MATERIALS AND METHODS

A detailed description of the methods used in this study has already been documented in earlier work.⁶ To summarise, all patients (new and existing) receiving a minimum of one hour's occlusion for all types of unilateral amblyopia were eligible for recruitment, following informed parental consent. Patients meeting the inclusion criteria were recruited over a 3–4 month period. On inclusion to the study patients were randomised into a leaflet group (who received educational information in the form of a leaflet) or a control group (who did not receive written educational material). The method chosen for randomisation was the tossing of a coin as this is a simple method, easily implemented by the clinicians, and requires the least instructions to follow so as to not unduly increase their workload. The dependent variable in this study therefore was concordance with occlusion therapy, the independent variable was parental understanding and knowledge of amblyopia, occlusion therapy, critical period, and prognosis. The independent variable was manipulated by educational material in the form of a leaflet. Patients were paired, being matched for age (allowing up to 2 years difference) and the length of prescribed occlusion time (allowing up to 1 hour difference). While it is disputed whether the age of the patient and severity of amblyopia are related to concordance, pairing of these variables was necessary to eliminate any possible effect on the concordance. In order to accurately match the children in both groups a quasi-randomisation technique was used. Children were randomly assigned to one of the groups and their length of prescribed occlusion and age was noted. The process continued and the children were paired in line with the criteria stated above. As the required number of patients in each group was reached any remaining patients who had not been paired were noted. The next patient who attended having met the matching requirements for that individual was then automatically placed in the other group in order to complete the final pairings.

The leaflet was constructed by printing in landscape on both sides of A4 size card; by folding into three, a six page leaflet was produced. The print chosen was Times New Roman, with a minimum of 12 point throughout, as recommended by the RNIB²⁰ to ensure the document was legible. The front of the leaflet was designed to gain the attention of the reader using the title “IMPORTANT” followed by a brief explanation of the aim of the leaflet. The different types of refractive error were then simply explained together with the way in which they are treated by means of spectacle correction. At this point emphasis was placed on the importance of wearing glasses during the critical period. Strabismus was then explained, indicating possible aetiologies. The treatment of strabismus was also discussed with strong emphasis given to the fact that surgery would align a cosmetically poor strabismus but not improve the vision. Finally, amblyopia was addressed. Reasons for amblyopia occurring were given along with information on what the condition actually entailed. Occlusion therapy was stated as the main method of treatment for amblyopia and again strong emphasis was given to indicate that patching should commence as early as possible, vitally before 7 years of age, and that it does not correct the strabismus. The last section of the leaflet indicated the important role of the parents in ensuring the treatment is undertaken effectively and that the children themselves are unlikely at present to be aware of the importance. The final point stressed the potential negative effects of untreated amblyopia, including the risk of blindness following the loss of the non-amblyopic eye and career restrictions. Space was given in this section where details of the treatment regimen, eye to be occluded, and number of hours could be recorded. This could be used for reference by the parents during the treatment in cases of poor memory. Sentences were kept short and simple language used whenever possible.

Concordance was monitored by means of a parental diary. Parents were instructed to record the length of time the patch was worn (to the nearest half hour) each day for a 4 week period. Parental understanding of amblyopia and occlusion and parents' own reasons for non-concordance were determined by means of a questionnaire (Table 1). Parents could choose the reasons they felt most accurately described why they had failed to occlude or, if they wished, provide their own alternative if necessary.

