1.25D (P = 0.054) subgroup and significantly better with the prescribed lens than the control in the 1.50–1.75D (P = 0.017) subgroup. Further analysis (Kruskal-Wallis test comparing the improvement in WRRT in the three presbyopia groups)
confirmed a statistically significant difference between the WRRT result in the different groups (P = 0.020). Post hoc pairwise comparisons with a Mann–Whitney U test revealed no significant difference between groups 1 and 2 (P = 0.35), a more marked difference between groups 2 and 3 (P = 0.028) and strongly significant difference between groups 1 and 3 (P = 0.005). Since in this analysis each group is involved in more than one comparison, a Bonferroni adjustment is prudent, whereby only the comparison between groups 1 and 3 maintains significance (adjusted P = 0.015).
Astigmatism group Of the 118 subjects in the astigmatism group, the scattergram (Figure 2c) showed that 67 (57%) had an improvement on the WRRT with the prescribed lens. Twenty-six subjects (22%) had an improvement in the WRRT of >5% when reading with the prescribed lens and 14 subjects (12%) had a reading rate that was at least 5% slower. Twenty of the 26 subjects with an improvement of >5% and 11 of the 14 subjects with a decrease >5% had a cylindrical correction of less than –1.25DC. As there was a large number of subjects with astigmatism, the astigmatism data were divided into three subgroups of increasing cylindrical power: –0.50 to –0.75 DC (n = 73), –1.00 to –1.25 DC (n = 35) and –1.50 to –1.75 DC (n = 10). A Wilcoxon signed ranks test indicated the rate of reading was just significantly faster with the intervention in the –1.00 to –1.25DC group (P = 0.049), but the conditions did not differ significantly in the other two groups (P > 0.38). The significant finding was checked with a Kruskal-Wallis non-parametric one-way analysis of variance comparing the dependent variable of improvement in WRRT in the astigmatism groups. The difference between the groups was not statistically significant (P = 0.17). Therefore, post hoc pairwise comparisons were inappropriate.
Astigmatic decompensation The commonplace clinical convention of recording refractive errors as ‘sphere/cylindrical power x axis’ was adopted for the statistical analysis above. We checked the findings using an astigmatic decompensation method (Thibos et al. 1997) and these analyses confirmed the results described above (O’Leary 2009).
Axis direction Some authors state that it is important to consider the axis of astigmatism before deciding whether to prescribe and recent research with induced astigmatism supports the view that against-the-rule astigmatism has a greater impact on acuity than with-the-rule (Wills et al. 2012). For those individuals with oblique astigmatism, the effect is considered to be even greater (Rabbetts 1998), and Ball (1982a) suggests that correcting small amounts of oblique astigmatism may significantly improve vision. Recent experimental data support this view (Kobashi et al. 2012). The data were therefore reclassified into three categories: with-the-rule (negative axis 180° ± 15°), against-the-rule (negative axis 90° ± 15°) and oblique astigmatism (all other cases). The modal cylinder powers in these subgroups were –1.00DC (range 91
C O'Leary et al.
–0.50 to –1.75), –0.75DC (range –0.50 to –1.50) and –0.50DC (range –0.50 to –1.75) respectively.
an improvement of >5% in the WRRT and correctly identify 69% of those who did not have this improvement.
Subjects with oblique astigmatism performed significantly better on the WRRT with the intervention compared to the control (n = 38, Wilcoxon signed ranks test P = 0.03). In the with-the-rule and against-the-rule groups the WRRT results did not differ significantly with the intervention compared to the control (Wilcoxon signed ranks test P > 0.22).
