Changes in the knowledge, attitudes, and practices of school teachers following an educational intervention in Skardu: a quasi-experimental study

Abstract

Objective

To evaluate within-participant changes in knowledge, attitude, and practice (KAP) among school teachers following a structured educational intervention on children’s eye health in Skardu District, Gilgit-Baltistan (GB), Pakistan.

Methods

This single-group pre–post quasi-experimental study was conducted across 132 government schools in Skardu District, GB, Pakistan. A total of 264 full-time school teachers were purposively selected in consultation with district education authorities. A structured two-day paediatric eye health and vision screening training programme combined theory-based instruction with supervised hands-on practice and was delivered by ophthalmologists and optometrists. The primary outcomes were within-participant changes in knowledge (0–10), attitude (5–25) and self-reported practice (5–25) scores. Knowledge was reassessed immediately after training, while attitude and practice were reassessed 2–3 weeks later, after school-based screening activities had begun.

Results

Knowledge showed the largest within-participant change, increasing from a median of 6 to 9, with an estimated paired location shift of 2.5 points (95% CI 2.0–3.0; p < 0.001; standardised Wilcoxon effect size r = 0.66, large). Self-reported practice increased from a median of 22 to 23, with an estimated paired location shift of 2.0 points (95% CI 1.5–2.5; p < 0.001; r = 0.39, moderate), although the absolute change in the median was modest. Attitude scores were high at baseline and showed a statistically significant but small-magnitude change; the median remained 23 (IQR 21–24 to 22–25), with an estimated paired location shift of 1.0 point (95% CI 0.5–1.5; p < 0.001; r = 0.27, small). Item-level knowledge gains were greatest for childhood cataract and glaucoma (+ 50.4% points) and recommended eye examination (+ 33.3% points). No evidence of differential change was detected across the examined teacher subgroups (all p > 0.05).

Conclusion

Following the structured two-day teacher-training programme, teachers demonstrated substantial within-participant gains in knowledge, a moderate improvement in self-reported practice, and a small change in attitude from an already high baseline. Because the study had no control group, used short-term follow-up, and relied on self-reported attitude and practice measures, the findings should not be interpreted as proof of causal effectiveness or sustained behavioural change. The results support further controlled and longer-term evaluation of teacher training as one component of school eye-health programmes in remote settings.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12886-026-05373-9.

Introduction

Childhood visual impairment remains a major but largely preventable public health challenge worldwide, with a disproportionate burden in low- and middle-income countries (LMICs) [1]. The World Health Organization (WHO) estimates that more than 2.2 billion people globally live with vision impairment or blindness, of whom at least one billion have conditions that are either preventable or remain unaddressed [2]. Among school-aged children, uncorrected refractive error is the leading cause of visual impairment and represents one of the most readily detectable and treatable conditions [3, 4]. Globally, approximately 19 million children aged 5–15 years are visually impaired, with an estimated 12 million cases attributable to uncorrected refractive error, a condition that can be easily detected and corrected with effective screening methods and education [5]. Other important causes of childhood visual impairment include amblyopia, corneal opacity, cataract, retinal disorders and congenital ocular conditions, although their relative contribution varies by region, age group and access to eye care services [5].

Vision is fundamental to child development and learning. Untreated visual problems can disrupt reading, writing, classroom participation and academic performance [5]. The global increase in myopia has further increased concern for future service needs. Recent estimates suggest that myopia prevalence among individuals aged 5 to 19 years increased from 24% in 1990–2000 to 36% in 2020–2023, with projections indicating that more than 740 million children and adolescents may be affected by 2050 [6]. Early identification and correction during the school years is therefore critical [3, 4].

Pakistan faces both a high burden of correctable childhood visual problems and unequal access to child eye care services [7–9]. School-based studies from different regions of Pakistan have reported refractive error prevalence ranging from approximately 8% to 30%, reflecting differences in geography, socioeconomic conditions, age groups and screening methods [7, 8, 10]. These challenges are amplified in GB, where scattered communities, difficult terrain, and limited specialist eye care services can delay diagnosis [11]. In Skardu district, high-altitude conditions and dispersed communities make school-based approaches particularly important [12].

