BACKGROUND AND OBJECTIVES:

School health programs are frequently attempted in low- and/or middle-income countries; however, programmatic scope and reach is limited by human resource constraints. We sought to determine if trained community members could implement a school health program that improved outcomes in rural primary schools in India.

METHODS:

This was a mixed-methods, stepped-wedge, cluster-controlled study of schools pragmatically assigned to receive a multicomponent, comprehensive school health program delivered by lay field-workers.

RESULTS:

All students in 22 primary schools (9 government schools and 13 low-cost private schools) participated in this study. A total of 3033 student-years were included in the analysis (2100 student-years in the intervention period and 933 student-years in the control period). Qualitative feedback was collected from 38 teachers, 49 parents, and 4 field-workers. In low-cost private schools, the diarrhea incidence was lower in students receiving the intervention (incidence rate ratio 0.58; 95% confidence interval [CI] 0.47 to 0.71; P < .001). There was no difference in diarrhea incidence for students in government schools (incidence rate ratio 0.87; 95% CI 0.68 to 1.12; P = .29). Health-knowledge acquisition was higher in intervention schools (mean difference 12.6%; 95% CI 8.8 to 16.4; P < .001) and similar in both school types. Intervention coverage rates were high (mean 93.9%; SD 2.0%), and performance assessment scores indicated fidelity (mean 3.45; SD 0.69). Stakeholders revealed favorable perceptions of the field-workers and high levels of perceived impact.

CONCLUSIONS:

Lay field-worker–led school health programs offer a promising alternative for improving school health delivery in resource-constrained settings.

What’s Known on This Subject:

School-aged children in low- and/or middle-income countries have unmet health needs. School health programs that follow the World Health Organization health-promoting school framework improve a range of health outcomes. However, the impact of such programs is limited by human resource constraints.

What This Study Adds:

In Indian primary schools, trained community members successfully implemented a school health program. Compared with government schools, greater impact was seen in low-cost private schools. Delivery of school health programs by lay field-workers is a promising approach in low- and/or middle-income countries.

School-aged children in low- and/or middle-income countries (LMICs) are burdened by unmet health needs that threaten their ability to develop and thrive. These challenges include iron deficiency anemia, helminth infections, and untreated mental illness.1,3 Worldwide, preventable infectious diseases, such as diarrheal diseases, are the primary cause of mortality in children <14 years of age.4 Additionally, 70% of adult premature deaths are linked to behaviors developed in childhood.5 

Children spend a large portion of their lives in school; thus, schools can provide a safe and consistent venue in which to promote positive change in health behaviors. In addition, there is a strong connection between children’s well-being and their capacity to learn, so it is in a school’s interest to invest in programs that promote health.6,8 Health promotion occurring within a positive school environment helps students develop healthy beliefs and behaviors, enabling them to achieve improved health and education outcomes.9,10 To leverage this potential, the World Health Organization (WHO) developed the health-promoting school (HPS) framework.11 Systematic reviews suggest that this model can positively affect a range of health outcomes, and HPS initiatives are widely pursued in LMICs.12,14 

Despite strong guidelines, successful implementation of comprehensive school health programs has been a challenge in LMICs. Few programs have successfully reached the target population and delivered the HPS approach in its entirety.15,18 A lack of skilled and available human resources has been cited as a barrier to implementation.17,19,20 Traditional delivery agents include teachers, midlevel health providers, and physicians.21 However, in many communities, there is a shortage of medical professionals, and teachers are overburdened with competing responsibilities. Thus, the identification of effective delivery agents is a pressing need.

A proposed alternative approach is to leverage existing community members to deliver school health programs.22,23 We developed a task-shifting model that uses lay field-workers to impact children in primary schools in the Darjeeling Himalayas. The novel Comprehensive Health and Hygiene Improvement Program (CHHIP) was piloted from 2012 to 2016. Here, we present the results of an assessment of the viability of this model for achieving comprehensive school health in resource-constrained primary schools.

This evaluation was conducted between 2012 and 2016 in the Darjeeling Himalayas, a region of the state of West Bengal in India. Within India, the region is geographically and ethnically distinct. A population of >800 000, the majority of whom are ethnically Nepali, lives in villages scattered across the mountainous region.24 Similar to the rest of rural India, economic conditions are poor. The majority of workers are tea plantation laborers earning daily wages of 120 Indian rupees ($1.84).25 In contrast to the economic conditions, literacy rates (79.6%) exceed the national average (74%).26 The perceived value of education is prompting families to increasingly select private education for their children, yet many still rely on government schools for their child’s education. To reach this target population, we implemented the intervention in both rural government and low-cost private primary schools. Further discussion of this education system, along with student characteristics by school type, is presented in the Supplemental Information and Supplemental Table 6.

In the study, we aimed to determine if trained community members could deliver school health programs in rural primary schools in LMICs. We hypothesized that a lay field-worker–led program would result in improved health and education outcomes. To test this, we used a mixed-methods, stepped-wedge, nonrandomized cluster-controlled design. Additionally, we coupled an evaluation of effectiveness with a secondary analysis of implementation.27,28 For both effectiveness and implementation, quantitative measures were selected, and qualitative feedback was integrated via a process of triangulation.

In stepped-wedge cluster-controlled trials, an intervention is introduced over time with sequential crossover of clusters from control to intervention.29 This design was chosen to facilitate robust evaluation while accounting for the pragmatic constraints affecting intervention delivery. Given the nature of the intervention, each school was a cluster, and each step was an academic year. Three geographic regions representative of rural Darjeeling were identified, and all (33) primary schools in these regions were approached for participation. Project staff pragmatically chose a convenience sample of 22 schools: 13 government schools and 9 low-cost private schools. On the basis of organizational capacity, a control arm was introduced in the second year, and subsequently, the number of schools in the intervention arm was determined on the basis of available funding. Schools were sequentially enrolled and transitioned between study conditions; assignment to intervention was nonrandom, although project staff informally matched schools on the basis of geographic location and school type to maintain comparability. Program implementation occurred in 16 schools and was led by 4 lay field-workers. The target for the intervention was all students enrolled in primary grade levels (kindergarten through class 4). The parents of these students consented to their children’s participation. Less than 5% of parents chose to opt their children out of receiving specific individual health interventions.

The intervention was implemented as a community development program with a rigorous evaluation component, and outcome data were collected prospectively. Sociodemographic data for study subjects were collected ex post facto in 2018 and matched to the original study subjects by using 5-point validation (including name, age, sex, class, and school). Participants, field-workers, and members of the research team were not blinded to study assignment. The research protocol was approved by the Colorado Multiple Institutional Review Board and is registered with ClinicalTrials.gov (identifier NCT03423615). Guidelines for reporting nonrandomized evaluations were followed.30 

CHHIP is an intense, multicomponent, comprehensive school health program based on the guidelines of the Government of India,21 the international Focusing Resources on Effective School Health framework,31 and the WHO HPS framework.32 The program is structured around 3 interrelated components: (1) health education, (2) primary health services, and (3) a healthy school environment. As illustrated in the intervention logic model (Supplemental Fig 4), activities include health classes, school-based treatment, screening and facilitated referrals, infrastructure improvements, and modeling a positive psychosocial environment.