Discussion
Receiver operating characteristic (ROC) curves ROC curves show the sensitivity of an intervention (ability to identify correctly subjects with the anomaly) plotted against 1-specificity (specificity is the ability to identify correctly subjects who do not have the anomaly). The closer the curve is to the top left-hand corner of the graph, the more effective the criterion is for identifying those subjects with an improvement in the WRRT. We generated ROC curves to determine which lens power would best predict a >5% improvement in the WRRT. In other words, for each lens power we plotted the sensitivity vs 1-specificity for delivering a >5% improvement. This analysis was confined to variables where the repeated measures Wilcoxon signed ranks test showed a significant effect of some levels of refractive correction on reading speed; the ROC curves for all variables are reproduced elsewhere (O’Leary 2009). As noted in the Introduction, a >5% improvement criterion has been used in previous work (Wilkins 2002), and this precedent was followed here (see Discussion). For presbyopia, the WRRT results showed a significant improvement in the WRRT for powers above +1.00DS. Figure 3a shows that a prescribing criterion of +1.00DS or above would correctly identify 74% of those subjects with a >5% improvement in WRRT, but would only correctly identify 51% of those subjects who did not have a >5% improvement. If this criterion is increased to +1.25DS or above, the sensitivity decreases to 48%, but the specificity improves to 77%. The ROC curve for oblique astigmatism (Figure 3b) shows that correcting oblique astigmatism of 1.00DC or over would correctly identify 67% of those subjects who had (a)
Hypermetropia Much of the previous research on hypermetropia has investigated whether correcting hypermetropia in young children prevents strabismus or amblyopia (Anker et al. 2004; Ingram et al. 1985; Lyons et al. 2004; Shea and Gaccon 2006). One refractive criterion for abnormal hypermetropia in young children has been considered to be +2.00DS (Ingram et al. 1985). This value was therefore used as the maximum limit in our criteria for the selection of subjects, in an effort to determine whether correcting smaller amounts of hypermetropia would improve reading performance as assessed by the WRRT. There has been very little previous research into the benefits of correcting smaller amounts of hypermetropia, but some notable papers found that hypermetropic spectacles of +0.50DS resulted in an improved reading performance (Pierce and Karlin 1957) and reduced physiological activity (Greenspan 1970). However, they cautioned that the effect was small and had only been tested on a short-term basis. Greenspan added that excessive plus lens power caused a decrement in performance, and that plus lenses should not be prescribed unless indicated by the clinical findings (Greenspan 1970). In a review of the literature in 1985, Press concluded that low-powered plus lenses may help some individuals and hinder others and that there was no single clinical measure able to predict the appropriate power. He also suggested that the physiological effects of plus lenses may be secondary to the postural changes induced by the lens, such as a closer working distance. There is also some evidence suggesting that hypermetropes have worse visual perceptual skills than myopes and emmetropes (Rosner and Gruber 1985) and that early correction of hypermetropia may result in better visual perceptual skills than later correction (Roch-Levecq et al. 2008; Rosner and Rosner 1986). Our research showed that correcting small hypermetropic refractive errors up to +2.00DS did not significantly improve performance on the WRRT, and this supports the opinion that small hypermetropic refractive errors need not be corrected unless clinically necessary based on the professional (b)
Figure 3. Receiver operating characteristic curves showing 1-specificity (x axis) and sensitivity ( y axis) for each refractive power: (a) presbyopic group (in dioptre spheres), (b) oblique astigmats (in dioptre cylinders). In graph (b) the point labelled ≥1.25 represents overlying points signifying ≥1.25, ≥1.50 and ≥1.75. 92
The effect of low refractive corrections on rate of reading
judgement of the optometrist. In a previous study looking at current prescribing habits we found that most practitioners would not correct less than 2D of hypermetropia unless there were symptoms (O’Leary and Evans 2003), and the WRRT results support this approach. Clearly, age is relevant to any discussion of hypermetropia and young children may require correction of lower degrees of hypermetropia if clinical tests or history raise concerns over the risk of esotropia. Adults approaching presbyopia are more likely to report symptoms associated with hypermetropia.
Accommodation The amplitude of accommodation is one of the clinical tests that may help determine whether a low hypermetropic correction would be beneficial. Patients with low amounts of hypermetropia rarely notice a deficit in VA (Brookman 1996), and this is particularly true of younger patients, where the large amount of accommodation available enables the image to be clearly focused on to the retina. The hypermetropic subjects in our study had a mean age of 14 years and mean amplitude of accommodation of 11D. Millodot and Millodot (1989) found that the maximum accommodation used when reading is between 50% and 75% of the total amplitude, and therefore it is likely that the subjects in our study had an adequate accommodative reserve to overcome +2.00DS of hypermetropia comfortably. It is, we believe, likely that patients with less than the expected amount of accommodation for their age would have a greater improvement on the WRRT than those with adequate accommodation. One possibility would have been to subdivide our hypermetropic group further according to age, amount of accommodative lag or accommodation reserve (amplitude of accommodation – uncorrected hypermetropia), but we felt that the sample size (n = 32) and young mean age (14 years) of our present sample meant that this type of analysis was unlikely to be productive. This is an area for further research.
