Visual Outcome of Preterm Infants Screened in a Tertiary Care Hospital

Abstract

OBJECTIVE:

The aim of this study is to assess the incidence of retinopathy of prematurity (ROP) in preterm infants and to compare the visual outcomes in babies with and without ROP.

MATERIALS AND METHODS:

A consecutive cohort of 812 preterm babies were recruited with gestational age ≤32 weeks and or birth weight ≤1500 g. The outcome was assessed at the end of 15 months by determining fixation behavior, cycloplegic refraction, and vision by Cardiff cards. Incidence of visual outcomes with 95% confidence limits and relative risks were estimated. Chi-squared test and t-test were used as tests of significance.

RESULTS:

The incidence of ROP was 25%. The incidence of myopia, hypermetropia, astigmatism, and strabismus were 15.8% (14.3–17.3), 6% (5.1–7.1), 55.6% (53.6–57.7), and 1.8% (1.4%–2.5%), respectively, in the cohort. The most common refractive error in terms of spherical equivalence was myopia (19.8% in ROP and 14.4% in non-ROP group). The mean visual acuity measured by Cardiff Acuity cards was 0.282 and 0.27 logarithm of the minimum angle of resolution units (right eye) and 0.293 and 0.277 (left eye) in patients with and without ROP, respectively. Strabismus was present in 5% of ROP group and 0.8% of non-ROP group babies. Babies with ROP had six times (risk ratio-6.02; 95% confidence interval 2.8–12.8) higher chance of developing strabismus than those without ROP.

CONCLUSIONS:

Ophthalmological morbidities in premature infants such as refractive errors and strabismus are high in addition to complications like ROP. The incidence of these conditions is more in infants with ROP when compared to non-ROP group.

Introduction

Vision-related problems are one of the major sequelae of prematurity. In middle-income countries, high rates of premature birth and increasing resuscitation of premature infants, often with suboptimal standards of care, have resulted in the third epidemic of retinopathy of prematurity (ROP).[1] The visual compromise from increased incidence of refractive errors, strabismus, and cerebral vision impairment has significant impact on visual functions and development such as psychological and educational aspects.[2]

The primary objective of the study was to assess the incidence of ROP in preterm infants. The secondary objective was to compare the visual outcomes (refractive errors, functional outcomes, and visual acuity) in infants with and without ROP.

Materials and Methods

A hospital-based prospective screening study was done for 24 months during the period April 2013–April 2015. A cohort of 812 consecutive premature babies ≤1.5 kg and or ≤32 weeks of gestational age were included in the study. None of these babies were excluded for any reason since we were interested in knowing the visual outcomes of preterm infants. The eyes of these preterm infants were screened after dilating with 1% tropicamide and 2.5% phenylephrine eye drops applied 10 min apart 1 h before the examination. Classification of ROP was done according to the International Classification of ROP[3] guidelines. These preterm infants were followed up periodically until maturation of retina or regression of the disease.

For those babies with ROP depending on the severity of the disease, observation was done for mild cases and treatment with diode laser (810 nm) photocoagulation was given for severe disease. Interventions and follow-up examinations were done according to the Early Treatment of ROP (ETROP)[4] guidelines. All these babies were followed up until 15 months of age and visual outcomes analyzed in 1596 eyes of 798 babies.

Visual outcomes were assessed by

1. Cycloplegic refraction done at 15 months of age. For this, the pupils were dilated using atropine 1% eye ointment applied twice daily for 3 days prior to refraction. In our study, refractive errors have been classified in terms of spherical equivalence calculated from the retinoscopy readings. A conversion to spherical equivalent (SE) was made, as the algebraic sum of the value of the sphere and half the cylindrical value, for statistical purposes.[5]

   Myopia is classified into three categories, according to the value of the SE, as low myopia 0–−3 diopters, moderate myopia − 3–−6 diopters, and high myopia >−6 diopters. Emmetropia is classified as 0–4 diopters and hypermetropia as >4 diopters.[6] Astigmatism is defined as cylindrical correction of ≥1 D[7] Astigmatism is calculated in plus cylinder form and classified as with-the-rule (WTR) 75°–105°, against-the-rule (ATR) 0°–15° and 165°–180°, and oblique (OBL) 16°–74° and 106°–164°.

2. Presence of strabismus or nystagmus

   Strabismus was detected using the Hirschberg test and cover tests.

3. Visual acuity testing by Cardiff visual acuity cards.[8] Monocular noncycloplegic visual acuity was measured in each eye in sitting position at a test distance of 1 m at 15 months of age. Babies requiring refractive correction were given glasses. It uses a target design of vanishing optotypes and employs the principle of preferential looking.