Four lay field-workers, school health activists (SHAs), delivered the intervention. A community committee led the hiring process. Recruitment criteria included a minimum age of 25 years, completion of secondary school, and residency within the targeted catchment area. The SHAs received 14 days of initial training, plus 5 days of continued professional development annually. For an average 30-hour workweek, they received a regionally competitive salary. The SHAs spent 1 day per week in each of the assigned 4 schools, where they taught the health curriculum, delivered interventions, supported access to medical care, and served as counselors. Additionally, SHAs promoted a healthy school environment by modeling positive behavior reinforcement, activity-based learning, and alternative discipline methods. The per-school cost was ∼$2300 per year.

Within this complex intervention with multiple potential outcomes, diarrheal illness was identified as a key target on the basis of the results of a community assessment. Hence, reduction in diarrhea incidence was chosen as the primary outcome to assess intervention effectiveness. Similarly, health-knowledge acquisition was identified by the community as a key result of the curriculum and an important mediator of impact. This led to health-knowledge acquisition being a secondary outcome. Given that anthropometric data were collected during delivery of health services, growth outcomes for children only in the intervention group were analyzed. Diarrheal incidence data were obtained by a 14-day calendar survey administered 2 to 3 times per year to parents.33 Health-knowledge acquisition was assessed via an internally developed assessment administered at the start and end of each year.

Implementation was evaluated in terms of reach and fidelity. Reach was assessed by coverage rates for select components of the primary health services package. A predefined benchmark of 90% of eligible students receiving the intervention was set to indicate successful engagement of the target population. Data for coverage rates was obtained at the time of the intervention. To assess fidelity, 2 project members conducted performance observations of the SHAs (n = 92). Guided by internal rubrics, observers independently rated the SHA on a 1 to 5 scale at 0.5 intervals before collaborative determination of a final score. A predefined benchmark of 3 was set to indicate quality and consistency of implementation as planned.

To capture opinions of both parents and teachers, we conducted focus groups in all study regions. A pragmatic sampling strategy was used. During a site visit, the first 15 parents who were encountered and all teachers were invited to separate discussions. Semistructured interviews were also conducted with each SHA. Focus group and interview guides are presented in the Supplemental Information. Project staff conducted interviews and focus groups in December 2016. Verbal informed consent was obtained from participants.

The analysis for the primary outcome (diarrhea incidence) and secondary outcome (health-knowledge acquisition) was conducted as a comparison between the intervention and control arms within the context of the stepped-wedge framework. The analysis was based on child-level data over a single year without linking data between years. Multilevel generalized linear mixed models were used to account for clustering at the school level and repeated measures within the same year. To account for imbalance, exposures of interest were explored with the primary and secondary outcomes, and a forward stepwise procedure was used to select variables for the final regression models. Because tea plantations are characterized by crowded living conditions and because transmission of diarrhea varies by season in monsoon climates, we tested community type and season to account for potential effect on diarrheal disease transmission.34,35 Because of effect modification and to facilitate comparison by context, results were stratified by school type. All P values were 2-tailed, and significance was set at P < .05.

Weight-for-age and height-for-age z scores were obtained on the basis of the National Center for Health Statistics and WHO international reference population.36 Mean results were compared based on the cumulative time of intervention exposure at the school level and a multilevel mixed regression model used to assess for intervention effect.

To study implementation outcomes, we conducted a series of descriptive analyses. Coverage rates were expressed as the percentage of eligible students receiving the individual intervention in a given year. Performance evaluation scores were expressed as means and SDs and subanalyzed by intervention component (education and health) and by a field-worker.

Coding and analysis of the qualitative data was an iterative process of reading, coding, summarizing, and rereading. Coding began with a deductive coding method with added emergent codes. Data were segmented by code, and a case ordered matrix was created. Key constructs were summarized across the domains of analysis, and a provisional set of themes was generated. These themes were reviewed, and important contrary opinions were identified. Ultimately, a final set of common themes and illustrative examples were identified.

All quantitative analyses were done in SPSS (version 24.0; IBM SPSS Statistics, IBM Corporation) and qualitative analysis was completed in CATMA (version 5.0; http://catma.de/). Wealth quintiles were calculated by using The EquityTool (www.equitytool.org).37 

Over 5 academic years (2012–2016), 22 primary schools were pragmatically enrolled and assigned to the study period. Figure 1 provides a schematic representation of the flow of schools and participants during the study period. The attrition of 3 schools occurred over the course of the study because of school closure, staff safety concerns, and voluntary withdrawal from the program. All data collected from students enrolled in a school at the time of their participation were included in the intention-to-treat analysis.

FIGURE 1

Trial diagram showing the flow of schools and students by study period over time. The number of students enrolled in each school for a given year is included within the figure. Tables provide the total number of students and schools by school type and year for each study condition.

FIGURE 1

Trial diagram showing the flow of schools and students by study period over time. The number of students enrolled in each school for a given year is included within the figure. Tables provide the total number of students and schools by school type and year for each study condition.

Close modal

A total of 3033 student-years were included in the analysis (2100 student-years in the intervention period and 933 student-years in the control period). The demographic characteristics of these students are shown in Table 1. Within this pragmatic sample, there was no significant difference in age (P = .43), sex (P = .79), class (P = .98), household size (P = .64), and birth order (P = .36). However, the proportion of student-years in low-cost private schools was significantly higher in the intervention period (1339; 66%) than in the control period (532; 57%). Additionally, there was a significant difference in several characteristics, including wealth status, parent’s education status, and year of participation. Qualitative data were obtained from 38 teachers and 49 parents who participated in a total of 8 focus groups; all 4 SHAs participated in semistructured interviews.

TABLE 1

Demographic Characteristics of Participating Children by Study Period and School Type

