Visual screening of Swedish children: An ophthalmological evaluation



ABSTRACT. Purpose: This study describes the various ophthalmological conditions detected in the Swedish visual screening program for children.

Methods: The study was longitudinal and retrospective. All children (3126) born in 1982 in three Swedish municipalities have been followed from birth to ten years of age. Visual acuity was examined at the ages of 4, 5.5, 7 and 10 years. Before the age of 4, a gross examination of the eyes was performed.

Results: The prevalence of ametropia in the population was 7.7%, the prevalence of strabismus 3.1%, and the prevalence of organic lesions 0.6%. Seven children (0.2%) were visually handicapped (visual acuity ≤0.3 in the better eye). Refractive errors and microtropias were mainly detected at the age of 4, when the first visual acuity test was performed, while manifest strabismus was in many cases detected before this age. With this screening and subsequent diagnosis and treatment, the prevalence of deep amblyopia (visual acuity ≤0.3) has been reduced from 2% to 0.2%. With treatment, 47% of the amblyopic children achieved a visual acuity better than 0.7.

Conclusion: Visual screening is effective in detecting visual and ocular disorders. Most conditions are discovered before the age of 6. Compared to an unscreened population, the prevalence of amblyopia is greatly reduced.

A comprehensive screening program for eye disease and visual dysfunction in children has been in operation in Sweden for the last 20 years. Up to the age of 4, inspection of the eyes and ocular alignment is performed at least five times at the Child Health Care Centres. At the age of 4 years (in some areas also at 5.5 years) visual acuity is measured at the centres and at school at the ages of 7 and 10 years. The vision screening, which is nationwide, is part of a general examination of the child.

An evaluation of the system with visual acuity screening at 4 years of age has been performed (Kvarnström et al. 1998). All children born in 1982 in three Swedish municipalities were included. The visual acuity screening was characterized by a very high attendance rate, 99.2%. The average sensitivity and specificity of the 4-year screening test in the three municipalities was 89.5% and 98.9% respectively. Thus, this part of the screening was very effective in detecting a decrease in visual acuity in 4-year-old children.

The total screening program provides unique information on visual impairment in a population of 10-year-old children. Diagnoses of the ocular and visual conditions in this population were obtained from the records of the schools and the eye clinics. The aim of this study was to describe and analyse the various ophthalmological conditions in children that were detected in the different phases of the screening tests in the period between 1982 and 1993. Studies of this type of total population are very few (Attebo et al. 1998; Frandsen 1960; Vinding et al. 1991) and no other study of children has been performed in this particular respect.

Snowdon and Stewart-Brown (1997) have produced a report summarizing papers on preschool vision screening and amblyopia at the request of the National Coordinating Centre for Health Technology Assessment (NCCHTA). They were critical to many studies as they were incomplete in many ways. Among other things, they questioned evidence regarding the true effect of amblyopia treatment since it had not been compared to the outcome in a non-treated control group. They also pointed out the negative effects of amblyopia treatment and the impact of orthoptic treatment on family life and the psychological well-being of the child.


The cohort consisted of all children born in 1982, who in 1993 were living in the municipalities of Huddinge (908 children out of a total population of 75537), Lund (1033 children – population 94234) and Linköping (1330 children – population 128610). Out of the total number of 3271, 145 were excluded from the study. The excluded children were mostly immigrants who had arrived in Sweden after the age of 4 years, and in whom previous visual screening had not been performed in Sweden. Other reasons for exclusion were that the child had moved from another municipality and no records from previous tests were available. In a few cases, the child had been tested at the Child Health Care Centre but no data could be found. Consequently, 3126 children were included in the study. Each child has been followed from birth to 10 years. By using the Swedish national registration system, it was possible to obtain health care information about the children who were born elsewhere and had moved to these three municipalities. The total population in Sweden at the time was about 8.5 million.