Schools provide repeated and organized access to children, and teachers are well placed to observe functional signs of visual difficulty such as squinting, poor attention or difficulty copying from the board [13, 14]. However, teachers often have low-to-moderate knowledge and inconsistent practices related to child eye health, even when their attitudes are generally positive [5, 15, 16]. Without structured training, practical tools and clearer referral pathways, teachers may recognise that eye health is important but remain unable to contribute effectively.

Educational interventions have shown promise in strengthening teachers’ knowledge, confidence and perceived responsibility for children’s eye health in India, Nepal and other LMIC settings [17, 18]. However, there is limited evidence of intervention from remote mountainous regions of Pakistan, where health system constraints differ substantially from those in urban areas. This study therefore evaluated within-participant changes in teachers’ KAP following a structured educational intervention in Skardu District and explored whether patterns of change differed across teacher characteristics. The findings are intended to inform the design and further evaluation of school-based eye-health programmes in similar settings.

Methodology

Study design and setting

This was a single-group pre–post quasi-experimental study without a control group, conducted as part of the “School-Based Vision Screening Programme in Skardu, Pakistan-A Clear and Bright Future” implemented from September to December 2025 by the Department of Ophthalmology and Visual Sciences (DOVS), The Aga Khan University Hospital (AKUH), Karachi, Pakistan, across 132 government schools in Skardu district, GB. The study was designed and reported in accordance with the Transparent Reporting of Evaluations with Nonrandomised Designs (TREND) statement [19]. The study was conducted with formal approval from the Government of Baltistan, School Education Department, which granted permission to access government schools and facilitated coordination with district education authorities.

Participants and sampling

The study population comprised full-time school teachers from participating government schools. A purposive sampling approach was adopted in consultation with school administrations and district education authorities. At least one teacher was nominated from each school; the final number (one to four per school) was adjusted based on school enrollment to ensure adequate screening capacity for subsequent vision-screening activities. This resulted in a final sample of 264 teachers from 132 schools. The participant flow is shown in Fig. 1.

Fig. 1.

Participant flow and timing of pre- and post-intervention assessments

Inclusion and exclusion criteria

Teachers were eligible if they were full-time employees with at least six months of teaching experience and willing to participate in both the training programme and school-based vision-screening activities. Teachers unavailable during scheduled training dates or who declined participation were excluded.

Educational intervention

All participants attended a structured two-day training programme, delivered in 12 batches of 20–25 teachers at various locations across Skardu district. To mitigate geographic access challenges inherent to the mountainous terrain, both in-person sessions and virtual platforms were used to ensure all enrolled teachers could complete the programme. Day 1 comprised didactic sessions covering basic ocular anatomy, common childhood eye disorders, preventive strategies, and an introduction to vision-screening techniques. Day 2 focused on hands-on training, including visual acuity assessment (Snellen/E-charts), near vision testing (reading charts), external eye exam using torchlight and colour vision assessment using the digital pseudoisochromatic plate test within the Eye Handbook application [20]. Training was delivered by optometrists and ophthalmologists from DOVS, AKUH and combined theoretical instruction with supervised practical demonstrations. Each teacher received a PDF guide for future reference (Supplementary File 1).

Competency assessment

At the end of Day 2, each teacher’s practical vision-screening competency was assessed using a standardised vision-screening checklist covering five domains: screening environment and preparation, vision-screening procedure, external eye examination and symptom assessment, determination of screening outcome and referral, and documentation (Supplementary File 2). Each checklist item was rated as “Completed,” “Needs Improvement,” or “Not Performed”. Those who passed were certified to conduct screening under the programme. Teachers were considered competent and eligible for certification only when they successfully completed 100% of the required checklist items. Teachers who did not initially demonstrate competency in one or more items received additional targeted training in the relevant domain(s) and were reassessed. Certification for programme screening was granted only after 100% competency had been demonstrated.