CharacteristicsAll SchoolsGovernment SchoolsLow-cost Private Schools
Intervention (n = 2100)Control (n = 933)PaIntervention (n = 712)Control (n = 400)PIntervention (n = 1388)Control (n = 533)P
Age, y, mean (SD, minimum, maximum)b 7.16 (2.18, 2, 14) 7.24 (2.38, 2, 12) .43 7.29 (2.46, 2, 14) 8.03 (2.41, 2,11) <.001 7.10 (2.03 2,14) 6.72 (2.22, 2, 11) .002 
Female sex, No. (%) 988 (53) 402 (47) .79 360 (51) 236 (59) <.001 602 (43) 258 (48) .71 
Class, No. (%)c          
 Kindergarten 723 (34) 319 (34) .98 503 (36) 221 (42) .13 220 (31) 98 (25) .10 
 Class 1–2 658 (31) 294 (32) — 450 (32) 159 (30) — 208 (29) 135 (34) — 
 Class 3–4 714 (34) 316 (34) — 431 (31) 149 (28) — 283 (40) 167 (42) — 
Scheduled caste or scheduled tribe, No. (%)d 866 (41) 413 (46) .01 293 (41) 213 (53) <.001 572 (41) 200 (41) .81 
Wealth status, No. (%)e          
 Quintile 2 128 (6) 31 (4) .001 96 (14) 23 (6) <.001 32 (2) 8 (2) .003 
 Quintile 3 369 (19) 143 (16) — 175 (26) 99 (26) — 194 (15) 44 (9) — 
 Quintile 4 1319 (66) 641 (72) — 386 (57) 245 (63) — 933 (71) 396 (78) — 
 Quintile 5 179 (9) 78 (9) — 23 (3) 20 (5) — 155 (12) 58 (12) — 
Annual income, $, mean (SD)f 1255 (1631) 1168 (1047) .08 803 (705) 914 (618) .01 1486 (1899) 1358 (1244) .15 
Children in household, mean (SD) 2.21 (1.16) 2.10 (1.03) .008 2.67 (1.35) 2.37 (1.21) <.001 1.97 (0.96) 1.89 (0.82) .11 
Total household size, mean (SD) 5.04 (1.65) 5.01 (1.53) .640 5.38 (1.87) 5.16 (1.4) .03 4.86 (1.50) 4.89 (1.62) .68 
Mother’s educational status, No. (%)          
 Some primary 646 (31) 313 (34) <.001 338 (49) 205 (52) <.001 308 (22) 108 (20) <.001 
 Primary 214 (10) 37 (4) — 100 (15) 17 (4) — 114 (8) 20 (4) — 
 Secondary 1081 (52) 543 (59) — 242 (35) 167 (43) — 839 (61) 376 (71) — 
 Higher secondary 93 (5) 30 (3) — 3 (0) 2 (1) — 90 (6) 28 (5) — 
 Undergraduate or higher 33 (2) 2 (0) — 6 (1) 2 (1) — 27 (2) 0 (0) — 
Father’s educational status, No. (%)          
 Some primary 384 (19) 190 (21) <.001 198 (28) 133 (35) .001 191 (14) 57 (11) <.001 
 Primary 287 (14) 88 (10) — 138 (20) 57 (15) — 149 (11) 31 (6) — 
 Secondary 1134 (55) 546 (60) — 310 (44) 177 (47) — 824 (60) 369 (70) — 
 Higher secondary 188 (9) 76 (8) — 39 (6) 11 (3) — 149 (11) 65 (12) — 
 Undergraduate or higher 83 (4) 9 (1) — 16 (2) 1 (0) — 67 (5) 8 (2) — 
Birth order, No. (%)          
 First born or only child 1002 (48) 428 (46) .36 237 (33) 120 (30) .26 765 (55) 308 (58) .27 
School type, No. (%)          
 Government 712 (34) 400 (43) <.001 — — — — — — 
 Low-cost private 1339 (66) 532 (57) — — — — — — — 
Year, No. (%)          
 2012 416 (21) 0 (0) <.001 82 (12) 0 (0) <.001 293 (21) 0 (0) <.001 
 2013 346 (18) 283 (33) — 100 (14) 90 (23) — 249 (18) 236 (44) — 
 2014 422 (22) 230 (27) — 163 (23) 113 (28) — 305 (22) 141 (27) — 
 2015 378 (19) 207 (24) — 210 (30) 105 (26) — 243 (18) 45 (9) — 
 2016 401 (20) 136 (16) — 157 (22) 92 (23) — 298 (22) 45 (9) — 
Community type, No. (%)g          
 Tea plantation 285 (15) 295 (35) <.001 128 (18) 277 (69) <.001 196 (14) 49 (9) .003 
 Agricultural village 1673 (85) 561 (65) — 584 (82) 123 (31) — 1189 (86) 483 (91) — 
Season, No. (%)h          
 Premonsoon 1169 (23) 570 (27) <.001 371 (23) 248 (28) .06 779 (23) 352 (28) .01 
 Monsoon 1970 (39) 751 (36) — 611 (39) 319 (36) — 1313 (39) 4561 (36) — 
 Postmonsoon 1961 (39) 765 (37) — 602 (38) 328 (37) — 1253 (38) — — 
CharacteristicsAll SchoolsGovernment SchoolsLow-cost Private Schools
Intervention (n = 2100)Control (n = 933)PaIntervention (n = 712)Control (n = 400)PIntervention (n = 1388)Control (n = 533)P
Age, y, mean (SD, minimum, maximum)b 7.16 (2.18, 2, 14) 7.24 (2.38, 2, 12) .43 7.29 (2.46, 2, 14) 8.03 (2.41, 2,11) <.001 7.10 (2.03 2,14) 6.72 (2.22, 2, 11) .002 
Female sex, No. (%) 988 (53) 402 (47) .79 360 (51) 236 (59) <.001 602 (43) 258 (48) .71 
Class, No. (%)c          
 Kindergarten 723 (34) 319 (34) .98 503 (36) 221 (42) .13 220 (31) 98 (25) .10 
 Class 1–2 658 (31) 294 (32) — 450 (32) 159 (30) — 208 (29) 135 (34) — 
 Class 3–4 714 (34) 316 (34) — 431 (31) 149 (28) — 283 (40) 167 (42) — 
Scheduled caste or scheduled tribe, No. (%)d 866 (41) 413 (46) .01 293 (41) 213 (53) <.001 572 (41) 200 (41) .81 
Wealth status, No. (%)e          
 Quintile 2 128 (6) 31 (4) .001 96 (14) 23 (6) <.001 32 (2) 8 (2) .003 
 Quintile 3 369 (19) 143 (16) — 175 (26) 99 (26) — 194 (15) 44 (9) — 
 Quintile 4 1319 (66) 641 (72) — 386 (57) 245 (63) — 933 (71) 396 (78) — 
 Quintile 5 179 (9) 78 (9) — 23 (3) 20 (5) — 155 (12) 58 (12) — 
Annual income, $, mean (SD)f 1255 (1631) 1168 (1047) .08 803 (705) 914 (618) .01 1486 (1899) 1358 (1244) .15 
Children in household, mean (SD) 2.21 (1.16) 2.10 (1.03) .008 2.67 (1.35) 2.37 (1.21) <.001 1.97 (0.96) 1.89 (0.82) .11 
Total household size, mean (SD) 5.04 (1.65) 5.01 (1.53) .640 5.38 (1.87) 5.16 (1.4) .03 4.86 (1.50) 4.89 (1.62) .68 
Mother’s educational status, No. (%)          
 Some primary 646 (31) 313 (34) <.001 338 (49) 205 (52) <.001 308 (22) 108 (20) <.001 
 Primary 214 (10) 37 (4) — 100 (15) 17 (4) — 114 (8) 20 (4) — 
 Secondary 1081 (52) 543 (59) — 242 (35) 167 (43) — 839 (61) 376 (71) — 
 Higher secondary 93 (5) 30 (3) — 3 (0) 2 (1) — 90 (6) 28 (5) — 
 Undergraduate or higher 33 (2) 2 (0) — 6 (1) 2 (1) — 27 (2) 0 (0) — 
Father’s educational status, No. (%)          
 Some primary 384 (19) 190 (21) <.001 198 (28) 133 (35) .001 191 (14) 57 (11) <.001 
 Primary 287 (14) 88 (10) — 138 (20) 57 (15) — 149 (11) 31 (6) — 
 Secondary 1134 (55) 546 (60) — 310 (44) 177 (47) — 824 (60) 369 (70) — 
 Higher secondary 188 (9) 76 (8) — 39 (6) 11 (3) — 149 (11) 65 (12) — 
 Undergraduate or higher 83 (4) 9 (1) — 16 (2) 1 (0) — 67 (5) 8 (2) — 
Birth order, No. (%)          
 First born or only child 1002 (48) 428 (46) .36 237 (33) 120 (30) .26 765 (55) 308 (58) .27 
School type, No. (%)          
 Government 712 (34) 400 (43) <.001 — — — — — — 
 Low-cost private 1339 (66) 532 (57) — — — — — — — 
Year, No. (%)          
 2012 416 (21) 0 (0) <.001 82 (12) 0 (0) <.001 293 (21) 0 (0) <.001 
 2013 346 (18) 283 (33) — 100 (14) 90 (23) — 249 (18) 236 (44) — 
 2014 422 (22) 230 (27) — 163 (23) 113 (28) — 305 (22) 141 (27) — 
 2015 378 (19) 207 (24) — 210 (30) 105 (26) — 243 (18) 45 (9) — 
 2016 401 (20) 136 (16) — 157 (22) 92 (23) — 298 (22) 45 (9) — 
Community type, No. (%)g          
 Tea plantation 285 (15) 295 (35) <.001 128 (18) 277 (69) <.001 196 (14) 49 (9) .003 
 Agricultural village 1673 (85) 561 (65) — 584 (82) 123 (31) — 1189 (86) 483 (91) — 
Season, No. (%)h          
 Premonsoon 1169 (23) 570 (27) <.001 371 (23) 248 (28) .06 779 (23) 352 (28) .01 
 Monsoon 1970 (39) 751 (36) — 611 (39) 319 (36) — 1313 (39) 4561 (36) — 
 Postmonsoon 1961 (39) 765 (37) — 602 (38) 328 (37) — 1253 (38) — — 