Content and face validity of the diary, questionnaire, and leaflet were assessed by several orthoptists. No revisions were deemed necessary but it was suggested that a reading ease score (Flesch Reading Ease)²¹ should be calculated using Microsoft Word to ensure that the documents were easily readable. A score of 85 on a 100 point scale was achieved for the questionnaire and a score of 80 for the leaflet. This is well above the recommended minimum score of 70, with a score 80 and above indicating that the text would be understood by a child and 95% of adults between 25 and 65 years.²² The diaries and questionnaires were piloted on parents of 20 children receiving occlusion and the leaflet was piloted on several laypeople. Following this, an extra question (“would occlusion be more effective at 2 years or 6 years?”) was added to the questionnaire to ensure all the elements of the critical period were explored. The laypeople were questioned regarding the material they had read in the leaflet to ensure the desired knowledge had been obtained. No difficulties were encountered regarding ability to understand the content or assimilate the knowledge and thus no changes were made to the leaflet. Analysis of the level of parental knowledge from the pilot study revealed that written information could be expected to increase the proportion of concordant parents by 25% and up to 35% if the effect of the written information on memory was included. A sample size calculation based on the estimated increase in concordance of 25%, with confidence of preventing a type I error at 0.05 and type II error at 0.20, revealed that 55 parents would be required in each group.

Diaries and instructions were issued to the parent at the clinic as soon as they had agreed to join the study and informed consent had been obtained. Those in the leaflet group were also given the educational material aimed to enhance their concordance. After a 4 week period the patients were routinely reviewed at the clinic and at this time the parent was asked to complete the questionnaire. The questionnaires were immediately collected to maximise return rates. The parents were advised that neither the diary nor the questionnaire would be seen by the orthoptist treating their child and, if they wished, the documents could be mailed directly to the author. This would encourage the parents to be open and honest, and reduce any anxiety or feelings that they were in some way being judged. It was not considered necessary to formally mask the orthoptist who collected the questionnaire and diary, as they could not influence the outcome. The number of hours of occlusion wear over the 4 week period was calculated for each patient. Concordance was analysed by means of a concordance index²³ where occlusion was recorded in the diary (hours)/prescribed occlusion (hours). Statistical comparison of means was achieved by using a related t test. The proportion of non-concordant parents was also calculated. In order to achieve this it was necessary to set a threshold of concordance. For the purposes of this and previous research⁶ parents with a concordance index greater than 0.80 were deemed to be concordant. The proportion of parents in the leaflet and control group were analysed using McNemar's test.

RESULTS

A total of 130 patients (65 pairs) were recruited, but eight pairs of patients were lost to follow up. Data on five pairs of patients were lost as a result of diaries being lost by parents and data were lost on three pairs of patients because the diaries were not returned because parents failed to attend appointments. The leaflet and control group were well matched for the variables of prescribed occlusion wear and age. The mean amount of prescribed occlusion for the leaflet group was 3.2 hours (median 3 hours, range 1–10 hours) with a mean age of 4.4 years (median 4 years, range 1–7 years), compared to 3.3 hours (median 3 hours, range 1–10 hours), and 4.6 years (median 5 years, range 2–7 years) for the control group. The age and length of prescribed occlusion of the patients that were excluded were not markedly different from the sample that remained.

Parental knowledge

The questionnaires were analysed for both the leaflet and control group. The first four questions explored parental knowledge of their awareness that their child had amblyopia, the treatment for this condition, and the prognosis (Table 2). Parental knowledge in the key area of the critical period was poor in the control group. To determine the effect of the leaflet in improving knowledge, the proportion of knowledgeable parents (achieving all correct answers) was calculated for both groups. For this analysis parents who gave reasons for non-concordance such as “I will wait until my child is older and more cooperative,” “I will wait until my child understands the reasons for the treatment,” and “the treatment does not work” were included with parents who gave incorrect responses to the questions and were deemed to have poor knowledge. Of the leaflet group, 50 parents (88%) had complete knowledge, compared to 28 (49%) of the control group (p <0.001). The next two questions of the questionnaire assessed the information exchange regarding the treatment regimen, which had been recorded on the leaflet. The proportion of parents giving a correct response to the length of time they have been told to occlude was 100% for both groups. However, only 54 (95%) parents in the control group indicated they were occluding the correct eye compared to 57 (100%) for the leaflet group (p >0.05). The reasons given by parents for their non-concordance were evaluated for both groups (Table 3). Two parents with a concordance index of more than 0.80 gave reasons but were not included given the negligible amount of their non-concordance. Each non-concordant parent provided a reason and no parent gave more than one reason. A variety of reasons were given for the non-concordance but comparison of both groups revealed that the most common reasons in the control group were attributable to poor knowledge, accounting for 68% of the non-concordant parents compared to only 15% of the leaflet group.