Presbyopia Previous research has shown that non-tolerance to optical prescriptions is most commonly the result of an incorrectly prescribed reading addition, often one that is too strong (Freeman and Evans 2010; Hanlon et al. 1987). Hanlon and colleagues evaluated which presbyopic reading additions were most likely to be successful for each age group, and suggested a first reading addition for patients aged 40–45 of +1.00DS (Hanlon et al. 1987). Hofstetter’s review of prescribing for presbyopia suggests that practitioners will delay prescribing a reading addition until a prescription of +1.00DS is reached (Hofstetter 1949), and O’Leary and Evans (2003) found that practitioners would prescribe a reading addition of +0.75DS if symptoms were present, and of +1.50DS in the absence of symptoms. The results of our study agree with Hanlon’s advice, as we found a significant improvement in the WRRT with a reading addition of +1.00 DS to +1.25 DS, but not with lower powers. However the mean age of the subjects in this group was a little higher than Hanlon’s, at 46 years of age (range 39–58 years). The ROC curve (Figure 3a) shows that there is no criterion that will adequately identify those subjects that are likely to have an improvement in their rate of reading from
those that do not, and the results may be more useful if considered in combination with a patient’s symptoms. For example, if the patient attends an eye examination with symptoms that may, equivocally, be related to reading it may be more important to consider the sensitivity of a criterion in identifying an improvement. In this case, if a reading addition of +1.00 or greater is required, the ROC curve indicates that there is a 74% chance that prescribing the addition will improve that patient’s rate of reading (and hopefully reduce symptoms). However if a reading add of +1.00 or greater is found in a patient without symptoms there would be a 49% chance that there would not be an improvement in the rate of reading, and therefore it may be wise to defer prescribing reading glasses until a reading addition of +1.25 or greater was reached when the figure drops to 23%. In conclusion, the final decision as to whether to prescribe a reading addition is likely to be influenced by a number of factors, including symptoms, performance, occupation, working distance, environmental factors (eg lighting, working distance) and patient motivation; however the above guidelines should provide a useful starting point.
Astigmatism The literature provides a wide range of recommendations as to when to correct astigmatism. Previous research has suggested that a ‘normal’ amount of astigmatism in children up to the age of 3 years is between 0.75 DC and 1.00 DC (Atkinson et al. 1980; Fulton et al. 1980) and that only 9% of adults have astigmatism of over 0.75 DC (Fulton et al. 1980). We found in a previous study that most UK practitioners would correct 0.75DC of astigmatism if symptoms were present and 1.50DC in their absence (O'Leary and Evans 2003). Our results showed a marginally significant improvement when astigmatism was corrected in the middle group (–1.00 to –1.25DC), but not in the group with higher astigmatism. The fact that significance was lost in the Kruskal-Wallis analysis also indicates that this effect is not robust. Our most interesting result concerning astigmatism is that, even though most of the subjects with oblique astigmatism had a relatively small cylindrical error (39 out of the 43 subjects with oblique axis had astigmatism of 1.00DC or less), correcting this astigmatism resulted in a statistically significant improvement in the mean WRRT. This finding supports a recent study from Japan (Kobashi et al. 2012). The likely cause of this finding is that oblique astigmatic blur has a greater effect on Snellen acuity than horizontal or vertical blur (Rabbetts 1996). As correcting oblique astigmatism resulted in a greater improvement in the WRRT than correcting either of the other types of astigmatism, the axis of astigmatism should be considered before deciding when to prescribe an astigmatic correction. The ROC curve for these subjects (Figure 3b) showed that correcting oblique astigmatism of 1.00DC or over would correctly identify 67% of those subjects who had an improvement of >5% in the WRRT and would correctly identify 69% of those who did not have this 93
C O'Leary et al.
improvement. Combining our results, therefore, indicates that, having established that a patient has oblique astigmatism, correcting it is likely to be beneficial to rate of reading when the power reaches 1.00DC.
(a)
Does the WRRT provide information that differs from visual acuity? Figure 4 shows fairly weak correlations between the improvements in WRRT and VA with refractive correction. The inadequacy of VA for assessing the effect of hypermetropic refractive corrections is highlighted by Figure 4a, in which there is no strong tendency for participants to demonstrate improved VA with hypermetropic corrections. Conversely, the graphs for astigmatism and presbyopia (Figures 4b–d) do indicate improved VA with refractive corrections, which supports clinical experience. Although Figure 4 shows that WRRT and VA are not measuring the same function, it does not indicate which is better at predicting an improvement with refractive correction. This was investigated by calculating Spearman rank correlation coefficients between refractive error, WRRT improvement and improvement in VA for the hypermetropia, presbyopia, astigmatism and oblique astigmatism groups. These correlations were compared with the partial Spearman correlations for refractive error and WRRT improvement whilst controlling for the improvement in VA (Altman 1991). These analyses were carried out for all the conditions, not just those for which the WRRT showed a significant improvement with higher degrees of refractive error. This is because of the possibility that there may have been a stronger relationship between VA improvement and refractive correction than between WRRT improvement and refractive correction. In view of the number of comparisons made, an adjusted P-value of P < 0.015 was taken as the cut-off value for statistical significance (Table 3). In the analyses in Table 3, all refractive errors have been coded as positive values (although the astigmatism was recorded in negative cylinder notation). For the WRRT improvement (WRRT) and the (logMAR) VA improvement, positive values indicate better performance with the prescribed lens than with the control. Therefore, the non-significant correlation between prescribed lens power and WRRT in the hypermetropia group is in an unexpected direction (larger improvement with weaker lenses). In all other conditions the directions of the relationships are as expected. It is clear from Table 3 that the relationship between the strength of the prescribed lens and the VA improvement is stronger than that between the strength of the prescribed lens and the improvement in WRRT. For the one condition (presbyopia) where there was a significant correlation between prescribed lens strength and improvement in WRRT, this relationship lost statistical significance when the improvement in VA was controlled for.