Percentages have been rounded and may not total 100. —, not applicable.

a

Calculated by using the independent sample t test for continuous variables and Pearson’s χ2 test for categorical variables.

b

All students in primary grade levels were enrolled in the study. Typical students in these grades ranged from 4- to 10-y-old students; however, enrollment was at the discretion of the school leadership, and the age range varied by individual school.

c

Classes are grouped as is typical for many rural primary schools in the study setting.

d

Scheduled caste and scheduled tribe are standard terms used in Indian demographic surveys to refer to officially recognized groups of historically disadvantaged peoples by the Government of India and State of West Bengal.

e

Wealth status is indicated by using quintiles of a continuous wealth index (The EquityTool) based on household characteristics as reported in 2018.

f

Annual income as reported by family in 2018. Presented in US dollars (1 US dollar = 65 Indian rupees).

g

Schools were categorized into 2 community types on the basis of whether the majority of enrolled families resided on a tea plantation or in an agricultural village.

h

Season is presented for the outcome of diarrheal incidence. Season is used to reference the time of year in which diarrheal illness was assessed.

The unadjusted diarrhea incidence among students in the intervention period (4.70 episodes per 100 person-weeks at risk; 95% confidence interval [CI] 4.50 to 5.34) was lower than that observed in the control period (8.05 episodes per 100 person-weeks at risk; 95% CI 7.80 to 9.20; Supplemental Tables 7 and 8). An analysis revealed evidence of effect modification by school type. Figure 2 shows that in low-cost private schools, the adjusted incidence of diarrheal illness was substantially lower (incidence rate ratio [IRR] 0.58; 95% CI 0.47 to 0.71; P < .001). However, there was little evidence of impact in government schools (IRR 0.87; 95% CI 0.68 to 1.12; P = .29).

FIGURE 2

Forest plot showing diarrhea incidence rates by school type. Results are graphically represented as point estimates of IRR and 95% CIs and are based on multilevel generalized linear mixed regression models, accounting for clustering at the school level and repeated measures within the same year and adjusting for school type (when appropriate), year, season, scheduled caste or scheduled tribe status, number of children in a household, wealth quintile, mother’s educational status, and father’s educational status.

FIGURE 2

Forest plot showing diarrhea incidence rates by school type. Results are graphically represented as point estimates of IRR and 95% CIs and are based on multilevel generalized linear mixed regression models, accounting for clustering at the school level and repeated measures within the same year and adjusting for school type (when appropriate), year, season, scheduled caste or scheduled tribe status, number of children in a household, wealth quintile, mother’s educational status, and father’s educational status.

Close modal

Health-knowledge acquisition differed by trial period. Table 2 shows that students exposed to the health education curriculum scored higher on the posttest assessment (mean difference 12.6%; 95% CI 8.8 to 16.4; P < .001). The magnitude of effect was similar when stratified by school type.

TABLE 2

Health-Knowledge Assessment Scores by School Type

Intervention Period Posttest Mean, %Control Period Posttest Mean, %Mean Difference, % (95% CI)PEffect Size, d
All schools 67.9 54.3 12.6 (8.8 to 16.4) <.001 0.67 
Government schools 66.4 51.8 14.5 (10.6 to 18.5) <.001 0.74 
Low-cost private schools 68.2 56.3 11.9 (6.4 to 17.3) <.001 0.68 
Intervention Period Posttest Mean, %Control Period Posttest Mean, %Mean Difference, % (95% CI)PEffect Size, d
All schools 67.9 54.3 12.6 (8.8 to 16.4) <.001 0.67 
Government schools 66.4 51.8 14.5 (10.6 to 18.5) <.001 0.74 
Low-cost private schools 68.2 56.3 11.9 (6.4 to 17.3) <.001 0.68 

Based on multilevel generalized linear mixed regression model, accounting for clustering at the school level and adjusting for pretest score, school type (when appropriate), year, class, wealth status, and scheduled caste or scheduled tribe status. d, Cohen’s d.

An anthropometric analysis revealed modest improvement in height z scores (β .017; 95% CI 0.007 to 0.026; P < .001) and weight z scores (β .012; 95% CI 0.004 to 0.020; P = .005) with increased exposure to the intervention (Table 3).

TABLE 3

Anthropometry by Cumulative Time School Is Exposed to the Intervention

OutcomeT0T1T2T3T4β (95% CI)aP
Height z score, mean (SD)        
 All schools −1.41 (1.50) −1.56 (1.23) −1.35 (0.97) −1.26 (1.16) −0.84 (0.92) .017 (0.007 to 0.026) <.001 
 Government schools −1.58 (1.49) −1.77 (1.27) −1.59 (1.14) −1.79 (1.40) −0.74 (0.93) .009 (0.002 to 0.019) <.001 
 Low-cost private schools −1.31 (1.50) −1.44 (1.19) −1.27 (0.88) −1.10 (1.02) −0.87 (1.12) .065 (0.041 to 0.089) <.001 
Wt z score, mean (SD)        
 All schools −1.37 (1.21) −1.13 (1.09) −1.26 (0.95) −1.28 (1.04) −1.04 (0.96) .012 (0.004 to 0.020) .005 
 Government schools −0.77 (1.11) −0.19 (0.98) −0.71 (0.94) −0.83 (1.10) −0.72 (0.93) −.008 (−0.029 to 0.013) .45 
 Low-cost private schools −1.37 (1.20) −1.09 (1.10) −1.26 (0.93) −1.25 (1.07) −1.00 (1.02) .004 (0.019 to 0.063) <.001 
OutcomeT0T1T2T3T4β (95% CI)aP
Height z score, mean (SD)        
 All schools −1.41 (1.50) −1.56 (1.23) −1.35 (0.97) −1.26 (1.16) −0.84 (0.92) .017 (0.007 to 0.026) <.001 
 Government schools −1.58 (1.49) −1.77 (1.27) −1.59 (1.14) −1.79 (1.40) −0.74 (0.93) .009 (0.002 to 0.019) <.001 
 Low-cost private schools −1.31 (1.50) −1.44 (1.19) −1.27 (0.88) −1.10 (1.02) −0.87 (1.12) .065 (0.041 to 0.089) <.001 
Wt z score, mean (SD)        
 All schools −1.37 (1.21) −1.13 (1.09) −1.26 (0.95) −1.28 (1.04) −1.04 (0.96) .012 (0.004 to 0.020) .005 
 Government schools −0.77 (1.11) −0.19 (0.98) −0.71 (0.94) −0.83 (1.10) −0.72 (0.93) −.008 (−0.029 to 0.013) .45 
 Low-cost private schools −1.37 (1.20) −1.09 (1.10) −1.26 (0.93) −1.25 (1.07) −1.00 (1.02) .004 (0.019 to 0.063) <.001 

