Effectiveness of a post-deworming education intervention to reduce soil-transmitted helminth infections and absenteeism in grade 5 school-children in a

Objectives

Primary

The primary objective is to determine the effectiveness of a school-based health hygiene education intervention to reduce re-infection by STH after a routine deworming intervention.
Secondary

The following secondary research objectives will also be addressed: i) to determine intervention effects on absenteeism rates; ii) to assess gender-specific determinants of STH infection, re-infection and absenteeism rates; iii) to measure STH species-specific cure rates and egg reduction rates; and iv) to examine behavioural and social determinants of STH infection, re-infection and absenteeism rates.
Methods

Study design

A prospective cluster-randomized controlled trial design will be used (Figure 1). Schools will be randomly allocated to one of the two arms using a stratified randomization technique (by school size). Because the total number of students enrolled in Grade 5 varies from 20 to more than 200 per school, the stratified design will ensure a balanced proportion of children in each of the two arms. A simple baseline questionnaire will be administered to all Grade 5 students to collect information on gender-specific risk factors and to evaluate the level of knowledge regarding STH infections. Each child will be weighed and measured (for anthropometric measures), and a stool specimen will be collected and examined for STH eggs (to estimate STH infection rates and intensity). The health hygiene education intervention will begin immediately after the routine deworming program in schools in the intervention arm. Schools in the control arm will follow their standard curriculum. (These schools will be offered the health hygiene education intervention after the 4-month follow-up period.)

Two weeks after the deworming, an efficacy assessment will be undertaken in a sub-sample of 300 previously-infected children (ie. Stool specimens will be collected to estimate cure rates for each STH. This is needed to provide a correction factor in the data analyses.). At the end of the 4-month follow-up study period, the baseline questionnaire will be re-administered to all children in both arms, the children will be weighed and measured, and re-infection rates will be assessed by collecting a stool specimen from each child. Absenteeism rates, based on routine school records, will be assessed from baseline to the end of the follow-up period. An intention-to-treat analysis will be used to compare outcomes in the two arms.
Study population

Multidisciplinary and participatory workshops conducted to ascertain local health priorities in the urban slum of Belén demonstrated that intestinal parasites were among the top-five health priorities and that targeted health hygiene education interventions were required [1]. In fact, in a recent survey we conducted in Belén, we found 86% of Grade 5 school children infected by at least one STH species (Ascaris=60.4%, Trichuris=77.9%, and hookworm=21.2%) [2].

We have obtained current information on the number of children enrolled in Grade 5 in schools of Belén, Peru. For efficiency purposes, only schools with a minimum enrolment of 10 boys and 10 girls in Grade 5 will be considered eligible. Eighteen schools, totalling more than 1438 students, are therefore considered eligible to participate in this trial.

Inclusion criteria

• 
  Girls and boys enrolled in Grade 5 in one of the 18 eligible schools in Belén.
  
• 
  Informed written consent obtained from the parents or legal guardian of the child.
  
• 
  Verbal assent from child.
  

Exclusion criteria

• 
  Parents (or guardians) refusing participation of their child.
  
• 
  Child who refuses to participate.
  

Figure 1: Flowchart of the cluster-randomized controlled trial

Health hygiene education intervention

Lack of knowledge on the transmission and prevention methods of STH infections is common factor in affected population groups of Latin America [27]. Therefore, the specific objectives of the health hygiene education intervention are:

i) To implement a gender-specific health hygiene education strategy which will result in a sustainable intervention that encourages the pro-active role of both teachers and students.

ii) To encourage changes in attitudes and practices with the aim of keeping the level of parasite infection low, through increased knowledge.

iii) To identify modifiable gender-specific behavioural and social determinants which favour the acquisition and transmission of STH infections to inform prevention activities.
Workshops will be organized for and by Grade 5 teachers to train them on the health hygiene education strategy. The education material will be provided by the Pan American Health Organization (in Spanish) and will be adapted to the local culture. The health hygiene education intervention will take into account the gender-specific perceptions, values, social constructs, and beliefs of students and teachers relevant to STH infections in order to maximize behavioural changes. After the workshop, Grade 5 teachers will be able to implement the health hygiene education intervention in their classroom and engage their students. Bi-monthly visits with the teacher by our research team and once monthly with the class will be performed to monitor the application of the education program during the 4-month follow-up period. Specifically, the intention of the health hygiene education program is to modify key gender-specific modifiable behavioural determinants that have been shown to be associated with STH infections from our review of the literature:

i) Not using soap for hand washing [28-30]

ii) Not hand washing (no reference to the use of soap) [31-33]

iii) Eating unpeeled or unwashed fruits [31, 34]

iv) Walking barefoot [30, 33]

v) Open air defecation [33-35]

vi) General unhygienic behaviours [36]
Measurements

Weight of the child will be measured using a portable flat scale (seca 869, seca corp., Baltimore, MD) and a mobile stadiometer (seca 217, seca corp., Baltimore, MD) will be used to measure their height. Individual risk factors, potential confounding variables, level of knowledge on STH transmission, and behavioural determinants will be measured using an interviewer-administered questionnaire at baseline. After the 4-month follow-up period, the same questionnaire will be re-administered.