Programme screening and referral pathway

Following completion of training and competency certification, teachers conducted school-based vision screening using the standardised screening protocol taught during the programme. Children who failed teacher-led screening were referred for further assessment during screening validation visits conducted by the DOVS team. Optometrists reassessed referred children, and those requiring further evaluation were referred to ophthalmologists at the Regional Headquarter Hospital (RHQ), Skardu, for comprehensive eye examination and management. Children requiring refractive correction were provided spectacles free of cost with support from the funding partner.

Participant compensation

Each teacher received a one-time honorarium of PKR 3,000 (approximately USD 10) as compensation for training and travel. The honorarium was disbursed after training completion and was not contingent on questionnaire responses or post-intervention performance. Honoraria and programme operational costs were funded by National Sight Restoring Project (NSRP) 2019–2025 implemented by The Fred Hollows Foundation (FHF), funded by Australia’s Department of Foreign Affairs and Trade (DFAT) through its Australian NGO Cooperation Programme (ANCP).

Study instrument

A structured KAP questionnaire was developed following a comprehensive literature review and adapted from guidelines issued by the World Health Organization, International Agency for the Prevention of Blindness, Lions Club International and the School Health Integrated Programme [21–23]. Items were selected and adapted to reflect the objectives of the teacher-training programme and the local school eye-health context in Skardu. The questionnaire was available in both English and Urdu. The complete questionnaire is provided as Supplementary File 3.

The questionnaire comprised four sections: (A) baseline characteristics and professional characteristics; (B) knowledge (10 multiple-choice items on refractive errors, signs and symptoms, visual acuity testing and common childhood ocular conditions); (C) attitudes (a 5-point Likert scale i.e. Strongly Disagree = 1 to Strongly Agree = 5); and (D) self-reported practices (a 5-point Likert scale i.e. Never = 1 to Always = 5).

Knowledge items were scored as correct (1) or incorrect (0) and summed to produce a total knowledge score ranging from 0 to 10. Attitude and practice items were summed separately, producing total scores ranging from 5 to 25 for each domain.

Instrument reliability

The questionnaire was pilot tested among 30 teachers before study implementation. Internal consistency was assessed separately for the three KAP domains using Cronbach’s α, with α ≥ 0.70 considered acceptable. Internal consistency was acceptable for knowledge (α = 0.75), attitude (α = 0.72), and practice (α = 0.70).

Data collection procedures

Written informed consent was obtained from each participant. Trained data collectors administered the pre-intervention questionnaire on Day 1 immediately before training. Knowledge was reassessed at the end of Day 2 using the same 10 items. Attitude and practice subscales were re-administered 2–3 weeks after the training, after teachers had started screening activities. This interval was chosen so that newly acquired knowledge could be reflected in attitudes and classroom practice. The present analysis focuses on teacher-level KAP outcomes. Diagnostic accuracy of teacher-led screening and student-level outcomes, including validation findings, referral completion and spectacle provision, were evaluated separately within the same programme.

Outcome measures

The primary outcomes were within-participant changes in total knowledge, attitude, and self-reported practice scores from pre- to post-intervention. Higher scores represented greater knowledge, more positive attitudes, and more frequent self-reported eye-health practices. Domain-level KAP scores were analysed as the prespecified outcomes. Item-level changes in individual knowledge questions were also examined as secondary outcomes.

Sample size considerations

A formal a priori sample size calculation was not undertaken because the study was embedded within an implementation-focused school vision-screening programme, and the number of participating teachers was determined primarily by operational requirements. Specifically, teacher recruitment was based on the need to ensure adequate screening capacity across the 132 participating government schools, with one to four teachers nominated per school according to school enrolment.

Ethical considerations

The study protocol was reviewed and approved by the Ethical Review Committee in accordance with the Declaration of Helsinki (Ref no. 2025-11851-36671) at AKU, Karachi. Permission was obtained from the Government of Baltistan Education Department. Written informed consent was obtained from all participants and data confidentiality was maintained throughout.