The children included have all taken part in the general Swedish screening system, which includes the following examinations:

  • • In many municipalities (in this study Huddinge and Linköping) an examination of the ocular media is performed in the maternity ward at the age of 1–3 days by a paediatrician.

  • • A general examination by a physician (paediatrician or general practitioner) of the ocular media and ocular alignment is performed at the Child Health Care Centres on the following occasions: 6–12 weeks, 6 months, 18 months and 36 months. The medical history is recorded and hereditary factors such as strabismus, ametropia and congenital ocular disorders are noted.

  • • Monocular vision is tested at 4 years (and at 5.5 years in Lund and Linköping) with the HVOT-chart at 3 meters by the nurses at the Child Health Care Centres.

  • • At 7 years, the children are tested in school by school nurses with a line E-chart (Oculus) at 5 m in Linköping and Huddinge, and with the HVOT-chart at 3 m in Lund.

  • • At 10 years, school nurses test the child’s vision in school. All three municipalities use Monoyer’s linear letters at 5 m.

The health records in school provided a complete overview of the results of visual screening for each child at the different ages. Other information, such as referral, was also available in the records. For each child referred to a department of paediatric ophthalmology, data were collected from the patient records concerning diagnosis, visual acuity, date of referral, refraction, ocular alignment and other results of the orthoptic and ophthalmological examination.

The criteria for referral at ages below 4 years were: suspected strabismus, uncertain fixation behaviour, ocular disease, or family history of ocular disorders. From 4 years, the visual acuity criteria for referral to an eye clinic were: at 4 years <0.8 (20/25), at 5.5 years <1.0 (20/20), at 7 years <0.8 (20/25) with the E-chart and <1.0 (20/20) with the HVOT-chart, and at 10 years <1.0 (20/20). In the event of failure to meet the requirements in the initial test, the child was re-examined twice before referral. Other criteria were suspected strabismus, uncertain fixation behaviour, ocular disease, or heredity of ocular disorders.

At the department of ophthalmology, an ordinary diagnostic investigation was performed, including retinoscopy in cycloplegia, monocular visual acuity testing, cover test for near and distance, prism-cover test, binocular tests/stereo tests, inspection of ocular movements and examination of the ocular media and fundus.


The age distribution of the 413 children referred between 1982 and 1993 to the eye clinics, 329 from Child Health Care Centres and 84 from schools, is shown in Fig. 1. As expected, there was a peak of referrals at the age of 4, when the visual acuity screening was performed. After the age of 8, the patients are mostly referred to opticians. The number of known referrals to the opticians after the age of 8 years was 158. Since the total number of referrals to the opticians is unknown, these children are not included in this diagram.

Figure 1.

Age at referral to the eye clinic. The false positives represent the children who were found normal after examination at the eye clinic. The graph does not display the children who are referred to the optician. This means, for instance, that children older than 8 years who are developing myopia do not show up in this diagram.

Fig. 1 also shows the portion of false positive referrals at each age, i.e. the children who were found to be normal in the examination at the Eye Clinic. They are not included in the subsequent presentation.

Fig. 2 presents the distribution of ages when refractive errors were detected. Children with similar refractive errors in each eye are labelled as ametropes and those with different refractive errors as anisometropes. Ametropia is defined here as hyperopia ≥2D, myopia ≥1D and astigmatism ≥1D. Anisometropia is defined as a difference in refraction between the eyes of ≥1D spherical or ≥1.5D cylinder. The majority of the ametropes were detected at the age of 4. The ametropes detected earlier were mainly those with a family history of refractive errors.

Figure 2.

Age at detection of ametropia. The graph does not display the children who are referred to the optician. This means, for instance, that children older than 8 years who are developing myopia do not show up in this diagram.

After the age of 4, mostly minor refractive errors were detected, except for 4 children with hyperopia exceeding +4.5 D. None of them had an anisometropia of more than 1.5D.