Stool specimens will be obtained in small plastic containers at baseline and at the end of the 4-month follow-up period by trained research personnel. Trained technologists will use the Kato-Katz method to examine the specimens for the presence, species and intensity of STH eggs [37]. This technique is the preferred method for assessment of the prevalence and intensity of intestinal parasitic infections in the field [4, 38]. The technologists will be blinded to the group assignment.

Data collection activities during fieldwork will be regularly supervised. Measurement bias will be kept to a minimum by ensuring calibration of study instruments (balance and scale), quality control assessments, and daily review of completed questionnaires. At each assessment, the coordinator will check all completed forms immediately after data collection at the end of each day. Entries on the questionnaires will be checked and if discrepancies or unclear answers appear, a discussion and explanation of the question and eligible answers will take place to avoid making errors the next day. Corrections will be made, if warranted. These types of errors should be kept to a minimum, as there will be a pre-testing phase with all interviewers to ensure correct data collection procedures and records. The consistency of the microscopic readings will be verified both during the examination and afterward as follows. The consistency of egg counting will be evaluated among the laboratory technologists using standard quality control methods [38]. The laboratory supervisor will read 10% of the slides of each microscopist without prior knowledge of the results. In the case of a discrepancy larger than 10%, the two readers will discuss the slides and further slides will be examined to avoid repeated errors. It should be noted that trained microscopists from our previous study will perform the examinations. They have an excellent performance record.
Sample size justification

Based on our survey of schools in Belén, 18 schools are eligible and have already demonstrated an interest in participating in the study. They have a total of 1438 students enrolled in Grade 5 (Table 2). Power calculations are based on the formula for logistic regression with a single binary covariate (i.e. the education intervention) and the Wald test was used as the basis for computations [39]. Based on our past parasitological Belén school surveys a response rate of 80.6% is expected [2]. A 5% loss to follow-up is also assumed during the 4-month period. Thus, we estimate a sample size of 1101 participating students, distributed in the 18 schools.

In addition, the power calculations must take into account within-school clustering. The intraclass correlation coefficient (ICC) represents the proportion of the total variance that is accounted for by between-cluster rather than within-cluster variation and can be used to calculate the Effective Sample Size (ESS). Because we could not find any reported ICC of within-school clustering of STH re-infection in the literature, we re-analysed the original database of Casapía et al. [2], which contained the parasitological results of 1074 Belén students, and calculated the within-school ICC of STH prevalence (as an estimation of ICC of re-infection rates). The ICC was calculated using STH prevalence as the outcome and the school as the grouping factor, while adjusting for age, sex, and the school ecological zone (floodplain or upland). The obtained ICC had a value of 0.028 which resulted in an ESS of 412 and an associated design effect of 2.7. Power was estimated assuming that 50% of children will be exposed to the intervention and that the re-infection rate in the control group will be 49.5%. The latter is based on the reported 3-month re-infection rate of Malaysian school-children treated with albendazole [40]. Thus, the proposed study has good power (80.3%) to detect a moderate effect (i.e., OR=1.75) and has an excellent power (93.4%) to detect a strong effect, corresponding to an OR of 2.00.

The efficacy of the deworming program will be determined using a sub-sample of 300 previously infected children. Assuming the most conservative value of a 50% cure rate, this sample size of 300 will result in precise species-specific cure rates estimated with a maximum 95% confidence interval of ± 5.7%. No within-school clustering is assumed for cure rates because variables affecting cure rates are intrinsic to the parasites and the hosts themselves, not the schools.

Table 2: Number of Grade 5 students enrolled in 2007 and 2008 in the 18 eligible schools located in the Belén district of Iquitos, Peruvian Amazon

School name	Nb. grade 5 students 2007	Nb. grade 5 students 2008	Nb. classrooms
Bagazan	42	49	2
Santo Cristo Bagazan	134	125	4
Diego Natal Juan	109	87	3
Ruy Guzmán Hidalgo	181	160	5
Participación	147	116	3
Enry Herve Linares Soto	26	24	1
Ramón Castilla	266	203	6
Sara Alicia Saberbein Pinedo	175	186	5
Sachachorro	65	56	2
San Lucas	104	113	3
Violeta Correa	47	53	2
Sagrada Familia	53	79	2
Isla Iquitos	34	32	1
San José	43	25	1
San Francisco	43	42	2
Nuevo Liberal	63	37	1
La Victoria de Jesús	33	31	2
Nuevo Campeón	25	20	1
Total	1590	1438

Planned statistical analyses

All children in both arms will be included in the analysis (i.e., an intent-to-treat analysis). Preliminary analyses will include descriptive statistics (means and standard deviations for continuous variables; frequencies for categorical variables) for all measurements in each intervention group separately. Gender-disaggregated data will be presented in tables, whenever pertinent.