94
(b)
(c)
(d)
Figure 4. Scattergrams for the change in visual acuity (x axis) related to the change in WRRT result ( y axis) measured in words per minute where positive values represent an improvement in the rate of reading. Visual acuity is measured in logMAR units, with an improvement in visual acuity represented by positive values. (a) hypermetropia; (b) presbyopia; (c) astigmatism; (d) oblique astigmatism.
The effect of low refractive corrections on rate of reading
Reassuringly, the analyses in Table 3 indicate that the time-honoured clinical approach of evaluating the effect of refractive errors on VA is generally valuable and should not be replaced by a lengthier assessment with the WRRT. A less reassuring finding is that, although they are statistically significant, none of the correlations between refractive correction and VA is strong (Spearman rho 5% were also those who improved in VA by at least 0.1 logMAR units. We chose 0.1 logMAR units because Atchison and colleagues (2009) found that this was the limit at which blur became noticeable. Atchison et al.’s research studied distance acuity, but we know of no equivalent research for near acuity. This analysis was carried out for the presbyopia group because they showed a significant relationship between refractive error and WRRT improvement and this group had an adequate number of participants (n = 58). In the low presbyopia group nine participants improved at both VA and WRRT, nine at VA only, two at WRRT only, and six improved at neither. In the group with higher degrees of presbyopia there was more concordance between the two tests. Five improved at both, one at VA only, one at WRRT only, and one in neither. In both groups a subject who improved only in the WRRT did so by more than 13%, suggesting that for some participants the WRRT does seem to measure a different aspect of visual performance to VA. However, the high degree of concordance between the two different methods in the subgroup with higher degrees of presbyopia does raise doubts over whether the WRRT adds useful information clinically for investigating refractive errors.
Although the WRRT and VA are different tests, there is clearly considerable overlap in the skills that they measure. Our analyses support the traditional clinical approach of using VA to measure the effect of optometric interventions. Although it is possible that in some cases the WRRT may add useful additional information, our results do not support the widespread use of this test to predict a benefit from the correction of refractive errors. This could be investigated further with longitudinal studies relating the WRRT results to current tests of visual function, and to changes in symptoms and quality of life (including workplace-related indices) following refractive correction. It is interesting to note that, when used with coloured overlays, the WRRT can predict participants who are likely to continue to use an overlay voluntarily for reading over a period of time (Wilkins et al. 1996). It is not known whether a similar finding would occur with refractive errors.
Limitations of the research In order to attract a large number of subjects who were routinely consulting a primary eyecare practice, we chose not to carry out a cycloplegic refraction as part of the research. This was unlikely to be of any consequence for the presbyopic and astigmatic groups, whose mean age was 46 years. However, for the hypermetropic group, whose mean age was 14 years (range 6–35 years), it is possible that there were some subjects who had a significantly higher degree of hypermetropia than was detected by non-cycloplegic refraction. There are advantages and disadvantages associated with using non-cycloplegic data. As already mentioned, a disadvantage is that latent hypermetropia will not always be detected and in these cases the full degree of hypermetropia is likely to be underestimated. However, an advantage is that, in the age range under investigation, it is likely that most community optometrists would not have carried out a cycloplegic refraction, so our results relate directly to the refractive error that is most likely to have been detected in UK clinical practice. Another advantage is that our research used the presenting refractive error, which is indicative of the degree of hypermetropia for which each subject’s accommodative system was not adequately compensating (see below). Baldwin (1990) compared means for refractive errors (at age 5–6 years) obtained from normative studies using cycloplegic refraction
Table 3. Spearman rank correlation coefficients illustrating the associations between refractive error (Rx), Wilkins Rate of Reading Test (WRRT) improvement and logMAR improvement in visual acuity (VA) for subjects with hypermetropia, presbyopia and astigmatism. P-values are given in parentheses and correlations that are significant at the P < 0.015 level are marked in bold. See text for more details Correlation
Hypermetropia
Presbyopia
Astigmatism
Oblique astigmatism
n
32
58
118
38
Rx and WRRT
–0.260 (0.15)
0.328 (0.013)
0.066 (0.48)
0.211 (0.20)
Rx and logMAR VA
0.378 (0.033)
0.467 (0.000)
0.299 (0.001)
0.445 (0.005)
WRRT and logMAR VA
0.147 (0.42)
0.402 (0.002)
0.194 (0.035)
0.274 (0.096)
Rx and WRRT controlling for VA
–0.345 (0.058)
0.173 (0.20)
0.009 (0.93)
0.103 (0.54)
95
C O'Leary et al.
