Introduction

Amblyopia screening and treatment firstly prevents the costs of insufficiently treated amblyopia patients who lose the function of their better eye later in life. Cost-effectiveness evaluations of screening and treatment should also consider amblyopia patients who are insufficiently treated but retain their better eye. They experience only a minor decrease in the quality of life, but the condition occurs much more frequently, in approx. 1% of the population [1]. Rahi recently could not detect a functionally or socially significant decrease in the quality of life of amblyopes as based on gross outcomes of quality of life, like educational achievement, employment, and socioeconomic achievement [2]. Such decrease can, however, be detected at the level of daily functional restrictions in amblyopia patients.

We have therefore designed a questionnaire—analogous to the National Eye Institute Visual Function Questionnaire-25 [3] (VFQ-25) [4]—specific for amblyopia and strabismus in order to assess the decrease in quality of life of such patients, the Amblyopia and Strabismus Questionnaire (A&SQ) [5]. The A&SQ is similar to the VFQ-25 in design, format and methods of development and of scoring. The A&SQ contains twenty-six questions that were formulated within five domains: “distance estimation”, “visual disorientation”, “fear of losing the better eye”, “social contact and cosmetic problems” and “diplopia”. The A&SQ has been translated into English [6]. Content validity and criterion validity have been previously evaluated by comparison with two existing and validated quality of life questionnaires, the VFQ-25 and the Short-Form 12 Health Survey (SF-12) [5]. In that study, 53 healthy controls, 68 outpatients and a historic cohort of 174 patients had filled out the questionnaires. Construct validity of the A&SQ has been demonstrated by factor analysis [7].

The clinical validity of the A&SQ was now ascertained by correlating the A&SQ domains to past and current orthoptic parameters of amblyopia patients who had been treated by occlusion therapy 30–35 years ago. These adult amblyopia patients were recruited using the original patient charts kept by the orthoptist who had treated them at childhood more than 30 years ago. Population-based vision screening was operational at the time, and all children with insufficient visual acuity from the entire region of Waterland had been referred to a single ophthalmologist and orthoptist.

In 2003, we re-examined the amblyopia patients to correlate the A&SQ domains to past and current orthoptic parameters.

We expected binocular vision to be correlated to the domain “distance estimation” as well as to the domain “visual disorientation”, and the angle of strabismus to the domain “social contact and cosmetic problems”. Contrary to this, the current visual acuity of the amblyopic eye was found to be the overall dominating parameter for all five A&SQ domains.

Methods

The study was approved by the Medical Ethical Committee of the Erasmus MC and the Medical Ethical Committee of the Waterland Hospital, and was performed according to the standards of the 1964 Declaration of Helsinki.

Patients

The historic cohort was derived from all 1250 subjects, mostly children, who had consulted the orthoptic outpatient clinic of the Waterland Hospital in Purmerend between 1968 and 1974. The hospital served during this period the entire region of Waterland, a rural area north of Amsterdam, bordered on its eastern side by the IJsselmeer. It contains one town—Purmerend—and eight villages. There was little migration in this area across communities.

The original orthoptic charts of these 1250 patients had been kept since 1968 by the orthoptist (HvK). 471 of these patients were children who had been treated for amblyopia with occlusion therapy (Fig. 1). They were born between 1962 and 1972. Of the 471 occluded children, 203 could be traced and were sent the questionnaires; 2 of the 203 patients had died in the meantime. 268 of the 471 patients treated with occlusion therapy could not be traced. In 133 cases, these patients had no known phone numbers. Ninety-nine patients could not be reached on the given phone number even after repeatedly trying to reach them. For 36 patients, the given phone numbers were not longer in service. Of the traced 203 patients, 174 filled out and returned the SF-12, the VFQ-25, and the A&SQ [5], whereas the questionnaires for four patients were sent to an old address.

Fig. 1
figure 1

Flowchart showing the number of patients (children) that had been occluded, been traced at adulthood, had filled out the questionnaires and had undergone the orthoptic examination in 2003

Full size image

137 patients of the historic cohort who had filled out the questionnaires gave their informed consent to undergo an orthoptic examination. Thirty-seven patients who had filled out the questionnaires (N = 174) did not undergo the examination. Ten could not be reached, nine refused to participate, seven did not meet the appointment for the examination, eight were unable to attend the examination because they were hospitalized, abroad or living too far from Purmerend, and four had omitted their name on the filled-out questionnaires.

