High school learners in the Free State aged 13–19 years, corresponding to grades 8 to 12, with no corrective lenses, were eligible to participate. Moreover, a grade two to four anterior chamber angle²⁶ graded using the Van Herrick technique was necessary for participation. Participants had to have no diseases or medications that were contraindicated for the stimulation of the parasympathetic system. Signed consent and assent forms were also a prerequisite for participation. The formula to calculate a single population proportion was employed to determine the sample size of the learners, as cited in other prevalence studies. It is indicated below:^16,27,28 n=(Z1−α)2(P(1−P))d2[Eqn 1]

The formula is composed of Z1 – α (the 95% confidence interval) of 1.96, P (prevalence determined in a previous South African study¹) of 0.206, and d (the margin of error, the absolute precision) of 0.05. In consultation with a statistician and using the multistage random sampling technique, 1008 participants from 64 Q1–Q3 high schools across the four districts were invited to participate in this study. To ensure equitable representation, a proportional stratified random sampling method was used, with Q1 schools contributing 40%, Q2 schools 28% and Q3 schools 32% of the sample. Using the 2021 provincial Department of Education schools list, schools were arranged in alphabetical order per quintile, then subjected to a systematic random selection after the sampling interval (kth term) calculation.^25,29,30 Systematic random sampling from class registers was also undertaken to select an average of three participants from each grade to ensure the representation of all grades.²⁵ Systematic random sampling is considered an enhanced method compared to simple random sampling and is recommended when dealing with large populations with available lists (as in this study) because of its ease, inexpensiveness and convenience.³⁰ A feasibility study was conducted at a conveniently selected secondary school, which subsequently did not form part of the main study. Data were collated by one researcher.

The examination included unaided visual acuity using the Tumbling E Bailey-Lovie logarith of the minimum angle of resolution (LogMAR) chart (recorded in LogMAR notation). Subsequently, the Hirschberg test at 0.4 m and Cover test at both 4 m and 0.4 m for binocular motor fusion assessment were conducted using a penlight touch, a wolf ball, an occluder and a prism bar. The slit lamp was employed to assess the anterior chamber (AC) angle and the health of the anterior segment, while the posterior segment ocular health assessment was performed with a Welch Allyn direct ophthalmoscope.

Two drops of 1% cyclopentolate were administered in each eye. To allow the drops to be fully effective, a waiting period of about 20 min – 25 min was observed. If pupillary reflexes were still observed after this period, a third drop was administered. Cycloplegia was also considered achieved in the absence of pupillary responses.^1,28,32 Once complete cycloplegia was attained, participants underwent autorefraction examination using a PlenOptika handheld unit (model nr: MRE0010), which was factory calibrated. Numerous studies have employed this unit among school-going children, including preschoolers, thus demonstrating reliable and accurate measurements of RE.^35,36,37,38 An average measurement of the participant’s refractive status was displayed on the screen for the researcher to note down after 30 s of measurement for each eye.

The URE diagnoses were made as follows: myopia ≤ −0.50 dioptre sphere (DS), hyperopia ≥ +2.00 DS and astigmatism ≤ −0.75 dioptre cylinder (DC) in at least one eye.^1,39 The diagnoses were further dissected into mild (−0.50 DS to −3.00 DS), moderate (−3.25 DS to −6.00 DS), and high myopia (< −6.00 DS). Furthermore, hyperopia is classified as mild (+2.00 DS to +3.00 DS), moderate (+3.25 DS to +5.75 DS) and high hyperopia (> +6.00).^1,39 Astigmatism was also categorised as mild (−0.75 DS to −2.00 DC), moderate (−2.00 DS to −4.00 DC) and high (< −4.00 DC). Astigmatism meridian classifications are with-the-rule (WTR), against-the-rule (ATR), and oblique astigmatism (OA).⁴⁰ Moreover, WTR is the axis correcting minus cylinder 180° ± 30°, ATR 90° ± 30°, while OA is divided into two meridian ranges, that is, 31° to 59° and 121° to 149°.⁴⁰ Although an association between RE and academic performance has been established,⁴¹ the precise clinical threshold at which RE begins to significantly impact learning remains undefined; hence, all levels of VI were considered.