with a study using non-cycloplegic data and found only 0.50D difference. Accommodation is a confounding variable in any analysis of the effect of hypermetropia in non-presbyopic patients. One possible approach would be to calculate the available amplitude of accommodation, for example by subtracting the degree of hypermetropia from the amplitude of accommodation. We feel that this approach would imply a homogeneity in patients’ adaptation to hypermetropia which is not clinically valid. For example, clinicians encounter some children with high amplitudes of accommodation and no eso-deviation who benefit from correction of low degrees of hypermetropia, and other cases who do not seem to benefit from correction of high degrees of hypermetropia, yet who are otherwise clinically similar. It is possible, although we accept debatable, that by using presenting (non-cycloplegic) hypermetropia we have obtained a measure of hypermetropia that is more likely to reflect the degree of hypermetropia that the patient is having difficulty compensating for than a calculated measure of the theoretically available amplitude of accommodation. It is possible that other measures of accommodative function (eg accommodative lag and accommodative facility: Evans 2007) might have produced valuable co-variates for the hypermetropic group, but these variables were not assessed in the present research. Another issue that future research could investigate is the effect of accommodative compensation on vergence in view of recent work on the relationship between fixation disparity and reading speed (Jainta et al. 2010). Accommodation is also likely to be a confounding variable in the astigmatism group. Astigmatism can be coupled with a myopic, hypermetropic or emmetropic refractive error, or a mixture when one principal meridian is myopic and the other hypermetropic (mixed astigmatism). In hypermetropic compound astigmatism, accommodation can be used to bring elements of the text into focus, at least in pre-presbyopes, whereas this is not possible in myopic astigmatism during distance vision. The impact of the type of astigmatism on the rate of reading is therefore likely to vary depending on the available accommodation, and the astigmatic axis relative to the orientation of the letters in the text. However, it has also been suggested that the use of accommodation to overcome an astigmatic error may be the cause of asthenopia, especially in the case of small astigmatic errors (Rabbetts 1998). A multivariate analysis of these factors may be useful for future work. It was noted in the Methods section that the term ‘prescribed lens’ is used in this research to describe the lens that was prescribed for use in the WRRT, based on the refractive findings during the eye examination. In cases where patients were prescribed spectacles, their optometrist followed usual clinical practice and may have modified the refractive findings, for example giving a par tial correction (Howell-Duffy et al. 2010). When designing the study, we chose to use the refractive findings rather than a final spectacle prescription to avoid the confounding variable of individual prescribing philosophies (Freeman and Evans 2010). Also, the practice of modifying refractive findings seems to 96
be aimed at making spectacles more comfortable or easier to adjust to (Howell-Duffy et al. 2011), and the effect of this, if any, on visual performance is unknown. Similarly, we have not differentiated between those patients who were already wearing a refractive correction and those who were not. Analyses with ROC curves are dependent on the definition of the target condition, which was set in the present research as an improvement in WRRT of >5%. Although this is a commonly used criterion in the literature (Evans and Joseph 2002; Kriss and Evans 2005; Ludlow et al. 2006; O’Leary and Evans 2006; Northway 2003; Singleton and Trotter 2005; Wilkins et al. 1996), one paper has suggested that a >10% criterion might be more appropriate (Kriss and Evans 2005). This paper called for more research to investigate which criterion best predicts a sustained benefit from an optometric intervention; further research of this type is not yet available. In the absence of such research, we chose in the present paper to keep with precedent and use the >5% criterion. As a precaution, the ROC curve analyses were repeated using a >10% criterion. In most cases, this had little effect on the ROC curves (O’Leary 2009). In the WRRT the participant is only required to read the text for a maximum of 1 minute at a time. Research to investigate the effect of individually prescribed coloured filters initially indicated that the beneficial effect was only apparent with prolonged reading (Tyrrell et al. 1995). It later emerged that the WRRT was sensitive enough to determine the benefit from filters over a period as short as 1 minute (Wilkins 2002). Wilkins implied that it is the visually intensive nature of the WRRT that allows it to detect a beneficial effect that would otherwise require longer test periods. It is possible that the research with coloured filters is not in this respect analogous to the effect of refractive errors on performance at the WRRT. It may be that 1-minute test sessions are not long enough to demonstrate an improvement in reading efficiency or comfort. We would stress, therefore, that our results only explore one facet of the potential benefit from prescribing for refractive errors. Other factors such as the patient’s responses to clinical tests and the presence of symptoms should also be taken into account (Brookman 1996).