Orthoptic examination

The following data from the examination 30 years ago, noted on the orthoptic charts, were used: visual acuity at distance of both eyes at the start and end of the treatment, the degree of binocular vision and angle of strabismus. The criteria that the orthoptist had employed to diagnose amblyopia at the start of the occlusion therapy were, in preverbal children, the inability to hold fixation with one eye and, in verbal children, a difference >1 logMAR line between the visual acuities of both eyes.

In the current orthoptic re-examination, binocular vision was examined with Bagolini glasses at distance (6 m) and near (30 cm) and with tests for stereopsis: Titmus test (fly, animals and circles), Lang II test, and TNO test. Visual acuity was measured (at 6 m) by a projector and optotypes. A Dutch reading test (hereafter abbreviated: DRT) (Medical Workshop, the Netherlands) was performed at near (0.35 m) with the text at five sizes: D = 0.5; D = 0.8; D = 1; D = 1.25; D = 2 with D = 1 lettering equivalent to newspaper print. The current angle of strabismus was measured by the simultaneous and alternating prism covertests during fixation at distance (6 m) and at near (30 cm). Latent nystagmus and dissociated vertical divergence (DVD) were also noted, if present. Retinoscopy was performed without cycloplegia, but in the dark, followed by subjective refraction to obtain best corrected visual acuity. Anisometropia was defined as a difference between the spherical equivalences of both eyes larger than 1 D.

The correlation (r) between the clinical parameters (visual acuity at distance of both eyes, degree of binocular vision and angle of strabismus) and the five A&SQ domains was measured by the Pearson correlation test (two-tailed). These correlations could be determined accordingly because the continuous parameters had a more or less normal distribution. The level of significance was divided into * (P = 0.01–P = 0.05 and ** (P < 0.01). The correlations between the dichotomous clinical parameters (nasal fixation, latent nystagmus and anisometropia) and the A&SQ domains were analyzed by mean score differences; level of significance was determined with the unpaired t-test.

Results

Demographic data

The 1250 subjects, mostly children, consulted the ophthalmologist and/or orthoptist of the hospital either on their own initiative or after referral by a child health care center or the general practitioner. Ophthalmologic care was provided between 1968 and 1975 by the sole ophthalmologist and orthoptist (HvK) practicing in the region.

We determined whether the historic cohort was representative for the entire population in the region at that time. The prevalence of amblyopia is reported to be approx. 3.25% in adults [8]. We expected a lower rate of children occluded in the historic cohort because children could have been treated elsewhere or been lost to follow-up. In order to clarify this, a comparison was made between the 471 children who had been treated by occlusion therapy by HvK and the local birth rates.

Of the 471 occluded children, most had been born in the years 1965, 1966, and 1967 (66, 64, and 68 respectively; 198 children in total (Fig. 2). The number of births in the years 1965, 1966 and 1967 in the region was derived from the registers of births: 1286, 1328 and 1355, 3969 children in total. The 198 children out of these 3969 children that had been occluded resulted in a prevalence of 5.0% occluded children in our historic cohort. The diagnosis of true amblyopia at the start of the occlusion therapy could in hindsight not be confirmed for 7 of the 137 patients. Accordingly, our cohort represented a prevalence of 4.8% occluded children in that region after correction for this fraction.

Fig. 2
figure 2

The number of patients (children) per year of birth of the 471 amblyopic patients occluded between 1968 and 1975 in the Waterland Hospital and the 137 re-examined patients are indicated by gray bars and vertically striated bars, respectively. In addition, time (years) of start and end of occlusion treatment of the 137 re-examined patients are indicated by horizontally striated bars and black bars respectively

Full size image

The group of the 137 patients had the following characteristics: mean age was 35.9 years, >75% of patients had secondary education and higher education and university (Table 1).

Table 1 Demographic data patients
Full size table

Orthoptic data

The mean acuity of the amblyopic eye had improved by the occlusion therapy and remained at approximately the same level during the subsequent 30 years (Table 2).

Table 2 Results of orthoptic examination (acuity expressed in logMAR)
Full size table

The degree of binocular vision of the 137 patients had slightly improved during the 30 years between the end of the occlusion therapy and re-examination, as shown by the distribution of patients on the categories of binocular vision (Fig. 3).