Also, there were cases of compound RE, that is, compound myopic, hyperopic and mixed astigmatism in either eye; antimetropia and anisometropia between the two eyes – all from the raw data as recorded and picked up by the software. All URE cases, as well as ocular pathology cases, were referred to the nearest public health hospital offering eye care services for further treatment. Once the examination was completed, the participants were provided with the results, and the subsequent steps were briefly explained to them in layman’s terms.

Data were recorded on a recording sheet tailor-made for this study. Data were checked for errors and/or inconsistencies before they were entered into the Statistical Package for Social Sciences (SPSS version 27). In consultation with a statistician, data were analysed using descriptive statistics, distribution tables and proportions. The 95% confidence interval (CI) was calculated using binomial distribution for very low prevalence, and p < 0.05 was considered statistically significant based on the Pearson Chi-square test. Both the 95% CI and p-value were employed to interpret the results. The one-sample Kolmogorov–Smirnov test was used to assess the age distribution of the participants. The distribution of presenting VA categories was tabulated by gender, age, grades, quintiles, as well as districts. All variables, that is, gender, age, race, quintile and grade, were categorical and were summarised using frequency distribution tables, Odds ratios and proportions.

Ethical clearance to conduct this study was obtained from the University of KwaZulu-Natal Biomedical Research and Ethics Committee (No. BREC/00005522/2023).

The right eye (72.9% n = 633, 95% CI: 12.02–12.39) had 0.7% (n = 6) fewer cases of unaided visual acuity (VA_sc) ≤ 0.2 LogMAR than the left eye (73.6%, n = 639, 95% CI: 12.07–12.44). However, the difference between the two eyes was not statistically significant (p = 0.805), as both right and left eyes recorded an average VA_sc of 0.2 LogMAR. Furthermore, the odds ratio of VA_sc ranged from 0.517 (sex) to 0.908 (age), and all variables were associated with VA_sc ≤ 0.2 LogMAR as they excluded 1. Also, the logistic regression for all the variables was statistically significant (p < 0.05). Table 2 presents the results of the right eye.

The distribution of URE is tabulated in Table 3. A total of 27.1% (n = 233, 95% CI: 24.2–30.1) presented with URE, which was most prevalent among female participants (71.2%, n = 166), 16- to 19-year-olds (72.5%, n = 169) and Grade 10 participants (33.9%, n = 79). No statistical significance (p = 0.17) was observed between URE and age, but gender and grade were statistically significant (p = 0.045) each.

The logistic regression of URE prevalence increased by a factor of 1.099 (95% CI: 1.003–1.205) as age increased by one unit. Also, URE increased by a factor of 1.396 (95% CI: 1.007–1.937) for female participants compared with male participants. According to the odds ratio, only the age (older) and sex (female) variables were associated with an increase in URE and were statistically significant (p < 0.05). This is because the 95% confidence interval (95% CI) for both variables excluded one.

Astigmatism was the most prevalent type of URE, accounting for 22.3% (n = 192, 95% CI: 19.6 to 25.2) cases, with female participants being twice as many (2.14 times), albeit not a significant difference (p = 0.446). Moreover, astigmatism was most prevalent among the seniors (71.4%, n = 137) and Grade 10 (34.9%, n = 67) participants. With-The-Rule astigmatism was the most prevalent in both the right (55.5%, n = 91) and left eyes (49.4%, n = 86), while oblique astigmatism was the least prevalent in the right (15.2%, n = 25) and left eyes (20.7%, n = 36). In addition, Against-The-Rule accounted for 29.3% (n = 48) in the right eye and 29.9% (n = 52) in the left eye.