In 379 children, the cycloplegic refraction had been documented. This number includes children who have been referred to the Eye Clinic and those children who were referred to the optician, where the refraction had been noted in the records. Table 1 presents the refraction of both eyes in all these children.

Table 1.  The refraction of 758 eyes of children who had been referred to the eye clinic or to the optician, where the refraction was known. Thumbnail image of

Table 2 shows the degree of refractive difference between the eyes in the same children. Both spherical and cylindrical differences are included.

Table 2.  Number of children (totally 379) with different degrees of anisometropia. Thumbnail image of

Manifest strabismus was seen in 84 children (Table 3). The children with infantile esotropia and those with a family history were diagnosed in the first year of life. Between the ages of 1 and 4 years, additional cases of strabismus were discovered, but there is a peak at 4 years of age, due to the screening at this age. One third of the strabismus cases detected at 4 years were microtropes. After the age of 4, very few cases of strabismus are detected. The prevalence of manifest strabismus was 2.7%. Only phorias with clear subjective symptoms are included in the table. Ten children had eccentric fixation, 7 with esotropia/microtropia and 3 with exotropia. Eight children with strabismus had nystagmus. In 5 children, the strabismus was combined with an eye disease.

Table 3.  The number of children with strabismus detected at different ages. The prevalence figures for various kinds of strabismus are also shown. Thumbnail image of

Visual handicap, defined in Sweden as a corrected visual acuity of ≤0.3 (20/66) in the better eye, was seen in 7 children or 0.2% (calculated on the total population born in 1982 in the three municipalities). All these children had congenital disease. The causes were (age at discovery in parenthesis): 3 cerebral visual impairment (CVI) (<4 years, 6 years, 10 years) (2 with mental retardation), 1 choroidal coloboma and cataract (4 months), 1 congenital cataract (2.5 weeks) (Down’s syndrome), 1 optic atrophy (4 years, Polish origin) and 1 spherophakia (2 years) (Marfan’s syndrome). According to the WHO definition of visual handicap (visual acuity <0.3) only three children fell into this category, which means a prevalence of 0.092%.

Congenital diseases, with visual acuity >0.3 in the better eye, were found in 7 children (0.6%). Three of them had eyelid lesions (2 ptosis, 1 eyelid coloboma), in one case slightly affecting vision. One of them had bilateral congenital cataract (7.5 months, not dense) (of Czech and Yugoslavian origin), 1 bilateral cone-rod dystrophy (10 years) (of Arab origin), 1 unilateral iris synechiae with minor lens opacity (4 years) and 1 unilateral congenital toxoplasmosis (4 years) (of South American origin).

Another 5 children had acquired unilateral organic lesions: 2 keratitis, 1 vitreous haemorrhage, 1 corneal erosion and 1 perforation.

Out of the 413 referrals from the Child Health Care Centres and schools, at least 44 children complained of headache. In 21 children, dyslexia was reported. In 62 cases, there were hereditary factors for strabismus and in 96 cases for refractive errors. Nystagmus was reported in 17 children, Brown’s syndrome in 3 children, of whom 2 were bilateral. Three children had Down’s syndrome and one child Marfan’s syndrome.

Ten children had some form of mental handicap and 3 of these were visually handicapped.


Amblyopia with a visual acuity of ≤0.7 was detected in 91 children (2.9%) and the causes were: unknown 3%, ametropia 30%, anisometropia 16%, strabismus 21% and strabismus + ametropia/anisometropia 30%. In 16% of the initially amblyopic children, the amblyopia was bilateral with a high spherical or cylindrical refractive error and 72.2% of these children achieved vision better than 0.7 in both eyes.