Fig. 3
figure 3

Binocular vision of the 137 patients is indicated during occlusion, 30–35 years ago (black bars), and current re-examination (gray bars). The degree of binocular vision was divided into seven categories based on tests for binocular vision and for stereopsis (from worst to best on abscissa). 1: Bagolini negative; 2: Bagolini positive; 3: Bagolini and Titmus stereotest fly positive; 4: minimal Titmus stereotest circles 200″-140″ positive; 5: minimal Titmus stereotest circles 100″-40″ positive; 6: Lang stereotest or minimal TNO test Plate V positive; 7: TNO test Plate VI or VII. Number of patients is represented on ordinate

Full size image

The most frequent types of the current strabismus of the 137 patients were microstrabismus and infantile esotropia (Table 3). The strabismus types were defined as follows: infantile esotropia as manifest esotropia with onset between birth and 6 months; exotropia as manifest diverging of visual axes; microstrabismus as strabismus with angle of deviation 0.5°–5° with reduced binocular vision; small angle as strabismus with angle of deviation 5°–12° with reduced binocular vision; subnormal binocular vision as orthotropia with reduced binocular vision; intermittent exotropia as occasionally manifest exotropia. In addition to strabismus, several patients had other strabismus-related problems, like latent nystagmus and DVD.

Table 3 Patients per condition
Full size table

Correlations between quality of life and orthoptic parameters

The correlations between the A&SQ domains and continuous past (1968–1975) and current (2002–2003) orthoptic parameters were performed by (r) Pearson correlation test and between domains and dichotomous parameters by mean score difference (delta). The two-tailed levels of significance were ** (P < 0.01) and * (P = 0.01–P = 0.05 (Table 4).

Table 4 Past and current orthoptic parameters correlated to the five A&SQ domains
Full size table

The domain “fear of losing the better eye” was significantly correlated (P < 0.01) to the current acuity at distance of the amblyopic eye and DRT results of the amblyopic eye (Table 4).

The domain “distance estimation” was significantly correlated (P < 0.01) to the current acuity at distance of the amblyopic eye and DRT results of the amblyopic eye, and degree of binocular vision.

The domain “visual disorientation” was significantly correlated (P < 0.01) to the current acuity at distance of the amblyopic eye.

The domain “diplopia” was significantly correlated (P < 0.01) to the current DRT results of the amblyopic eye.

The domain “social contact and cosmetic problems” was significantly correlated (P < 0.01) to the current DRT results of the amblyopic eye, and the degree of binocular vision.

The current acuity at distance of the amblyopic eye correlated at the level of significance P < 0.01 to three of the five A&SQ domains. The DRT results of the amblyopic eye correlated at the level of significance P < 0.01 to four of the five A&SQ domains. The current degree of binocular vision correlated at the level of significance P < 0.01 to two of the five A&SQ domains. None of the other correlations were significant at the levels of P < 0.01 and P = 0.01–P 0.05.

Discussion

It was surprising that the current acuity at distance of the amblyopic eye and the current DRT results of the amblyopic eye proved to be overall dominant (P = 0.01–P 0.05) for all five A&SQ domains: “fear of losing the better eye”, “distance estimation”, “visual disorientation”, “diplopia”, and “social contact and cosmetic problems”. On the one hand, the deficit of an amblyopic eye in itself may have profuse effects on vision. On the other hand, intermediate factors could cause the effect—i.e. dominance of the adult visual acuity of the amblyopic eye on the A&SQ domains—but that these factors could not be detected by either the questions of the A&SQ or the tests used in the orthoptic examinations. Thirdly, development of the brain, especially at higher cortical levels, may affect together visual acuity, distance estimation and visual disorientation.

The correlation of the domain “distance estimation” to both acuity and degree of binocular vision may suggest that both are limiting factors in final reaching to and grasping of objects (prehension) [9]. The domain “visual disorientation” may be functionally analogous to positional uncertainty [10] and thus possibly be influenced by the acuity of the amblyopic eye. There is, to our knowledge, no prior research that might explain the correlation of the domain “social contact and cosmetic problems” with latent nystagmus. It is not known to what extent the latent nystagmus became manifest in daily life in the patients, something that could explain the correlation.

Our historic cohort showed a higher proportion of occluded children (4.8%) than the prevalence of amblyopia, determined as unilateral visual impairment in adulthood, in a population not treated for amblyopia (3.25%) [8]. This discrepancy can be explained by the tendency of orthoptists to overtreat a child’s amblyopia after screening: It is not known beforehand which cases of amblyopia will improve or deteriorate spontaneously between end of treatment and adulthood [11]. Orthoptists will therefore tend to prescribe occlusion therapy even in cases of mild amblyopia.

Considering the overall dominance of the current acuity of the amblyopic eye, our results emphasize the necessity of effective treatment of amblyopia.