Myopia ranked as the second most prevalent URE at 15.1% (n = 130, 95% CI: 12.9–17.6). Male participants with myopia accounted for 23.9% (n = 31). Similarly, myopia was most prevalent among the senior participants, 74.6% (n = 97). Also, 6 in 10 learners (60.0%, n = 78) from the Thabo Mofutsanyane district presented with myopia, compared with 7.7% (n = 10) from the Xhariep district. Similar to the URE logistic regression, myopia was associated with an increase in age, grade and the female sex, all of which were statistically significant (p < 0.05). Hyperopia was the least prevalent URE (5.4%, n = 46) (95% CI: 40–70), with 7 in 10 female participants (71.71%, n = 46) being hyperopic. About two-thirds of hyperopia cases (63.0%, n = 29) were recorded among the senior learners. Moreover, hyperopia was more prevalent among Grade 9 to Grade 11 participants (80.4%, n = 37). According to the odds ratio, hyperopia and astigmatism were not associated with any of the variables, as the 95% CI included one, and the odds ratios were not statistically significant (p > 0.05) for all variables. Importantly, URE or any of the types of ametropia were neither associated with nor statistically significant (p > 0.05) for the economic classification of the school (Q1–Q3). Table 4 demonstrates the classification of URE.

Heterotropia presented in 2.3% (n = 19) and 2.4% (n = 20) participants at 4 m and 0.4 m, respectively. Esotropia (1.1%, n = 11) was the most prevalent misalignment at 4 m followed by exotropia (0.8%, n = 7) and hypotropia associated with exotropia (0.4%, n = 2). Similarly, heterotropia at 0.4 m was esotropia (1.1%, n = 11), exotropia (0.9%, n = 8) and hypotropia associated with exotropia (0.4%, n = 2). Statistical significance was observed among those diagnosed with URE and had either latent or manifest deviation at 4 m (p = 0.037) and 0.4 m (p = 0.001). The only latent deviation detected at 4 m was exophoria (0.5%, n = 4) and at 0.4 m was exophoria (36.2%, n = 315) as well as esophoria (0.3%, n = 3).

Only two male cases of trauma had pupils responsive only to light and accommodation. Also, one female participant presented with a corneal scar, and 1.9% (n = 16) and 1.2% (n = 10) had varying forms of corneal opacity in the right and left eyes, respectively. Vernal keratoconjunctivitis (VKC) cases stood at 3.4% (n = 29). A total of 0.4% (n = 3) presented with glaucomatous signs, and 0.16% (n = 1) had optic atrophy in the left eye.

Stringent terms set by the local authority, compounded by the lack of data collectors, resulted in the retrospective exclusion of the Fezile Dabi district. Notwithstanding the aforementioned, this was a comprehensive study that was representative of the population. Various factors, including no parental consent, withdrawal of assent, and absenteeism, influenced the participation rate. Moreover, Grade 12 learners had to be excluded midway through the study because of examinations that had commenced, resulting in a participation rate of 9.86%, which skewed the results. The study did not measure VI associated URE, as well as the effects of URE on classroom performance.

A similar study among the same cohort as in this study ought to be undertaken in the Fezile Dabi district because it did not feature in this study. Data collection in similar large-scale research would best be conducted by more than one researcher. Furthermore, studies of this nature need to be conducted throughout the country, given the adverse, life-long effects of URE on children, which have deleterious effects on their education. Barriers to the uptake of spectacles in this population ought to be studied in an attempt to comprehend this prevalence. Future research should aim to integrate objective clinical measures of visual function with academic performance indicators to more precisely evaluate the functional impact of UREs on educational outcomes. This would be crucial, as similar studies were mostly assessed among primary school-going children.⁴¹ Such an approach would facilitate the identification of clinically significant thresholds of RE that impair learning, thereby informing the development of more targeted and effective interventions within school settings.