Concordance

The concordance indices were analysed to determine the effect of the increased knowledge obtained by those parents who had received the extra educational material in the leaflet group. The mean concordance index for the leaflet group was 0.85 (median 0.88, range 0.4–1.0) compared to 0.71 (median 0.75, range 0.1–1.0) for the control group (p < 0.001). The proportion of non-concordant parents in the leaflet group was 0.23 (95% CI 0.13 to 0.35) compared to 0.54 (95% CI 0.41 to 0.67) for the control group (p <0.005) (Fig 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Distribution of parents in relation to the concordance index (n =57 pairs). Open bars represent the control group (n = 57); solid bars represent the leaflet group (n = 57).

DISCUSSION

The patients recruited to the study were representative of the typical amblyopic population presenting at an eye unit. All types of amblyopia were recruited except for stimulus deprivation as no patients presented with this condition during the study period. This simply reflects the reduced incidence of this type of amblyopia and would not have introduced any form of bias to the findings. This is because the aim of the study was to assess the effect of written information on understanding and concordance and not to assess the effect on the outcome of the treatment. Patients in both groups were well matched for age and length of prescribed occlusion. This was considered necessary to eliminate any effect of these variables between the groups, as the effect of age on concordance has been variable between different studies.^4,7 Given that the majority of patients could be matched following randomisation, it is not felt that the automatic consecutive placement of some children into one of the groups, in order to achieve accurate matching, would have introduced any substantial bias. Subsequent analysis of the control group demonstrated no correlation between length of prescribed occlusion wear and non-concordance, as reported elsewhere,⁶ but at the time of the study period no evidence was available to rule out the effect of this potential bias. The level of visual acuity was not matched as the level of prescribed occlusion generally reflects the level of acuity with greater levels of occlusion being prescribed for more severe amblyopia. Also the introduction of another variable, if not absolutely essential, only serves to add methodological problems, making matching of the subjects more difficult.

Parental knowledge of the critical period and the importance of the age at which treatment commences were poor in the control group. Failure of parents to appreciate these facts means that there would be no sense of urgency to the treatment, which itself could lead to non-concordance. Parents may also actively choose to delay the treatment until their child reaches an age when either they can decide whether or not they wish to have the treatment themselves or reach an age when the rationale for the treatment can be more easily explained and understood. Knowledge in the leaflet group was far superior and the educational material was therefore successful in increasing the level of parental understanding in each area that was examined. In order to analyse the difference in knowledge between the two groups quantitatively it was decided to class the parents as knowledgeable only if all questions were answered correctly (and if none of the parental reasons indicated poor knowledge), rather than creating a scoring system. It was considered that a parent could obtain a high score if most questions were answered correctly despite, for example, being unaware of the prognosis being more favourable at 2 years rather than 6 years. Poor knowledge in just one area may have important consequences for non-concordance and a high score in this case would be misleading. It was deemed more useful to identify the proportion of parents who had adequate knowledge from the remainder of parents with inadequate knowledge in whom poor understanding may have contributed to their non-concordance. Three patients in the control group were occluding the incorrect eye compared to no errors in the leaflet group. While the difference between the groups was not statistically significant, it is of clinical significance. The confusion may have arisen owing to a misunderstanding between the clinician and the parent or perhaps because of poor memory, both of which would be improved by increased knowledge. The leaflet used in this study incorporated a space where details of the regimen could be recorded and used for reference at a later date if required. This would help to reduce this type of error but as the proportion of parents who occluded the incorrect eye was low it is not possible from the sample size used to draw any firm conclusions as to whether the leaflet group had no errors as a result of the educational material or simply by chance.