Mean anthropometry results are presented as age- and sex-adjusted z scores in relation to the National Center for Health Statistics and WHO reference population. Anthropometric data were collected as a routine component of intervention activities. Thus, results are presented here only for children exposed to the intervention.; T0, first year of intervention exposure at the school level; T1, second year of intervention exposure at the school level; T2, third year of intervention exposure at the school level; T3, third year of intervention exposure at the school level; T4, fifth year of intervention exposure at the school level.

a

β represents the expected change in the mean z score per each additional year of intervention exposure. Results are based on the multilevel generalized linear mixed model, accounting for clustering at the school level and adjusting for age, sex, and school type (when appropriate).

Qualitative feedback revealed a strong perception of positive impact (Table 4). Parents reported improved physical health, as manifested by weight gain and positive behavior change. Teachers attested to the effect of positive role modeling by the SHAs and resulting changes in the school learning culture and norms. Specifically, they highlighted reduced use of corporal punishment. One area in which stakeholders identified limited impact was increased access to health care. In each teacher focus group and interview, economic barriers were cited as a common cause of incomplete referrals.

TABLE 4

Representative Themes and Quotes From Qualitative Focus Groups and Semistructured Interviews

Themes and SubthemesQuotes
Effectiveness  
 Physical health Caregiver: “I have observed changes and seen results. My child’s weight has gone up from 12 to 18 kg. It hasn’t even been a complete year of him coming to CHHIP and receiving the supplements, but there has been changes, and ultimately, everyone will notice changes in their children too.” 
 Corporal punishment Classroom teacher: “I used to be a very strict teacher; I would sometimes even hit the children if they were not studying properly. But after [the SHA] started taking lessons, I saw that the children are very happy and excited for her class and they get along. Whereas in my class, the children do not even come close. So, I stopped taking a stick to class and I understood that it is better to be friendly with the children, and they learn better that ways.” 
 Access to health care Classroom teacher: “When you refer cases to various doctors, some of the children still do not go...this of course could be because of the financial difficulty faced by the parents.” 
 Health knowledge Parent: “Last time I had carried an apple, I didn’t wash it before packing it. When I gave it to my daughter, she got very angry. She asked me why I would give her an unwashed apple. They are full of germs. To calm her down, I had to rewash the apple in front of her. This proves and shows how they have developed and learnt through the lessons taught by the SHA.” 
 Behavior change SHA: “The children have learned a lot regarding water, saving water, and safe water. They do not drink unboiled water; they bring boiled water to school.” 
Implementation  
 Facilitator-school engagement Classroom teacher: “We consider the SHA to be a part of the staff too. We see her coming and teaching. She teaches her class properly, and we feel she is doing it properly. When we see her, we know she is doing her job properly.” 
 Facilitator trust Parent: “Those who conduct the examinations are individuals who have training of some sort and are reliable. We believe in them and we trust them with our children, and that is why we are giving our approval and sending them for examinations.” 
 Facilitator enthusiasm for program Classroom teacher: “Health is wealth. The children have learnt basic health knowledge.” 
 Facilitator professional conduct Classroom teacher: “Yes, 100%, she does her work full heartedly. She does the work given to her in the allotted time. She implements the trainings she has got from the program. She is very dedicated in her work, be it games or lessons.” 
 Facilitator training Classroom teacher: “Knowing that she is from CHHIP and she has received health training from CHHIP automatically makes us want to trust her.” 
 Barrier: parental engagement SHA: “For CHHIP to improve and better itself, it needs to find new interventions so that parents will do what CHHIP advises them to do regarding their health.” 
 Barrier: communication Parent: “We don’t know much about what you can do or cannot do. So, if you tell us the areas you can deal with, then we can give suggestions and learn more about how you function as well.” 
 Barrier: traditional beliefs Parent: “There are still some people who believe that if you eat the iron medicine, their child will start seeing gods and spirits. So, it is really difficult to get through to other parents.” 
Themes and SubthemesQuotes
Effectiveness  
 Physical health Caregiver: “I have observed changes and seen results. My child’s weight has gone up from 12 to 18 kg. It hasn’t even been a complete year of him coming to CHHIP and receiving the supplements, but there has been changes, and ultimately, everyone will notice changes in their children too.” 
 Corporal punishment Classroom teacher: “I used to be a very strict teacher; I would sometimes even hit the children if they were not studying properly. But after [the SHA] started taking lessons, I saw that the children are very happy and excited for her class and they get along. Whereas in my class, the children do not even come close. So, I stopped taking a stick to class and I understood that it is better to be friendly with the children, and they learn better that ways.” 
 Access to health care Classroom teacher: “When you refer cases to various doctors, some of the children still do not go...this of course could be because of the financial difficulty faced by the parents.” 
 Health knowledge Parent: “Last time I had carried an apple, I didn’t wash it before packing it. When I gave it to my daughter, she got very angry. She asked me why I would give her an unwashed apple. They are full of germs. To calm her down, I had to rewash the apple in front of her. This proves and shows how they have developed and learnt through the lessons taught by the SHA.” 
 Behavior change SHA: “The children have learned a lot regarding water, saving water, and safe water. They do not drink unboiled water; they bring boiled water to school.” 
Implementation  
 Facilitator-school engagement Classroom teacher: “We consider the SHA to be a part of the staff too. We see her coming and teaching. She teaches her class properly, and we feel she is doing it properly. When we see her, we know she is doing her job properly.” 
 Facilitator trust Parent: “Those who conduct the examinations are individuals who have training of some sort and are reliable. We believe in them and we trust them with our children, and that is why we are giving our approval and sending them for examinations.” 
 Facilitator enthusiasm for program Classroom teacher: “Health is wealth. The children have learnt basic health knowledge.” 
 Facilitator professional conduct Classroom teacher: “Yes, 100%, she does her work full heartedly. She does the work given to her in the allotted time. She implements the trainings she has got from the program. She is very dedicated in her work, be it games or lessons.” 
 Facilitator training Classroom teacher: “Knowing that she is from CHHIP and she has received health training from CHHIP automatically makes us want to trust her.” 
 Barrier: parental engagement SHA: “For CHHIP to improve and better itself, it needs to find new interventions so that parents will do what CHHIP advises them to do regarding their health.” 
 Barrier: communication Parent: “We don’t know much about what you can do or cannot do. So, if you tell us the areas you can deal with, then we can give suggestions and learn more about how you function as well.” 
 Barrier: traditional beliefs Parent: “There are still some people who believe that if you eat the iron medicine, their child will start seeing gods and spirits. So, it is really difficult to get through to other parents.” 