Statistical analysis

Data were analysed using R version 4.3.1. Categorical variables were summarised using frequencies and percentages. Normality of continuous variables was assessed using the Shapiro–Wilk test alongside visual inspection of histograms, Q–Q plots, and boxplots. Age was summarised using both mean ± SD and median (IQR) owing to mild deviation from normality. KAP scores were not normally distributed (all p < 0.001) and are reported as median (IQR).

Item-level changes in knowledge were assessed using McNemar’s test. Overall KAP score changes were assessed using the Wilcoxon signed-rank test. Effect sizes were calculated using r = Z/√N and interpreted as small (r = 0.1), moderate (r = 0.3), and large (r = 0.5). Subgroup analyses compared change KAP scores (post minus pre) using the Mann–Whitney U test for binary variables and the Kruskal-Wallis test for variables with three or more categories. A complete-case approach was used. Missing data were minimal (1–5 participants per outcome; <3%) and did not differ systematically by participant characteristics.

Because multiple teachers were recruited from some schools, a sensitivity analysis was conducted to assess the potential influence of school-level clustering. For each KAP domain, a linear mixed-effects model was fitted to the pre- and post-intervention scores, with time (pre- versus post-intervention) included as a fixed effect and random intercepts used to account for clustering of observations within schools and repeated measurements within teachers. These models were restricted to participants with complete paired observations for the respective KAP domain and were used as sensitivity analyses to determine whether accounting for the hierarchical data structure materially altered the conclusions of the primary paired analyses.

All tests were two-sided with p < 0.05 considered statistically significant.

Results

Baseline characteristics

Table 1 summarises the baseline characteristics of the 264 teachers. The mean age was 40.2 ± 7.0 (median 39.5, IQR: 35–45) years. The sample was approximately equally divided by gender (49.2% male and 50.8% female). Most had completed a bachelor’s degree (57.6%). Most teachers (78.8%) were class teachers and the remaining 21.2% were subject-specific teachers who taught across multiple class sections. Nearly half (48.1%) taught the primary level (Grades 1–5), followed by secondary level (35.2%) (Grades 6–10). More than half (52.7%) taught both science and non-science subjects. Most had more than 10 years (65.5%) of teaching experience. Notably, only 5.7% had received prior eye-health training and 14.8% had previously participated in school-based screening.

Table 1.

Baseline characteristics of school teachers (n = 264)

Characteristic	n (%)
Age (years)
Mean ± SD	40.2 ± 7.0
Median (IQR)	39.5 (35–45)
Gender
Male	130 (49.2)
Female	134 (50.8)
Highest education level
Graduate	152 (57.6)
Postgraduate or above	112 (42.4)
Currently a class teacher
Yes	208 (78.8)
No	56 (21.2)
Grade level taught
Primary (Grades 1–5)	127 (48.1)
Secondary (Grades 6–10)	93 (35.2)
Both	44 (16.7)
Teaching experience
≤ 10 years	91 (34.5)
> 10 years	173 (65.5)
Prior eye-health training*
Yes	15 (5.7)
No	246 (94.3)
Prior participation in school screening**
Yes	39 (14.8)
No	224 (85.2)
Subject category
Science	42 (15.9)
Non-science	83 (31.4)
Both	139 (52.7)

Values are presented as n (%) unless otherwise indicated. Age is summarised using both mean ± SD and median (IQR) due to mild deviation from normality

*Percentages for prior eye health training are based on valid responses (n = 261)

** Percentages for prior participation in school screening are based on valid responses (n = 263)

Item-level knowledge gains

All 10 items demonstrated a statistically significant improvement post-intervention (all p < 0.05; Fig. 2). The largest gains were in knowledge of paediatric cataract and glaucoma (K10) (22.7% to 73.1%; +50.4% points), recommended eye examination frequency (K5) (48.1% to 81.4%; +33.3% points), and Snellen chart interpretation (K7) (54.8% to 82.5%; +27.8% points). Knowledge of appropriate management of suspected visual problems showed a high baseline with smaller improvement, suggesting a pre-existing ceiling effect.