The targets are drawn with a white band bordered by two black bands, each of half-width of the white band, all on a neutral gray background, thus the average luminance of the target is equal to that of the grey background. In the Cardiff test, each target is positioned either in the top half or in the bottom half of the card. The Cardiff test includes three cards at each acuity level, although only two are usually presented. Beginning with the widest target (lowest acuity), the three cards are shuffled, and the first card is presented at the child's eye level. The calibration for the cards presents acuity levels for the two testing distances, in both equivalent Snellen acuity and logarithm of the minimum angle of resolution (LogMAR).[8]

Visual outcome in terms of ROP was measured as the incidence of ROP with its 95% confidence intervals (CI). All qualitative variables are summarized in percentages and quantitative variables as means with standard deviation. Refractive errors and functional outcome (strabismus) in infants with and without ROP were compared using Chi-squared test, since they were qualitative variables and expressed in %. The comparison of mean visual acuity in the two groups was done using Student's t-test, since it was a quantitative variable. Incidence of refractive errors, strabismus, and visual acuity in infants with ROP and without ROP was also compared using relative risks and its 95% CI.

Ethical clearance was obtained from the Institutional Ethics Committee and informed consent of parents was obtained. The study conforms to the guidelines of the Declaration of Helsinki. Statistical analysis was performed using SPSS version 22 (Statistical Package for the Social Sciences, IBM Corp., New York, NY, USA).

Results

The incidence of ROP in our study is 25% (95% CI 22.49%–27.52%). Among the 812 preterm babies enrolled in the study, the outcome was assessed in 1596 eyes of 798 babies, of which 199 babies had ROP. Fourteen cases (1.7%) were not included in the follow-up analysis, as eight of them lost their lives, of which four were with ROP. Six babies were lost to follow-up.

In our study, the severity of ROP shows, 209 eyes in Stage 1 ROP, 75 eyes in Stage 2 disease, 98 eyes in Stage 3 disease, and 16 eyes had aggressive posterior ROP (APROP). Laser photocoagulation was done in 56 babies with severe as well as Type 1 prethreshold ROP which accounted for 27.5%. More than one sitting of laser was required in five cases. Additional intravitreal antivascular endothelial growth factor therapy was given in one patient with severe APROP which showed prolonged activity. All babies with ROP had bilateral involvement. All types of ROP regressed at 6 months of follow-up.

The significant refractive error present on follow-up at 15 months of age in cases with and without ROP was myopia (19.8% in ROP cases and 14.4% without ROP). This difference was statistically significant (P = 0.01), and the relative risk was 1.37 (CI 1.08–1.99) [Table 1]. Among myopic babies, the incidence of high myopia was 12%. Hypermetropia was present in 8.5% of ROP and 5.2% of non-ROP patients (P = 0.02). Anisometropia of >1 diopter was present in 26% babies.

Table 1.

Analysis of refractive and functional outcome in the eyes of babies with and without retinopathy of prematurity

Refractive and functional outcome	ROP present eyes (n=398)	Incidence % (95% CI)	ROP absent eyes (n=1198)	Incidence % (95% CI)	χ2; P	RR (95% CI)	Outcome variables in cohort (n=1596)	Incidence % in cohort
Myopia
Yes	79	19.85 (16.72-23.30)	173	14.44 (12.83-16.18)	6.57; 0.01	1.37 (1.08-1.99)	252	15.79 (14.3-17.34)
No	319		1025
Hypermetropia
Yes	34	8.54 (6.45-11.08)	62	5.16 (4.2-6.31)	5.4; 0.02	1.65 (1.10-2.47)	96	6.02 (5.1-7.1)
No	364		1136
Astigmatism
Yes	241	60.55 (56.47-64.52)	647	54.01 (51.63-56.37)	5.19; 0.02	1.12 (1.02-1.23)	888	55.64 (53.6-57.68)
No	157		551
Strabismus
Yes	20	5.03 (3.45-7.09)	10	0.83 (0.48-0.137)	26.2; <0.1	6.02 (2.84-12.75)	30	1.88 (1.38-2.51)
No	378		1188

χ²: Chi-square, P: Probability, RR: Relative risk, CI: Confidence interval, ROP: Retinopathy of prematurity

Astigmatism was present in around 60.6% of eyes with ROP and 54% cases without ROP (P = 0.02). WTR astigmatism was present in 86% cases, ATR in 12%, and OBL astigmatism in 2% infants. The relative risks of all the refractive visual outcomes studied were significantly higher among babies with ROP compared to those without ROP [Table 1].

Regarding functional outcome, visual behavior by fixing and following light was good in all except three babies who had neurological sequelae and visual inattention due to global developmental delay. Functional outcome in terms of visual behavior was studied [Table 1]. Strabismus was present in 5% of ROP group and 0.8% of non-ROP group babies. Strabismus with associated nystagmus was present in one baby. All babies with strabismus had esotropia. Babies with ROP had six times higher chance of developing strabismus than babies without ROP (risk ratio = 6.02 and CI = 2.84–12.75).

Visual acuity was assessed by Cardiff Acuity Test (CAT) which showed good outcome with Snellen equivalent >6/12 in most patients of the cohort [Table 2]. The mean visual acuity measured by Cardiff Acuity cards was 0.282 and 0.27 logMAR units (right eye) and 0.293 and 0.277 logMAR (left eye) in patients with and without ROP, respectively. Summary statistics of visual acuities in logMAR units is shown in Table 3. Comparison of mean visual acuity in logMar between babies with and without ROP, excluding those with visual inattention due to neurological cause (logMAR 1) shows slightly lower visual acuity in babies with ROP which is not significant [Table 4].