The cumulative coverage rate for the assessed health interventions was 93.9% (SD 2.0%). Figure 3 reveals that the coverage rate for each individual activity exceeded the predefined benchmark of 90%. Similarly, the mean score from performance assessments exceeded the predefined threshold for quality and consistency of implementation (n = 92; mean 3.45; SD 0.69; Supplemental Table 5).

FIGURE 3

Coverage rates achieved by individual health intervention by year. Results are graphically represented with reference to the predefined benchmark (90%). A, Deworming. B, Iron supplementation. C, Vision screening. D, Growth monitoring. E, Oral health examination. F, Skin health examination. a The intervention was introduced in the 2013 school year. b The intervention was introduced in the 2014 school year.

FIGURE 3

Coverage rates achieved by individual health intervention by year. Results are graphically represented with reference to the predefined benchmark (90%). A, Deworming. B, Iron supplementation. C, Vision screening. D, Growth monitoring. E, Oral health examination. F, Skin health examination. a The intervention was introduced in the 2013 school year. b The intervention was introduced in the 2014 school year.

Close modal

Qualitative feedback revealed that the SHAs perceived minimal difficulty in conducting activities (Table 4). They attributed this to high levels of engagement at the school level. Teachers and parents held favorable perceptions of the SHAs as delivery agents. This was facilitated by trust, observation of professional conduct, and awareness that they received formal training.

Identified barriers to implementation included limited parental engagement. Parents felt that the program worked best when the SHA served as an intermediary, relaying important information about their child. However, limited communication with the SHA was noted as a barrier to parental engagement. All stakeholders noted the lack of intervention activities used to target parents and requested additional components to target their involvement.

The results of this nonrandomized cluster-controlled trial suggest that lay field-workers can effectively deliver HPS programs in rural primary schools in LMICs. Successful implementation occurred in all schools, with process indicators and qualitative feedback supporting the conclusion that the intervention was delivered as intended (fidelity) and to the target population (reach). In evaluating impact, some results differed by school type. For the outcomes of diarrhea and growth, the effect was greatest among children enrolled in low-cost private schools. Importantly, however, improvements in height and health-knowledge acquisition were observed in both subsamples.

It is likely that the impact of the intervention was influenced by differences between the government school and low-cost private school populations. This may be expected because a range of factors, from external to interventions, have been shown to impact child health and education outcomes.38,41 In this study, socioeconomic status was relatively higher among children in low-cost private schools. Additionally, there are broad differences between the 2 school types in terms of quality, culture, and community connectedness. Finally, the choice to direct limited economic resources to a private education likely reflects variance in family ethos and parental engagement. In light of these differences, it appears that the overall impact of the intervention was greater in more advantaged schools.

The results of this study reveal the importance of the context in which HPS programs take place while contributing to the evidence that such programs can improve student health outcomes in LMICs.42 Translating this evidence into effective services that reach underserved children has proved challenging. In most LMICs, there is widespread commitment to school health, with governmental policies in place. However, implementation of these policies is a challenge, and human resource constraints are a key barrier to effective delivery.43,46 Task shifting to lay community members represents a potential solution to this challenge.

Task shifting has been successful in many other areas of health care delivery, including maternal health, integrated case management of childhood illnesses, and mental health.47,51 Researches of a recent trial in India evaluated the effectiveness of a health promotion intervention in secondary schools. When delivered by lay counselors, there was strong evidence of impact; however, no effect was observed with teacher led delivery.52 To our knowledge, this study is the first application of this broadly generalizable approach to school health delivery in primary schools in an LMIC.

This study has a number of limitations. We obtained data from a program that was administered within real-world constraints, including nonrandomized allocation of clusters. Although we corrected for imbalances between study conditions, undetected bias may remain. We used a 14-day parental recall to assess diarrhea to match common practice in India.53 However, concern has been raised regarding the accuracy of parental report and underestimation of illness with extended recall periods.54,56 Finally, the field-workers were intensely supported, and this level of supervision may not be feasible at scale.

A lay field-worker–led school health program resulted in improved outcomes. This novel approach to administering a school health program was implemented successfully and generated favorable perceptions among stakeholders. This innovation holds immense promise for wide-scale impact. Our experience in a remote, rural region of an LMIC suggests that committed individuals who can be trained and empowered to deliver school health are widely available. Further research is required to directly compare delivery models, establish cost-effectiveness, and identify key implementation strategies required for adoption and sustainability. This research is critically important for the achievement of public health policy goals and to promote access to effective school health for disadvantaged children in LMICs.

Dr Matergia conceptualized and designed the study, conducted the quantitative data analysis, drafted the initial manuscripts, and reviewed and revised the manuscript; Mr Ferrarone and Ms Khan contributed to the design, implementation, and analysis of the qualitative assessment and reviewed and revised the manuscript; Ms Matergia, Ms Giri, and Ms Thapa contributed to the study concept, design of data collection instruments, and data collection and reviewed and revised the manuscript; Dr Simões contributed to the study design, data analysis, and drafting of the initial manuscript and reviewed and revised the manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

All deidentified individual participant data will be made available upon publication and with no end date. The data will be made available to researchers who provide a methodologically sound proposal for use in achieving the aims in the approved proposal. Proposals should be submitted to michael.matergia@ucdenver.edu.

This trial has been registered at www.clinicaltrials.gov (identifier NCT03423615).

FUNDING: Funded by Broadleaf Health and Education Alliance and supported by the Center for Global Health at the Colorado School of Public Health and by the American India Foundation through the William J. Clinton Fellowship for Service in India.

We are grateful to the staff of Darjeeling Ladenla Road Prerna and the SHAs for their dedication in implementing CHHIP. We thank Jean-Camille Kollmorgen and Fatima Salman for their contribution to the design of the evaluation framework and data collection tools. We thank Molly Lamb and Jonathon Murphy for their assistance in data processing and analysis. We acknowledge the support of Darjeeling Ladenla Road Prerna in providing access to the study data. We thank the children, families, and schools who participated in the program.