The effect of screening and treatment of amblyopia is shown in Fig. 3, where the bars represent the number of children in each visual acuity range and age. The visual acuity shown for the 4-year-old children was uncorrected vision, while the acuity at 7 and 10 years represented visual acuity with correction. The number of children below a certain visual acuity level is shown cumulatively at the ages of 4, 7 and 10 years. There were 79 children with amblyopia ≤0.7 at the age of 4 and they are all included in the diagram. Another 12 children probably had an acuity ≤0.7 at the age of 4, but 9 of them were missed (false negative), 2 could not be tested at the age of 4 years and 1 could not be tested at the age of 7 years. Of the 79 children, 37 or 47% achieved vision better than 0.7. The number of children with amblyopia is greatly reduced from 4 to 10 years at all levels. Amblyopia is somewhat more common (not significant in a paired t-test) at 10 years than at 7 years, showing that for some children, visual acuity had actually decreased during that time.

Figure 3.

The effect of screening and treatment on amblyopia. The figure shows 79 amblyopic children, who at the age of 4 had vision ≤0.7. The columns represent the number of children at different visual acuity levels and ages and the diagram is cumulative. Since all children had vision ≤0.7 at the age of 4 this column is not included in the group ≤1.0.

Discussion and Conclusions

This evaluation of the Swedish screening program for ocular disorders and visual acuity screening has comprised three municipalities: a large urban area (Huddinge) and two areas consisting of smaller cities in rural surroundings (Linköping and Lund). The samples are considered representative of a Swedish average population, at least from the more densely populated part of the country.

The prevalence of visual handicap (WHO definition) reported here is 0.092% or 9.2 in 10 000, which is in accordance with the prevalence of 4.8–10.9 in 10 000, that has been reported earlier for the age group 0–17 years in the Nordic countries (Blohmé & Tornqvist 1997a; Riise et al. 1992). There were 3 children with cerebral visual impairment (CVI), constituting 42.9% of the visually handicapped children. They also had a high rate of additional impairments. Blohmé and Tornqvist (1997b) have reported a similar prevalence for CVI. Genetic counselling, optimal care during pregnancy and delivery, and better neonatological care may have reduced the prevalence of these more serious visual disorders. While screening cannot reduce the incidence of these disorders, it can certainly reduce the incidence of visual impairment caused by them. It is therefore very important for these conditions to be discovered as early as possible. This is especially true for those conditions that can be treated surgically, such as congenital cataract, but it also applies to visual rehabilitation of the visually handicapped child.

The study showed that the screening system detected a high percentage of the children with strabismus before the age of 4 years. At the 4-year examination, the microtropes were detected. Microtropia often remains unnoticed by the parents, and this shows the importance of performing a visual acuity test to detect these children. After the age of 4, very few cases of strabismus were detected, which is also a sign of the effectiveness of the screening system at the earlier ages.

The total prevalence of strabismus was 3.1% and manifest strabismus was seen in 2.7% of the children. This is in accordance with other authors, who have reported a prevalence of 3.1%–6.7% (Frandsen 1960; Graham 1974; Kornder et al. 1974; Laatikainen & Erkkilä 1980; Ladenvall 1988; Preslan & Novak 1996; Rantanen & Tommila 1971). Esotropia is almost three times as common as exotropia in this study, which is a higher ratio than has been found in other Nordic countries (Laatikainen & Erkkilä 1980; Rantanen & Tommila 1971).

Reported prevalence of important refractive errors ranges from 7.4% (in children less than 7 years old) (Preslan & Novak 1996) to 22.6% (in children less than 15 years old) (Laatikainen & Erkkilä 1980). The prevalence of ametropia, including strabismus and ametropia, in the age group studied here was 7.7%. It must be remembered, however, that this number does not include children aged between 8–10 years who were referred to the opticians. They have presumably developed myopia and therefore the number of children with myopia is probably underestimated in this study, while the prevalence of hyperopia, astigmatism and anisometropia is more accurate. The major refractive error found here is hyperopia (≥+2D), with a prevalence of approximately 5%. This is substantially lower than has been found by Laatikainen and Erkkilä (1980) who found hyperopia in 19.1% of 7–8 year old children. However, the percentage found in our study may not reflect the true prevalence of hyperopia, since many of the children with mild hyperopia (2–4 D) may not have reduced visual acuity or any other visual problem and therefore go unnoticed.