Conclusions • For presbyopes there was a significant improvement in the mean WRRT performance with reading additions of +1.00DS and over, and the improvement in WRRT performance increased as the power of the reading addition also increased. • For astigmats, correcting oblique astigmatism resulted in a significant improvement in the mean WRRT outcome, even for small cylindrical powers. • Correcting with-the-rule astigmatism, against-the-rule astigmatism or any hypermetropic correction within the range tested did not result in a significant improvement in the WRRT result. • From ROC curve analysis, correcting oblique astigmatism of 1.00DC or greater was the best criterion to adopt when using an improvement of at least 5% in the WRRT performance as the desired outcome.
The effect of low refractive corrections on rate of reading
• These results should not be considered in isolation. It is important to take account of symptoms, working environment and other factors. • This work reconfirms the importance of VA in assessing the effect of refractive corrections.
Acknowledgements The authors are grateful to the patients who participated in the study and to colleagues at Cole Martin Tregaskis Optometrists who assisted in the masked testing. The authors were members of EyeNET, the primary care eye research network supported by the London NHS Executive, and the authors are grateful to EyeNET for funding this research. The views expressed in this publication are those of the authors and not necessarily those of the NHS Executive. We are also grateful to Professor Arnold Wilkins for helpful comments on the experimental design and analyses of the data and to Professor David Crabb and Dr Richard Armstrong for statistical advice. The authors do not have any financial interest in the Wilkins Rate of Reading Test or in any other materials described in this paper.
Summary Our aim was to investigate whether the correction of borderline refractive errors improves reading performance as assessed with the Wilkins Rate of Reading Test (WRRT). A total of 208 subjects with ‘borderline’ refractive errors had their reading performance assessed with the WRRT both with the prescribed lens in place and with a control lens using a double-masked randomised placebo-controlled design. There were 32 pre-presbyopic subjects with hypermetropia, 58 subjects with presbyopia and 118 subjects with astigmatism. Prescribing a reading addition of +1.00DS or more and correcting oblique low cylindrical errors were likely to result in an improvement in reading performance. Correcting with-the-rule astigmatism, against-the-rule astigmatism or any hypermetropic refractive error up to +1.75DS did not result in a significant improvement in mean WRRT results. The relationship between the magnitude of refractive error and the improvement that correction of the refractive error brings about in VA is stronger than the relationship between refractive error and improvement in rate of reading. Our results therefore support the traditional method of assessing the likely impact of refractive corrections via their effect on VA.
References Abrahamsson M, Fabian G, Sjostrand J (1988) Changes in astigmatism between the ages of 1 and 4 years: a longitudinal study. Br J Ophthalmol 72, 145–9 Allen PM, Evans BJW, Wilkins AJ (2012) The uses of colour in optometric practice to ameliorate symptoms of visual stress. Optom Pract 13, 1–8 Altman DG (1991) Practical Statistics for Medical Research. London: Chapman & Hall Anker S, Atkinson J, Braddick O et al. (2004) Non-cycloplegic refractive screening can identify infants whose visual outcome at 4 years is improved by spectacle correction. Strabismus 12, 227–45 Atchison DA, Guo H, Charman WN et al. (2009) Blur limits for defocus, astigmatism and trefoil. Vis Res 49, 2393–403 Atkinson J, Braddick O, French J (1980) Infant astigmatism: its disappearance with age. Vis Res 20, 891–3 Bailey IL, Lovie JE (1976) New design principles for visual acuity letter charts. Am J Optom Physiol Optics 53, 740–5 Baldwin WR (1990) Refractive status of infants and children. In: Rosenbloom AA, Morgan MW (eds) Pediatric Optometry. Philadelphia: J.B. Lippincott, pp. 104–54 Ball GV (1982a) Symptoms in Eye Examination. London: Butterworth Ball GV (1982b) Headache as a presenting symptom. In: Symptoms in Eye Examination. London: Butterworth, pp. 78–85 Brookman KE (1996) Low ametropias. In: Refractive Management of Ametropia. Boston: Butterworth-Heinemann, pp. 123–43 Cotter SA (2007) Management of childhood hyperopia: a pediatric optometrist’s perspective. Optom Vis Sci 84, 103–9 Elliott DB (1997) Clinical Procedures in Primary Eye Care. Oxford: Butterworth-Heinemann Evans BJW (2007) Pickwell’s Binocular Vision Anomalies. Oxford: Elsevier Evans BJW, Joseph F (2002) The effect of coloured filters on the rate of reading in an adult student population. Ophthal Physiol Opt 22, 535–45 Evans BJW, Wilkins AJ (2000) A new near vision test card. Optom Today 40, 38–40 Farbrother JE (2008) Spectacle prescribing in childhood: a survey of hospital optometrists. Br J Ophthalmol 92, 392–5 Firth AY, Machin J, Watkins CL (2007) Tilt and reading speed. J AAPOS 11, 52–4 Freeman C, Evans BJW (2010) Investigation of the causes of non-tolerance to optometric prescriptions for spectacles. Ophthal Physiol Opt 30, 1–11 Fulton AB, Dobson V, Salem D et al. (1980) Cycloplegic refractions in infants and young children. Am J Ophthalmol 90, 239–47 Greenspan SB (1970) Effects of Children's Nearpoint Lenses: Changes in Body Posture and Visually-centered Performance. MSc thesis. Indiana University 97
C O'Leary et al.