     
  • CHHIP

    Comprehensive Health and Hygiene Improvement Program

  •  
  • CI

    confidence interval

  •  
  • HPS

    health-promoting school

  •  
  • IRR

    incidence rate ratio

  •  
  • LMIC

    low- and/or middle-income country

  •  
  • SHA

    school health activist

  •  
  • WHO

    World Health Organization

1
Kassebaum
NJ
;
GBD 2013 Anemia Collaborators
.
The global burden of anemia.
Hematol Oncol Clin North Am
.
2016
;
30
(
2
):
247
308
[PubMed]
2
Brooker
S
,
Clements
AC
,
Bundy
DA
.
Global epidemiology, ecology and control of soil-transmitted helminth infections.
Adv Parasitol
.
2006
;
62
:
221
261
[PubMed]
3
Polanczyk
GV
,
Salum
GA
,
Sugaya
LS
,
Caye
A
,
Rohde
LA
.
Annual research review: a meta-analysis of the worldwide prevalence of mental disorders in children and adolescents.
J Child Psychol Psychiatry
.
2015
;
56
(
3
):
345
365
[PubMed]
4
Kassebaum
N
,
Kyu
HH
,
Zoeckler
L
, et al;
Global Burden of Disease Child and Adolescent Health Collaboration
.
Child and adolescent health from 1990 to 2015: findings from the global burden of diseases, injuries, and risk factors 2015 study.
JAMA Pediatr
.
2017
;
171
(
6
):
573
592
[PubMed]
5
Mathers
C
,
Stevens
G
,
Mascarenhas
M
.
Global Health Risks: Mortality and Burden of Disease Attributable to Selected Major Risks
.
Geneva, Switzerland
:
World Health Organization
;
2009
6
Bundy
DAP
,
de Silva
N
,
Horton
S
,
Jamison
DT
,
Patton
GC
, eds.
Optimizing Education Outcomes: High-Return Investments in School Health for Increased Participation and Learning
.
Washington, DC
:
World Bank Group
;
2018
7
Powney
J
,
Malcolm
H
,
Lowden
K
.
Health and Attainment: A Brief Review of Recent Literature
.
Edinburgh, Scotland
:
Scottish Council for Research in Education Centre
;
2000
8
Taras
H
.
Nutrition and student performance at school.
J Sch Health
.
2005
;
75
(
6
):
199
213
[PubMed]
9
Stewart-Brown
S
.
What is the Evidence on School Health Promotion in Improving Health or Preventing Disease and, Specifically, What is the Effectiveness of the Health Promoting Schools Approach?
Copenhagen, Denmark
:
World Health Organization Regional Office for Europe
;
2006
. Available at: www.euro.who.int/__data/assets/pdf_file/0007/74653/E88185.pdf. Accessed March 25, 2018
10
Busch
V
,
de Leeuw
JR
,
de Harder
A
,
Schrijvers
AJ
.
Changing multiple adolescent health behaviors through school-based interventions: a review of the literature.
J Sch Health
.
2013
;
83
(
7
):
514
523
[PubMed]
11
World Health Organization
.
Promoting Health Through Schools: Report of a WHO Expert Committee on Comprehensive School Health Education and Promotion
.
Geneva, Switzerland
:
World Health Organization
;
1997
12
Langford
R
,
Bonell
CP
,
Jones
HE
, et al
.
The WHO health promoting school framework for improving the health and well-being of students and their academic achievement.
Cochrane Database Syst Rev
.
2014
;(
4
):
CD008958
[PubMed]
13
Lister-Sharp
D
,
Chapman
S
,
Stewart-Brown
S
,
Sowden
A
.
Health promoting schools and health promotion in schools: two systematic reviews.
Health Technol Assess
.
1999
;
3
(
22
):
1
207
[PubMed]
14
St Leger
L
,
Young
IM
.
Creating the document ‘promoting health in schools: from evidence to action’.
Glob Health Promot Educ
.
2009
;
16
(
4
):
69
71
[PubMed]
15
Vince Whitnam
C
,
Aldinger
CE
.
Case Studies in Global School Health Promotion
.
New York, NY
:
Springer Publishing Co
;
2009
16
Rajaraman
D
,
Sachin
S
,
Patel
V
, eds.
School Health Promotion: Case Studies From India
.
New Delhi, India
:
Byword Publications
;
2015
17
Ademokun
OM
,
Osungbade
KO
,
Obembe
TA
.
A qualitative study on status of implementation of school health programme in South Western Nigeria: implications for healthy living of school age children in developing countries.
Am J Educ Res
.
2014
;
2
(
11
):
1076
1087
18
Yoshimura
N
,
Jimba
M
,
Poudel
KC
, et al
.
Health promoting schools in urban, semi-urban and rural Lao PDR.
Health Promot Int
.
2009
;
24
(
2
):
166
176
[PubMed]
19
Ippolito-Shepherd
J
,
Cerqueira
MT
,
Ortega
DP
.
Health-promoting schools regional initiative of the Americas [in Spanish].
Promot Educ
.
2005
;
12
(
3–4
):
220
229, 180
20
St Leger
L
.
Schools, health literacy and public health: possibilities and challenges.
Health Promot Int
.
2001
;
16
(
2
):
197
205
[PubMed]
21
Ministry of Health and Family Welfare, Government of India
. Guidelines on the school health programme. 2009. Available at: http://mdm.nic.in/files/school health programme/guidelines_shp_29th_jan_09-final_final.pdf. Accessed January 24, 2019
22
Shinde
S
,
Pereira
B
,
Khandeparkar
P
, et al
.
The development and pilot testing of a multicomponent health promotion intervention (SEHER) for secondary schools in Bihar, India.
Glob Health Action
.
2017
;
10
(
1
):
1385284
[PubMed]
23
Rajaraman
D
,
Travasso
S
,
Chatterjee
A
, et al
.
The acceptability, feasibility and impact of a lay health counsellor delivered health promoting schools programme in India: a case study evaluation.
BMC Health Serv Res
.
2012
;
12
:
127
[PubMed]
24
Gorkhaland Territorial Administration
. Gorkhaland Territorial Administration profile. Available at: www.gta-darjeeling.org/node/285. Accessed March 25, 2018
25
Rai
RP
. Small farmers: locating them in “Darjeeling Tea”. In:
Bag
N
,
Bag
A
,
Palini
LMS
, eds.
Tea Technological Initiatives
.
New Delhi, India
:
New India Publishing Agency
;
2017
:
19
38
26
Office of the Registrar General and Census Commissioner, India
.
Census of India, 2011
.
New Delhi, India
:
Ministry of Home Affairs, Government of India
;
2011
27
Craig
P
,
Dieppe
P
,
Macintyre
S
,
Michie
S
,
Nazareth
I
,
Petticrew
M
;
Medical Research Council Guidance
.
Developing and evaluating complex interventions: the new Medical Research Council guidance.
BMJ
.
2008
;
337
:
a1655
[PubMed]
28
Moore
GF
,
Audrey
S
,
Barker
M
, et al
.
Process evaluation of complex interventions: Medical Research Council guidance.
BMJ
.
2015
;
350
:
h1258
[PubMed]
29
Hemming
K
,
Haines
TP
,
Chilton
PJ
,
Girling
AJ
,
Lilford
RJ
.
The stepped wedge cluster randomised trial: rationale, design, analysis, and reporting.
BMJ
.
2015
;
350
:
h391
[PubMed]
30
Des Jarlais
DC
,
Lyles
C
,
Crepaz
N
;
TREND Group
.
Improving the reporting quality of nonrandomized evaluations of behavioral and public health interventions: the TREND statement.
Am J Public Health
.
2004
;
94
(
3
):
361
366
[PubMed]
31
World Education Forum