Table 2.

Visual acuity in logarithm of the minimum angle of resolution units

Visual acuity	Right eye, n (%)	Left eye, n (%)
0.1	1 (0.13)	1 (0.13)
0.2	234 (29.32)	230 (28.82)
0.3	540 (67.67)	536 (67.17)
0.4	7 (0.88)	15 (1.88)
0.5	6 (0.75)	6 (0.75)
0.6	4 (0.50)	4 (0.50)
1	6 (0.75)	6 (0.75)
Total	798 (100)	798 (100)

Table 3.

Summary statistics of visual acuity in babies with and without retinopathy of prematurity

Statistics of vision in logMAR units	Right eye		Left eye
	ROP present	ROP absent	ROP present	ROP absent
Mean	0.2822	0.2714	0.2925	0.2773
95% CI for mean
Lower bound	0.2758	0.2699	0.2786	0.2709
Upper bound	0.3031	0.2826	0.3063	0.2837
5% trimmed mean	0.2762	0.2714	0.2799	0.2722
Median	0.3000	0.3000	0.3000	0.3000
Variance	0.004	0.003	0.010	0.006
SD	0.06654	0.05245	0.09895	0.07941
Minimum	0.20	0.10	0.20	0.10
Maximum	0.60	0.60	1.00	1.00
Range	0.40	0.50	0.80	0.90
Interquartile range	0.10	0.10	0.10	0.10

SD: Standard deviation, CI: Confidence interval, ROP: Retinopathy of prematurity, logMAR: Logarithm of the minimum angle of resolution

Table 4.

Difference in mean visual acuity between babies with and without retinopathy of prematurity

ROP	n	Mean in logMAR units	SD	SEM	t
Present	197	0.2822	0.06654	0.00474	2.336
Absent	595	0.2714	0.05245	0.00215

SEM: Standard error mean, logMAR: Logarithm of the minimum angle of resolution, SD: Standard deviation, ROP: Retinopathy of prematurity

Discussion

The incidence of ROP in our study is 25%, in concordance with various studies, in which the reported incidence ranges from 19% to 26% among the low-birth-weight babies.[9,10,11]

Myopia was the predominant refractive error in the cohort comprising around 15.8% of cases when assessed at 15 months of age. Low birth weight and prematurity contribute to myopia in infancy. In addition to this, the incidence of myopia rises in infants with ROP as compared to those without ROP of same birth weight. The ETROP[4] study reported that 80% of threshold ROP infants and 65% of treated high-risk prethreshold infants had myopia. Studies show that laser treatment also contributes to myopia in ROP infants.[12] The incidence of myopia in our study is less compared to these studies[2,12] and comparable to study by Katoch et al.[6]

The increased incidence of myopia in ROP is distinct from that of full-term patients and is mostly associated with lens thickness with lesser contribution from corneal steepness and axial length. Eyes of premature infants have shorter axial lengths, shallower anterior chambers, and more highly curved corneas than eyes of full-term infants. These differences become more significant as the severity of ROP increases.[5] Furthermore, it has been shown that spontaneously regressed ROP has lesser myopia compared to treatment groups.[13] The risk factors for myopia are a greater number of clock hours of ROP, greater number of laser spots, and a longer time to regression of ROP similar to our study.[6]

According to study by Davitt et al.,[14] astigmatism was not influenced by zone of acute-phase ROP, presence of plus disease, or retinal residua similar to our study. WTR astigmatism was common in most of the preterms according to several studies.[7,14,15] The incidence of anisometropia more than 1 D SE was less compared to the study by Varughese et al.[15] The incidence of the same in their study was 31%.

Testing visual acuity in infants is difficult because of their short attention span and restlessness.[8] The CAT overcomes the symbolic demands and is a test of choice for young children.[16] There are only few Indian studies involving visual acuity estimation with Cardiff cards.[17] Cardiff acuity cards are enjoyable and child-friendly, though they are not as precise and tend to overestimate vision compared to Teller acuity cards.[18] As with other studies, the assessment of visual acuity with the CAT was seen to underdiagnose reduced acuity caused by refractive errors.[17]

Strabismus was present in thirty patients in the study, of which twenty were in the ROP group. All patients with strabismus had esotropia. The rate of strabismus was lower than that of ETROP[4] study which had 22.9% and also in many other studies which had higher incidence.[19] Infants with ROP show significant variability in the presence versus absence of strabismus in the 1^st year of life; thus, conservative management is recommended initially.

Conclusions

Incidence of ROP is 25% or one-fourth of preterm infants have ROP. There is no statistically significant difference in visual acuity between those with ROP and those without ROP. Vision-related morbidities were higher in babies with ROP compared to those without ROP. The study helps in reaching a normative data of visual outcome in preterm babies with and without ROP, from this part of the country.

Financial support and sponsorship

Kerala State Council for Science, Technology and Environment (KSCSTE), Thiruvananthapuram, Kerala, India.

Conflicts of interest

There are no conflicts of interest.