Fig. 2.

Item-level changes in knowledge responses following educational intervention. Dumbbell plot shows the percentage of teachers answering each knowledge item correctly before and after the educational intervention. Blue points represent pre-intervention responses and red points represent post-intervention responses. All knowledge items showed significant improvement based on McNemar’s test (p<0.05)

Change in KAP scores

All three KAP domains showed statistically significant changes following the intervention. (Fig. 3), although the magnitude of change differed across domains. Knowledge showed the largest change, increasing from a median of 6 (IQR 5–8) to 9 (IQR 8–10), with an estimated paired location shift of 2.5 points (95% CI 2.0–3.0; p < 0.001). Practice increased from a median of 22 (IQR 18–23) to 23 (IQR 22–24), with an estimated paired location shift of 2.0 points (95% CI 1.5–2.5; p < 0.001). Attitude scores had the same pre- and post-intervention median of 23, although the IQR shifted from 21-24 to 22–25; the estimated paired location shift was 1.0 point (95% CI 0.5–1.5; p < 0.001).

Fig. 3.

Changes in KAP Scores Following Educational Intervention. Raincloud plots illustrate the distribution of pre- and post-intervention KAP scores for each domain. Only participants with complete paired data are included (Knowledge: n=259; Attitude: n=263; Practice: n=262). Pre-intervention scores are shown in blue and post-intervention scores in red

Sensitivity analysis accounting for school-level clustering

Mixed-effects sensitivity analyses accounting for clustering of teachers within schools produced results consistent with the primary paired analyses. The estimated adjusted pre–post change was 2.05 points for knowledge (95% CI 1.76–2.34; p < 0.001), 0.77 points for attitude (95% CI 0.44–1.11; p < 0.001), and 1.68 points for self-reported practice (95% CI 1.21–2.15; p < 0.001). Thus, accounting for school-level clustering did not materially alter the statistical interpretation of the observed pre–post changes.

Screening competency

Of the 264 teachers assessed at the end of Day 2, 252 (95.5%) demonstrated 100% competency across all checklist items on the initial assessment. The remaining 12 teachers (4.5%) required additional training in specific competency domains. These teachers received targeted retraining in the areas requiring improvement and were subsequently reassessed. All 12 achieved 100% competency following reassessment and were then certified to conduct school vision screening. Thus, all 264 teachers ultimately met the required competency criterion before certification.

Subgroup analyses

Subgroup analyses revealed no statistically significant differences in KAP change scores across age group, gender, educational qualification, class teacher status, pre-existing eye health training, pre-existing screening experience, subject category and teaching experience (all p > 0.05; Table 2).

Table 2.

Stratified analysis of changes in KAP scores

Variable	Knowledge Δ Median (IQR)	p-value	Attitude Δ Median (IQR)	p-value	Practice Δ Median (IQR)	p-value
Age group
< 35 years	2 (1–3)	0.352	1 (0–2)	0.428	2 (0–4)	0.119
≥ 35 years	2 (0–4)		1 (-1–2)		1 (-1–4)
Gender
Male	2 (0–4)	0.681	0.5 (0–1)	0.95	1 (0–5)	0.156
Female	1.5 (0–4)		1 (-1–3)		1 (-1–3)
Education level
Graduate	2 (0–4)	0.911	0 (-1–2)	0.779	1 (-1–4)	0.897
Postgraduate or above	2 (1–3)		1 (-1–2)		1 (0–4)
Class teacher status
No	1 (0–3.5)	0.534	1 (0–2)	0.154	2 (0–5)	0.314
Yes	2 (0–4)		0 (-1–2)		1 (-1–4)
Prior eye health training
No	2 (0–4)	0.650	0 (-1–2)	0.334	1 (0–4)	0.249
Yes	1 (-1–5)		1 (-1–4)		1 (-2.5–3.5)
Screening experience
No	2 (0–4)	0.585	0 (-1–2)	0.429	1 (0–4)	0.680
Yes	2 (0–4)		1 (0–3.5)		1 (0–3)
Subject category
Science	2 (0–3)	0.692	0 (0–1)	0.388	2 (0–5)	0.121
Non-science	2 (0–4)		1 (-1–2)		1 (-1–4)
Both	2 (0–4)		0 (-1–2)		1 (-1–4)
Teaching experience
≤ 10 years	2 (0–3)	0.071	0 (-1–2)	0.160	1 (0–4)	0.744
> 10 years	2 (0–4)		1 (-1–2)		1 (-1–4)