Hanlon SD, Nakabayashi J, Shigezawa G (1987) A critical view of presbyopic add determination. J Am Optom Assoc 58, 468–72
Pierce JR, Karlin JE (1957) Reading rates and the information rate of a human channel. Bell Systems j 36, 497–516
Hofstetter HW (1949) A survey of practices in prescribing presbyopic adds. Am J Optom Arch Am Acad Optom 26, 144–60
Press LJ (1985) Physiological effects of plus lens application. Am J Optom Physiol Opt 62, 392–7
Howell-Duffy C, Umar G, Ruparelia N et al. (2010) What adjustments, if any, do UK optometrists make to the subjective refraction result prior to prescribing? Ophthal Physiol Opt. 30, 225–39
Rabbetts RB (1996) Matters arising: scalar representation of astigmatism. Ophthal Physiol Opt 16, 257–60
Howell-Duffy C, Scally AJ, Elliott DB (2011) Spectacle prescribing II: practitioner experience is linked to the likelihood of suggesting a partial prescription. Ophthal Physiol Opt 31, 155–67
Reiter C, Leising D, Madsen EM (2007) Survey of German clinical prescribing philosophies for hyperopia. Optom Vis Sci 84, 131–6
Hrynchak P (2006) Prescribing spectacles: reasons for failure of spectacle lens acceptance. Ophthal Physiol Opt 26, 111–15
Roch-Levecq AC, Brody BL, Thomas RG et al. (2008) Ametropia, preschoolers’ cognitive abilities, and effects of spectacle correction. Arch Ophthalmol 126, 252–8
Ingram RM, Walker C, Wilson JM et al. (1985) A first attempt to prevent amblyopia and squint by spectacle correction of abnormal refractions from age 1 year. Br J Ophthalmol 69, 851–3 Jainta S, Jaschinski W, Wilkins AJ (2010) Periodic letter strokes within a word affect fixation disparity during reading. J Vis 10, 2 Jeanes R, Busby A, Martin J et al. (1997) Prolonged use of coloured overlays for classroom reading. Br J Psychol 88, 531–48 Kobashi H, Kamiya K, Shimizu K et al. (2012) Effect of axis orientation on visual performance in astigmatic eyes. J Cataract Refract Surg 38, 1352–9 Kriss I, Evans BJW (2005) The relationship between dyslexia and Meares-Irlen syndrome. J Res Reading 28, 350–64 Leat SJ (2011) To prescribe or not to prescribe? Guidelines for spectacle prescribing in infants and children. Clin Exp Optom 94, 514–27 Ludlow AK, Wilkins AJ, Heaton P (2006) The effect of coloured overlays on reading ability in children with autism. J Autism Dev Disord 36, 507–16
Rosner J, Gruber J (1985) Differences in the perceptual skills development of young myopes and hyperopes. Am J Optom Vis Sci 62, 501–4 Rosner J, Rosner J (1986) Some observations of the relationship between the visual perceptual skills development of young hyperopes and age of first lens correction. Clin Exp Optom 69, 166–8 Shea SJ, Gaccon L (2006) In the absence of strabismus what constitutes a visual deficit in children? Br J Ophthalmol 90, 40–3 Singleton C, Trotter S (2005) Visual stress in adults with and without dyslexia. J Res Reading 28, 365–78 Thibos LN, Wheeler W, Horner D (1997) Power vectors: an application of Fourier analysis to the description and statistical analysis of refractive error. Optom Vis Sci 74, 367–75 Tyrrell R, Holland K, Dennis D et al. (1995) Coloured overlays, visual discomfort, visual search and classroom reading. J Res Reading 18, 10–23
Lyons SA, Jones LA, Walline JJ et al. (2004) A survey of clinical prescribing philosophies for hyperopia. Optom Vis Sci 81, 233–7
Viner C (2004) Refractive examination. In: Harvey W, Gilmartin B (eds) Paediatric Optometry. London: Butterworth-Heinemann, pp. 21–6
McGraw PV, Winn B (1994) Glasgow acuity cards – a new test for children’s acuity. Optician 207, 14–16
Wilkins AJ (2002) Coloured overlays and their effects on reading speed: a review. Ophthal Physiol Opt 22, 448–54
Millodot M, Millodot S (1989) Presbyopia correction and the accommodation in reserve. Ophthal Physiol Opt 9, 126–32
Wilkins AJ, Jeanes RJ, Pumfrey PD et al. (1996) Rate of Reading Test: its reliability, and its validity in the assessment of the effects of coloured overlays. Ophthal Physiol Opt 16, 491–7
Northway N (2003) Predicting the continued use of overlays in school children – a comparison of the Developmental Eye Movement test and the Rate of Reading test. Ophthal Physiol Opt 23, 457–64
Wills J, Gillett R, Eastwell E et al. (2012) Effect of simulated astigmatic refractive error on reading performance in the young. Optom Vis Sci 89, 271–6
O’Donoghue L, Rudnicka AR, McClelland JF et al. (2012) Visual acuity measures do not reliably detect childhood refractive error – an epidemiological study. PLoS ONE 7, e34441