Concordance was significantly greater in the leaflet group compared to the control group. The increase in knowledge achieved by the introduction of written material was therefore successful in improving parental concordance. Significant differences in concordance were present when both the mean concordance indices and proportion of non-concordant parents were compared between the two groups. Clearly the proportion of non-concordant parents will vary depending on where the threshold of non-concordance is set. The 80% threshold was considered a reasonable level to choose, as parents who begin to fall below this level (averaged over the whole treatment period) will be failing to occlude for a substantial period of time. Adjustment of the level of the threshold for non-concordance, either an increase or decrease, did not alter the finding of a statistically significant difference between the groups. The success of the effect of improving understanding and knowledge to increase concordance is important. The extent of the problem of non-concordance with occlusion therapy and the resulting consequences on the outcome has already been discussed and these issues have been well documented. The present study did not aim to assess the effect of concordance on outcome and a much longer follow up period would be required in order to assess this satisfactorily. Although lack of information on outcome may be considered to weaken the possible conclusions that can be made from this study, it has been consistently reported that non-concordance is one of the main factors responsible for treatment failure. While the difficulties in implementing the treatment have been explored, little research has been undertaken to investigate methods that could increase the level of concordance. This study has demonstrated a practical method by which concordance can be improved and as a direct result of this an improvement in acuity would clearly be expected. It is accepted that the level of parental concordance may vary during the treatment period and that this study only assessed a snapshot of one month of the total course of treatment. However, this is unlikely to have any influence on the differences found between the two groups. It is also accepted that written information cannot be expected to improve concordance in all parents. However, analysis of the parental reasons for non-concordance (Table 3) revealed that 68% of parents in the control group gave reasons indicating poor knowledge. It would appear therefore that a substantial number of parents would benefit from extra educational material to increase their understanding and subsequently their concordance. It could be argued that the leaflets would not in fact be read by many of the parents. Research in other areas though does not show this to be the case. Patients generally want to have more information and it has been shown that between 89% and 95% of patients read information booklets.²⁴ This could be maximised if parents were handed the leaflets by a clinician with a verbal reinforcement of the importance of the contents. In order for the leaflets to be effective attention must also be drawn to the presentation and design and a good readability score obtained. For this study the content of the leaflet was designed with the sole purpose of providing and emphasising only key information that would be likely to improve concordance. The aim was not to provide a general information booklet about eyes and strabismus.

The method used to assess concordance was by means of a parental diary. Discussion of its validity and good comparisons with objective techniques have already been made in earlier work.⁶ It is accepted that this method may lead to some underestimation of the degree of non-concordance as parents may not always fully admit to the extent of their non-concordance. However, steps were taken to minimise this possibility by ensuring confidentiality and anonymity from the orthoptist treating their child. Both groups used the same method to measure the concordance and therefore this area of potential bias would not have influenced the effect of the leaflet on concordance. As objective techniques become more readily available^23,25 it would be useful to conduct further research involving methods of improving non-concordance using both diaries and objective patch monitors for comparison.

In conclusion it would be beneficial to a large proportion of patients if their parents were issued with a simple leaflet aimed at improving knowledge in key areas such as the critical period, which also emphasised the importance of occlusion and possible negative consequences of failing to treat amblyopia, in conjunction with verbal reinforcement. It is necessary for parents to understand the aims and reasons for the treatment, perceive an element of urgency, and be aware of the seriousness of the condition (issues relating to the Health Belief Model variables¹¹) in order for the decision to be made that the positive benefits outweigh the negative difficulties of having to implement the treatment. Unlike behavioural techniques, which are often impractical for use on large numbers of people, written information is simple to produce, easy and inexpensive to implement, yet a highly effective method of increasing concordance.