Values are presented as median change (IQR). P-values were obtained using the Mann–Whitney U test or Kruskal-Wallis test, as appropriate. All comparisons were non-significant (p > 0.05), indicating no differential response to the intervention across subgroups

Discussion

In this single-group pre–post study, teachers showed substantial within-participant gains in knowledge, a moderate increase in self-reported practice, and a statistically significant but small change in attitude after a structured two-day training programme. The magnitude of change was largest for knowledge (r = 0.66), followed by self-reported practice (r = 0.39) and attitude (r = 0.27). Because there was no control group, these changes cannot be attributed solely to the intervention; repeated testing, concurrent influences, and participation effects remain plausible alternative explanations.

Pre-training knowledge among teachers in Skardu was moderate, with important gaps in specific clinical and screening-related concepts. This is consistent with evidence from other Pakistani settings. Habiba et al. found high knowledge in only 35.9% of Rawalpindi primary teachers, while Khan et al. found moderate awareness (66%) among primary school teachers in Peshawar, with gaps in recognition of some ocular conditions [5, 15]. Similar findings have been reported from Nigeria, Ethiopia, Ghana and Swaziland, where teachers’ knowledge of childhood eye problems has varied but has often remained incomplete in the absence of formal training [24–26].

The knowledge gaps observed in Skardu are understandable given the context. Most participants had no prior eye health training and had not previously participated in school-based screening. In remote districts, teachers may have limited access to eye health information, specialist services or continuing professional development opportunities related to child health. This differs from more urban settings, where teachers may have greater exposure to health campaigns, school health services or personal contact with eye care providers. Such contextual differences may explain why some urban studies from LMICs have reported relatively higher baseline knowledge than studies conducted in rural or underserved areas.

Following the intervention, all 10 knowledge items had higher proportions of correct responses, with the largest gains in areas of lowest baseline familiarity, including childhood cataract and glaucoma, recommended frequency of eye examinations and Snellen chart interpretation. These gains compare favourably with Thummalapalli et al., who reported knowledge improvement from 39.3% to 63.7% among Indian government teachers after structured educational sessions, and with a South African study that showed marked improvements in knowledge of visual acuity screening procedures after training [17, 27]. Other international evidence also suggests teachers only become reliable partners in school eye health when technical skills are backed by broader understanding of eye conditions and when to refer [13, 14, 27]. The breadth of knowledge gain observed here may reflect the curriculum, which included ocular anatomy, common childhood eye conditions, warning signs, prevention, referral, and practical screening skills. However, because the same knowledge items were repeated immediately after training, some of the observed short-term gain may reflect testing or recall effects.

Attitude scores were already high before the intervention and improved only modestly after training. This pattern is consistent with local, LMIC and international KAP literature, where teachers frequently express positive attitudes toward early detection and referral of children’s health problems, even when their knowledge and practices are less developed [5, 15, 16, 25, 26, 28]. Teachers are generally motivated to support children’s learning and wellbeing, and many understand that untreated vision problems can affect classroom participation and educational performance.

In Skardu, favourable baseline attitudes may also reflect the social role of teachers in remote communities, where schools often serve as trusted points of contact between families and public services, and teachers may be aware of the difficulties children face in accessing timely care. Therefore, teachers may already believe that child eye health is important, even without having the technical skills required for screening or referral.