Multivariate analyses will be used to compare the post-deworming (at 4-months) STH re-infection rates between the two groups. To account for the within-school clustering, hierarchical (random effects) logistic regression models will be used [41]. The model will include important confounding variables such as gender, sex, age, school ecological zone, as well as any baseline characteristics that may not be balanced between the two groups. Subgroup analysis will be performed for boys and girls separately as well as for each of the three helminth species studied (Ascaris, Trichuris, and hookworm).

All statistical tests will be two-sided and a 0.05 level of significance will be used throughout for determination of statistical significance. Point estimates of the intervention effect with 95% confidence intervals will be presented to demonstrate the precision of the study results.
Safety reporting

As this study is testing the impact of an educational intervention on selected non-invasive outcome measures, no adverse events or complications associated with this intervention are expected. We consider that the administration of the questionnaire, the health hygiene education intervention, and the collection of the stool specimen will have no safety effect.

With respect to the routine deworming program, it may be that the deworming pill has a minor transitory side effect. The proposed anthelmintic, albendazole, has an excellent safety record involving over billions of doses [7, 42]. This drug is considered very safe and its administration is not a concern. Mass deworming of school children with albendazole is recommended by WHO, UNICEF, and the World Bank, among other organizations. Because of its overwhelmingly benefits, in 2001, The World Health Assembly unanimously ratified the resolution that deworming should reach at least 75% of all school children in endemic areas by the year 2010. Nevertheless, we will ensure that the local health authorities are informed about our deworming activities. We have collaborated with them in the past and will continue to do so throughout this research.
Anticipated benefits

It is anticipated that all schools and students participating in the research will benefit from this study. Schools will benefit directly because the capacity-building workshops and the education intervention program will provide them with new knowledge and tools to address STH infections, and indeed, other health hygiene concerns. Students participating in the study will also benefit directly from the study because they will be given the opportunity of increased knowledge. School absenteeism is expected to be reduced, benefiting students, schools, families and their community. As our previous study demonstrated, 86% of grade 5 students in Belén are infected by at least one parasites species [2]. Participation in this study will therefore contribute to improved health status.
Potential risks

As the planned intervention will consist of a health hygiene education program, there are no associated risks. There are also no associated risks with the data collection procedures (questionnaires and provision of stool specimens).
Privacy and confidentiality

Participants and schools will be assigned an identification number for the duration of the study to ensure the confidentiality of any results. All original documents, including questionnaires and informed consent forms, will be kept (in Peru) in a locked cabinet and room. All electronic information, such as databases, will be stored on a password-protected computer in a locked room. Access to original documents and electronic information will initially be restricted to the Project Director, Principal Investigator and study coordinators only.
Details of trial team

The trial management on site in Iquitos (Peru) will be under the local and daily supervision of Dr. Martín Casapía (Project Director), together with the study coordinator, Salomé Chapiama. Ms. Chapiama is trained as a nurse and has collaborated on more than seven research projects with us. She is an excellent coordinator and has the confidence of the investigator team. The Canadian members of the research teams include Dr. Theresa Gyorkos (principal investigator) and her research coordinator, Mathieu Maheu-Giroux. M. Maheu-Giroux will also be present in Iquitos during the implementation phase and randomization phase of the trial. He holds two Master’s degrees from McGill University (Ecology and Epidemiology) and has already conducted research in Iquitos. Dr. Theresa Gyorkos is responsible for the overall study design including all methodological issues. Although involved at a distance, all study decisions will be taken in consultation with the entire team. Dr. Gyorkos has been working on a variety of research projects in this region of Peru over the last 8 years and has strong ties with local collaborators and health officials. Furthermore, she has already conducted a large RCT in this region of Peru. This is an international collaboration of Canadian and Peruvian researchers who have had extensive experience in conducting epidemiological research together.

The two major collaborating centres are the Research Institute of the McGill University Health Centre in Montreal (Canada) and the Asociación Civil Selva Amazónica in Iquitos (Peru).
Ethical oversight

The principal investigator, Project Director, study coordinators and all other research personnel will conduct the study in an ethical manner which is consistent with the international principles of good practice and will function according to institutional policies of the McGill University Health Centre and the Asociación Civil Selva Amazónica, governing human subjects’ research, applicable research guidelines and in compliance with the law. The research investigators understand that the study is subject to review and approval from both the Research Ethics Board in Canada and the Dirección de Salud de Loreto in Peru.