. Focusing resources on effective school health: a FRESH start to enhancing the quality and equity of education. 2000. Available at: https://www.unicef.org/lifeskills/files/FreshDocument.pdf. Accessed October 24, 2017
32
World Health Organization
.
Health Promoting Schools: A Framework for Action
.
Manilla, Philippines
:
World Health Organization, Regional Office for the Western Pacific
;
2009
33
Goldman
N
,
Vaughan
B
,
Pebley
AR
.
The use of calendars to measure child illness in health interview surveys.
Int J Epidemiol
.
1998
;
27
(
3
):
505
512
[PubMed]
34
Hashizume
M
,
Armstrong
B
,
Hajat
S
, et al
.
Association between climate variability and hospital visits for non-cholera diarrhoea in Bangladesh: effects and vulnerable groups.
Int J Epidemiol
.
2007
;
36
(
5
):
1030
1037
[PubMed]
35
Azage
M
,
Kumie
A
,
Worku
A
,
C Bagtzoglou
A
,
Anagnostou
E
.
Effect of climatic variability on childhood diarrhea and its high risk periods in northwestern parts of Ethiopia.
PLoS One
.
2017
;
12
(
10
):
e0186933
[PubMed]
36
World Health Organization
.
WHO Child Growth Standards: Length/Height-for-Age, Weight-for-Age, Weight-for-Length, Weight-for-Height and Body Mass Index-for-Age: Methods and Development
.
Geneva, Switzerland
:
World Health Organization
;
2006
37
Chakraborty
NM
,
Fry
K
,
Behl
R
,
Longfield
K
.
Simplified asset indices to measure wealth and equity in health programs: a reliability and validity analysis using survey data from 16 countries.
Glob Health Sci Pract
.
2016
;
4
(
1
):
141
154
[PubMed]
38
Jensen
R
.
The (perceived) returns to education and the demand for schooling.
Q J Econ
.
2010
;
125
(
2
):
515
548
39
Mendelsohn
AL
,
Huberman
HS
,
Berkule
SB
,
Brockmeyer
CA
,
Morrow
LM
,
Dreyer
BP
.
Primary care strategies for promoting parent-child interactions and school readiness in at-risk families: the Bellevue project for early language, literacy, and education success.
Arch Pediatr Adolesc Med
.
2011
;
165
(
1
):
33
41
[PubMed]
40
Lumeng
JC
,
Kaciroti
N
,
Frisvold
DE
.
Changes in body mass index z score over the course of the academic year among children attending Head Start.
Acad Pediatr
.
2010
;
10
(
3
):
179
186
[PubMed]
41
Bann
CM
,
Wallander
JL
,
Do
B
, et al
.
Home-based early intervention and the influence of family resources on cognitive development.
Pediatrics
.
2016
;
137
(
4
):
e20153766
[PubMed]
42
Langford
R
,
Bonell
C
,
Komro
K
, et al
.
The health promoting schools framework: known unknowns and an agenda for future research.
Health Educ Behav
.
2017
;
44
(
3
):
463
475
[PubMed]
43
World Health Organization
.
Global School Health Initiatives: Achieving Health and Education Outcomes. Report of a Meeting, Bangkok, Thailand, 23–25 November 2015
.
Geneva, Switzerland
:
World Health Organization
;
2017
44
Ministry of Health and Family Welfare, Government of India
. National health policy, 2017. 2017. Available at: https://www.nhp.gov.in//NHPfiles/national_health_policy_2017.pdf. Accessed October 30, 2018
45
The Partnership for Child Development, Imperial College of London
. Situational analysis. A summary of school health in India and in four states: Andhra Pradesh, Delhi, Gujarat, and Tamil Nadu. 2013. Available at: www.schoolsandhealth.org/Shared Documents/Downloads/Situation Analysis A Summary of School Health in India and in Four States.pdf. Accessed March 25, 2018
46
Santhya
KG
,
Jejeebhoy
SJ
.
Young People’s Sexual and Reproductive Health in India: Policies, Programmes and Realities
.
New Delhi, India
:
Population Council, Regional Office for South and East Asia
;
2007
47
Ciapponi
A
,
Lewin
S
,
Herrera
CA
, et al
.
Delivery arrangements for health systems in low-income countries: an overview of systematic reviews.
Cochrane Database Syst Rev
.
2017
;(
9
):
CD011083
[PubMed]
48
Jeet
G
,
Thakur
JS
,
Prinja
S
,
Singh
M
.
Community health workers for non-communicable diseases prevention and control in developing countries: evidence and implications.
PLoS One
.
2017
;
12
(
7
):
e0180640
[PubMed]
49
Lewin
S
,
Munabi-Babigumira
S
,
Glenton
C
, et al
.
Lay health workers in primary and community health care for maternal and child health and the management of infectious diseases.
Cochrane Database Syst Rev
.
2010
;(
3
):
CD004015
[PubMed]
50
Brenner
JL
,
Barigye
C
,
Maling
S
, et al
.
Where there is no doctor: can volunteer community health workers in rural Uganda provide integrated community case management?
Afr Health Sci
.
2017
;
17
(
1
):
237
246
[PubMed]
51
Patel
V
,
Weiss
HA
,
Chowdhary
N
, et al
.
Effectiveness of an intervention led by lay health counsellors for depressive and anxiety disorders in primary care in Goa, India (MANAS): a cluster randomised controlled trial.
Lancet
.
2010
;
376
(
9758
):
2086
2095
[PubMed]
52
Shinde
S
,
Weiss
H
,
Pereira
B
, et al
.
Impact of a school health promotion intervention on school climate and health outcomes for adolescents in Bihar, India: a cluster randomised trial.
In:
11th World Congress on Adolescent Health
;
October 27–29, 2017
;
New Delhi, India
53
International Institute for Population Sciences
;
ICF
.
National Family Health Survey (NFHS-4), 2015-16: India
.
Mumbai, India
:
International Institute for Population Sciences
;
2017
54
Schmidt
WP
,
Arnold
BF
,
Boisson
S
, et al
.
Epidemiological methods in diarrhoea studies–an update.
Int J Epidemiol
.
2011
;
40
(
6
):
1678
1692
[PubMed]
55
Melo
MC
,
Taddei
JA
,
Diniz-Santos
DR
,
May
DS
,
Carneiro
NB
,
Silva
LR
.
Incidence of diarrhea: poor parental recall ability.
Braz J Infect Dis
.
2007
;
11
(
6
):
571
579
56
Feikin
DR
,
Audi
A
,
Olack
B
, et al
.
Evaluation of the optimal recall period for disease symptoms in home-based morbidity surveillance in rural and urban Kenya.
Int J Epidemiol
.
2010
;
39
(
2
):
450
458
[PubMed]

Competing Interests

POTENTIAL CONFLICT OF INTEREST: Dr Matergia and Ms Matergia serve as directors for Broadleaf Health and Education Alliance, which funded the development and implementation of the intervention; Dr Matergia is also a former employee of the implementing agency, Darjeeling Ladenla Road Prerna, where he led the initial design of the Comprehensive Health and Hygiene Improvement Program; Ms Giri and Ms Thapa are current employees of Darjeeling Ladenla Road Prerna; and Mr Ferrarone, Ms Khan, and Dr Simões have indicated they have no potential conflicts of interest to disclose.

FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.

Supplementary data