Attitude scores were already close to the maximum possible score before training, leaving limited room for measurable improvement. The unchanged median (23 before and after training) and small effect size (r = 0.27) indicate that the statistically significant change was modest in practical magnitude. This pattern is compatible with a ceiling effect, but it also means that the data provide limited evidence of meaningful attitude change. In KAP studies, broad statements about the importance of early detection, referral, and communication with parents may also be susceptible to socially desirable responding. The main implication is that attitude may not be the main barrier to teacher participation in school eye health programmes. Instead, programme success is more likely to depend on training quality, clear protocols, feasible workload expectations, supervision and functioning referral pathways.

Self-reported practice scores increased after the intervention, with a moderate effect size. The practice assessment occurred 2–3 weeks after training, after teachers had begun programme screening activities; therefore, it captures reported early implementation behaviour rather than immediate post-training intentions. However, these outcomes were not based on direct observation, and social-desirability or recall bias may have led to overestimation of actual practice change. Objective screening performance was evaluated separately within the programme and is not an outcome of the present analysis.

Local evidence from Pakistan has shown that teachers’ eye health practices are often inconsistent. In Rawalpindi, a significant proportion of teachers had poor practice scores (66.8%) despite some baseline awareness of children’s eye health [5]. In Quetta, several barriers to teacher-led vision screening, including workload pressures, limited training, uncertainty around referral and insufficient follow-up support were reported [9]. These barriers are important because positive attitudes alone do not ensure that teachers can act effectively in school settings.

The improvement in practice in the present study may be explained by the intervention design. Teachers received didactic instruction, supervised hands-on training, practical demonstrations and competency assessment. They were not trained as passive recipients of information but were given a defined role within an active school vision screening programme. The practice subscale captured behaviours that are directly relevant to school eye health, including observation of children for visual difficulty, adaptation of classroom support, communication with parents, referral and follow-up of spectacle use.

The programme pathway may also have supported practice change. Teachers were linked to a defined referral and validation process in which children who failed teacher-led screening were reassessed by optometrists during validation visits by the DOVS, AKUH team. Children requiring further evaluation were referred to ophthalmologists at the RHQ, Skardu, and those requiring refractive correction were provided with spectacles free of cost with support from the funding partner. This linkage between teacher training, optometrist validation, ophthalmologist referral and spectacle provision likely made teachers’ role more concrete and actionable. It also distinguishes the intervention from stand-alone awareness sessions, where teachers may gain information but lack a practical system through which to apply it.

Findings from India, South Africa, Zambia and Peru similarly suggest that teachers can support school eye health activities when training is practical, protocols are clear and referral systems are available [13, 17, 27, 29]. However, practice change is usually more difficult to sustain than knowledge gain because it depends on workload, school leadership, parental cooperation, availability of services and follow-up mechanisms. The moderate improvement observed in this study is therefore encouraging, but it should be supported through refresher training, supervision and quality assurance.

Exploratory subgroup analyses detected no statistically significant evidence of differential change across the teacher characteristics examined, including age, gender, education level, teaching experience, subject category, prior eye-health training, and prior screening experience. These findings suggest that no clear signal of differential change was detected in this sample; however, they do not establish equal effectiveness across subgroups. The subgroup analyses may have had limited power, particularly for small categories such as teachers with prior eye-health training, and should therefore be interpreted cautiously. This finding contrasts with some previous studies that reported associations between teacher characteristics and baseline KAP levels [5, 30]. A cross-sectional study in Ghana reported a statistically significant association between academic qualification and knowledge of ocular conditions (p = 0.01) [30]. Such baseline associations may reflect differences in previous educational or health-related exposure.

From a programmatic perspective, the absence of detected differential change suggests that the training could be delivered to teachers with varied professional backgrounds. However, the present findings do not demonstrate equivalent effects across these groups or establish broad generalizability. The purposive sample was drawn from government schools in Skardu, and further studies involving larger and more diverse samples are needed to examine potential effect modification and determine whether similar findings are observed in other teacher populations and settings.