Refractive errors were mainly detected at the age of 4, when the first visual acuity test was performed. The 5.5-year examination also detected a few children with refractive errors. Out of the children with refractive errors, 73.1% were found before the age of 6 years. After 6 years, very few anisometropes or strabismics with ametropia were discovered.

The prevalence of amblyopia with a visual acuity ≤0.7 (≤ 6/9) without screening or treatment ranges from 2.9% to 3.9% (Attebo et al. 1998; Frandsen 1960; Level 1957; Preslan & Novak 1996; Vinding et al. 1991). We believe the condition is sufficiently frequent to warrant screening. The corresponding prevalence with screening and early treatment has been reported to be from 1.0% to 2.1% (Jensen & Goldschmidt 1986; Kvarnström et al. 1998; Ladenvall 1988). However, the reduction of amblyopia after screening and treatment is most pronounced for the lower visual acuities. For instance, in Denmark, the prevalence of severe amblyopia (visual acuity ≤0.3 or 6/18) has been reduced from 3.2% in 1941 (Knudtzon 1941), to 0.73% in 1983 (Jensen & Goldschmidt 1986) and in Sweden, the corresponding prevalence has been reduced from 2.0% in 1970 (Hauffman 1974) to 0.2% in 1992 (Kvarnström et al. 1998). Both this and our earlier study (Kvarnström et al. 1998) have also shown that screening and treatment is effective in reducing the prevalence of amblyopia, especially the more severe levels. With the aid of additional measures in the Child Health Care system for detecting ocular disorders, the rate of serious amblyopia has been substantially reduced (Hauffman 1974) in Sweden. The prevalence of amblyopia with visual acuity ≤0.5 has been reduced from 2.84% to 0.8% and with visual acuity ≤0.3 from 2.0% to 0.2%. The requirement for a driving license in Sweden 0.5 or better. A person with an amblyopia of 0.3 or worse would become visually handicapped if vision were lost or impaired in the better eye due to ocular disease or trauma. It has been estimated that about 1% of the individuals with this degree of amblyopia eventually become visually handicapped (Jakobsson et al. 1999). This visual handicap can largely be avoided by early screening and treatment of amblyopia.

The number of children with amblyopia at 4 years of age is much lower than would be expected from other studies of untreated amblyopia (Hauffman 1974; Vinding et al. 1991). The probable reason for this is that 55% (Table 3) of the strabismic children were referred and treated before the age of 4. The treatment of amblyopia has consisted of: glasses, occlusion, Atropine and Ryser-filters. In spite of screening and treatment, there were still 5 children with amblyopia ≤0.3. Three of these had eccentric fixation and one had microtropia. Non-compliance may also have been a factor in poor results of treatment, but the issue of compliance was beyond the scope of this study.

Some children improved their visual acuity between 7 and 10 years, but vision actually deteriorated in several children. This is the reason for the prevalence of amblyopia being somewhat higher at 10 years than at 7 years. This may be interpreted as indicating that treatment is concluded too early in many patients and that treatment and examination of these children should often be continued, at least to the age of 10 years.


The valuable assistance of the Senior School Physician, Dr. Marianne Ehinger, Professor Göran Stigmar of Lund, and school nurses in all three municipalities is gratefully acknowledged. The study was financially supported by the Sigvard & Marianne Bernadotte Research Foundation for Children’s Eye Care, the Swedish Medical Research Council (Project No. 12X-734), the Research Committee of Östergötlands Läns Landsting, the Solstickan Foundation and Crown Princess Margareta’s Foundation for the Visually Handicapped (KMA).

Gun Kvarnström
Department of Ophthalmology
University Hospital
SE-581 85 Linköping
Fax: +4613223065

Received on September 28th, 2000.
Accepted on December 30th, 2000.