Wolffsohn JS, Cochrane AL (2000) The practical near acuity chart (PNAC) and prediction of visual ability at near. Ophthal Physiol Opt 20, 90–7
O’Leary CI (2009) The Correction of Borderline Refractive and Heterophoric Anomalies. PhD thesis. City University O’Leary CI, Evans BJW (2003) Criteria for prescribing optometric interventions: literature review and practitioner survey. Ophthal Physiol Opt 23, 429–39 O’Leary CI, Evans BJW (2006) Double-masked randomised placebo-controlled trial of the effect of prismatic corrections on rate of reading and the relationship with symptoms. Ophthal Physiol Opt. 26, 555–65 98
Rabbetts RB (1998) Bennett & Rabbetts’ Clinical Visual Optics. Oxford: Butterworth
CET multiple choice questions This article has been approved for one non-interactive point under the GOC’s Enhanced CET Scheme. The reference and relevant competencies are stated at the head of the article. To gain your point visit the College’s website www.college-optometrists.org/oip and complete the multiple choice questions online. The deadline for completion is 31 October 2015.
The effect of low refractive corrections on rate of reading
1. Which one of the following statements regarding near vision charts is not true? (a) With most near charts the words presented can often be guessed from the context of the passage (b) The lowest line on most near charts is usually smaller than the limit of near visual acuity (c) Near point charts are typically used by optometrists in community optometric practice (d) The eyes are relatively steady when reading print on the typical near point charts used by community optometrists 2. Which one of the following statements regarding the Wilkins Rate of Reading Test (WRRT) is true? (a) Performance on the WRRT is strongly dependent on the level of reading skill of the participant (b) Words commonly used in the English language are avoided in the WRRT (c) The text is printed in Times font at N10 size (d) The test is set single-spaced with four-point spacing between words 3. Which one of the following statements regarding the WRRT is not true? (a) The WRRT was developed to assess the effect of coloured filters on reading in children with reading difficulties (b) The WRRT has been successfully used to demonstrate the effects of coloured overlays on the reading ability of children with autism (c) The most commonly used cut-off criterion for a clinically significant improvement in reading speed using the WRRT is greater than 10% (d) The WRRT has been successfully used to assess the effect of head tilt on reading speed in normal readers
5. When plotting an ROC curve which of the following statements correctly describes the x and y axes of the graph? (a) Sensitivity is x-axis and 1-specificity is the y-axis (b) 1-specificity is the x-axis and sensitivity is the y-axis (c) Specificity is the x-axis and 1-sensitivity is the y-axis (d) 1-sensitivity is the x-axis and specificity is the y-axis 6. What magnitude of hypermetropia was considered by Ingram et al. to be abnormal in young children? (a) +0.50D (b) +1.00D (c) +1.50D (d) +2.00D
CPD Exercise After reading this article can you identify areas in which your knowledge of the effect of low refractive corrections on rate of reading has been enhanced? How do you feel you can use this knowledge to offer better patient advice? Are there any areas you still feel you need to study and how might you do this? Which areas outlined in this article would you benefit from reading in more depth, and why?
Reflection 1. What impact has your learning had, or might it have, on: • your patients or other service users (eg those who refer patients to you, members of staff whom you supervise)?
4. Which one of the following statements is true? (a) A size lens has no refractive or prismatic power but does provide magnification (b) Horizontal blur caused by uncorrected astigmatism has a greater effect on acuity than the same amount of uncorrected oblique astigmatism (c) Vertical blur caused by uncorrected astigmatism has a greater effect on acuity than the same amount of uncorrected oblique astigmatism (d) When a patient has oblique astigmatism, correcting it is only likely to be beneficial when the power reaches 2.00DC
99
C O'Leary et al.
• yourself (improved knowledge, performance, confidence)?
• your colleagues?
2. How might you assess/measure this impact?
To access CPD Information please click on the following link: college-optometrists.org/cpd
100