This study has several strengths. It included 264 teachers from 132 government schools in a remote mountainous district, providing evidence from a region that is under-represented in the school eye health literature. The intervention combined theory-based instruction, supervised practical training, and formal competency assessment delivered by ophthalmologists and optometrists. The use of paired statistical analyses and effect sizes also provided a clear assessment of both statistical significance and magnitude of change. The timing of the practice assessment, conducted after teachers had begun screening activities, allowed early implementation behaviour to be captured rather than immediate post-training intention alone.

The study also has several limitations. The single-group pre–post design without a control group limits causal inference; therefore, the observed changes cannot be attributed solely to the training intervention and may partly reflect repeated testing, increased attention from participation, or the Hawthorne effect. The follow-up period was short: knowledge was reassessed immediately after training, while attitude and practice were reassessed after 2–3 weeks. Consequently, the findings reflect short-term changes and cannot establish long-term retention or sustained behaviour change. In addition, attitude and practice were self-reported and may have been influenced by social-desirability bias. This is particularly important for practice, where the median score increased only modestly from 22 to 23, while attitude showed no change in the median score despite statistical significance. Therefore, caution is warranted in interpreting the observed changes in attitude and self-reported practice as evidence of meaningful or sustained behavioural change. Finally, purposive sampling was operationally appropriate for the programme but limits generalizability beyond the participating schools and similar remote districts.

This study focuses only on teacher-level KAP outcomes. Teacher screening accuracy and student-level outcomes, including validation findings, referral completion and spectacle provision, were evaluated in a separate linked study within the same programme and will be reported separately. This separation allows the present analysis to focus on whether training improved teachers’ preparedness for school eye health activities, while the linked study examines whether trained teachers could accurately identify children requiring further eye care.

Future school eye health programmes in remote settings should combine teacher training with refresher training, periodic competency checks, supportive supervision and referral tracking. Training is most likely to have sustained public health value when it is embedded within a functional pathway that connects schools, optometrists, ophthalmologists, spectacle provision and follow-up of children who fail screening. These components will be essential for translating improved teacher KAP into long-term gains in child eye health and educational participation.

Conclusion

Following the structured two-day teacher-training programme, teachers demonstrated a large short-term improvement in knowledge, while self-reported practice showed a moderate effect size but a modest absolute increase in median score. Attitude scores were high at baseline and showed only a small-magnitude statistical change, with no change in the median score. No significant differences in KAP changes were observed across the examined participant subgroups, suggesting that the training approach may be broadly applicable across teachers with different demographic and professional characteristics; however, these findings do not establish equivalent effects across subgroups and require confirmation in larger studies. Given the single-group pre–post design, short follow-up period, and self-reported practice outcomes, the findings should be interpreted as short-term within-participant changes following training rather than evidence of causal effectiveness or sustained behavioural change. Further controlled studies with longer follow-up are needed to confirm the findings and assess their generalizability.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

KA contributed to programme design, conceptualisation, methodology, protocol development, development of training content and materials, delivery of training, investigation, data curation, analysis, interpretation, drafting of the original manuscript and overall supervision of the study and critical review of the manuscript. FA contributed to methodology, development of assessment tools, delivery of training, data cleaning and analysis, drafting of manuscript sections, and critical review and editing of the manuscript. HT contributed to methodology, co-supervision of the study, interpretation of findings, drafting of manuscript sections, and critical review and editing of the manuscript. KFD contributed to methodology, co-supervision of the study, interpretation of findings, drafting of manuscript sections, and critical review and editing of the manuscript.

Funding

This work was supported by the National Sight Restoring Project 2019–2025, implemented by The Fred Hollows Foundation and funded by Australia’s Department of Foreign Affairs and Trade through the Australian NGO Cooperation Programme.

Data availability

The datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request. The data are not publicly available because they contain information that could potentially compromise participant confidentiality.

Declarations

Ethical approval

This study was approved by the Ethical Review Committee at Aga Khan University, Karachi (Ref 2025-11851-36671) in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.