Positive deviance study of malnutrition

Published Date: 23 Mar 2015

Introduction

Malnutrition is the underlying cause of every one out of two deaths in children under 5 years of age.

Bryce J, Boschi-Pinto C, Shibuya K, Black RE, and the WHO Child Health Epidemiology Reference Group. WHO estimates of the causes of death in children. Lancet 2005; 365: 1147-52.

It is a largely preventable cause of over a third-3.5 million- of all child deaths. Four-Fifths of undernourished children live across 4 regions-Africa, Asia, Western Pacific and the Middle East-. These are high priority nations for action. The first of the millennium development goals was "to half between 1990 and 2015 the proportion of people who suffer from hunger."

U.N.Mellinium Project 2005. Halving hunger:It can be done. London and Sterling, VA:Task force on hunger, 2005.

Nutrition is a neglected aspect of child health which is not justifiable as we know that it is a major risk factor for disease.

Lopez AD, Mathers CD, Ezzati M, Jamison DT, Murray CJL. Global and regional burden of disease and risk factors, 2001:systematic analysis of population health data. Lancet 2006: 367:1747-57.

Inappropriate feeding practices and their consequences are major obstacles to sustainable socioeconomic development and poverty reduction. Governments will be unsuccessful in their efforts to accelerate development in any significant long-term sense until optimal child growth and development, especially through appropriate feeding practices, is ensured.

WHO Global Strategy on Infant & Young Child Feeding Report of the Secretariat 55th World Health Assembly, April 2002(5)

The indicator by which progress is measured is the prevalence of underweight in children under five, specifically, the percentage of children aged 0-59 months who fall below minus two standard deviations from the median weight for age of the standard reference population.

http://www.unicef.org/progressforchildren/2006n4/index_undernutrition.html

World Bank Report on 11th May 2006, there is a dominant focus on food supplementation that detracts from aspects that are more crucial for improving child nutritional outcomes, such as improving mothers' feeding and caring behavior - For example, only 40 percent of Indian mothers practice exclusive breastfeeding.

Also delivery of services is not sufficiently focused on the youngest children (under three), who could potentially benefit most from ICDS interventions - Growth-faltering starts during pregnancy, and approximately 30 percent of children in India are born with low birth weight, and by the age of two years most lifetime growth retardation has taken place, and is largely irreversible

The flowchart below depicts the interactions between underlying and immediate causes of malnutrition -

The Positive Deviance concept which forms the basis of this study focuses on two direct behavioural causes of malnutrition at the household level, namely inadequate dietary intake and disease prevention.

This study measures childhood malnutrition using the new growth standards that have been recently released by the World Health Organization. The new standards are based on children from Brazil, Ghana, India, Norway, Oman and the US and adopt a fundamentally prescriptive approach designed to describe how all children should grow rather than merely describing how children grew in a single reference population at a specified time.

Garza C, de Onis M. (for the WHO Multicentre Growth Reference Study Group). Rationale for developing a new international growth reference. Food Nutr Bull 2004; 25 (Suppl. 1): S5-14.

Magnitude of Malnutrition

Malnutrition is a widely prevalent problem in India and one of astonishing magnitude. According to the National Family Health Survey 3 (NFHS III, 2005-06), about a third of India's children are born underweight, about 44 percent of children under five are underweight, 48 percent are stunted, 20 percent are wasted and 70 percent are anaemic. NFHS II (1992-93), more than half (53%) of children below four years of age are under nourished. In 1998, 29.1% children between 1-5 years of age suffered from moderate and 12.3% from severe under nutrition. This shows only a gradual decrease in the prevalence of under nutrition in India over twelve years. Nutritional adequacy is one of the key determinants of the health and well being of the children. Under-nourishment not only retardsphysical developmentbut also hampers the learning and cognitive process, leading to sluggish educational, social and economic development, according to Sanjeev Kumar in his study - Malnutrition in Children of the Backward States of India and the ICDS Programme.

According to the NFHS III data, Tamil Nadu, although considered one of the better performing states, recorded to have nearly 35% of its rural children as undernourished. This value was certainly lower than the national rural average of 49%, but according to the NNMB survey based on the NCHS standards in 2001, there was a rise in the levels of severe malnutrition in both boys and girls in the under 5 age group in Tamil Nadu

Therefore the focus of this study is to investigate how some children living in the same surroundings escape the ill-effects of malnutrition and thrive in spite of poverty and limited resources, and use these practices to eliminate malnutrition among the rest.

The Positive Deviance Approach and its benefits

Positive Deviance is based on the fact that solutions to some community problems already exist within the community and need to be discovered. It is a "strength-based" or "asset-based" approach based on the belief that in every community there are certain individuals ("Positive Deviants") whose special, or uncommon, practices and behaviours enable them to find better ways to prevent malnutrition than their neighbours who share the same resources and face the same risks.

Through a dynamic process called the Positive Deviance Inquiry (PDI), these practices are discovered to contribute to a better nutritional outcome in under nourished child This intervention is designed to enable families with malnourished children to learn and practice these and other beneficial behaviours. The programme actively involves the mother and child in rehabilitation and learning in a home - like situation and work to enable the families to sustain the child's enhanced nutritional status at home. The typical session consists of nutritional rehabilitation and education over a twelve-day period followed by home visits.

The development of a community-based nutrition rehabilitation model called PD/Hearth was promoted by USAID and other international organizations such as UNICEF.

The Hearth approach

In the Hearth approach, caregivers of malnourished children practice new cooking, feeding, hygiene and caring behaviours shown to be successful for rehabilitating malnourished children. The selected practices come from both the findings of the Positive Deviance Inquiry and emphasis behaviours highlighted by public health experts. The Hearth session consists of nutritional rehabilitation and education over a twelve-day period followed by home visits.

The Hearth approach promotes behaviour change and empowers caregivers to take responsibility for nutritional rehabilitation of their children using local knowledge and resources. After two weeks of being fed additional high-calorie foods, children become more energetic and their appetites increase. Visible changes in the child, coupled with the "learning by doing" method, results in improved caregiver confidence and skills in feeding, child care, hygiene and health-seeking practices. Improved practices, regardless of mothers' education levels, enhance child growth and development. This approach successfully reduces malnutrition in the target community by enabling community members to discover the wisdom of Positive Deviant mothers and to practice this wisdom in the daily Hearth sessions.

Positive Deviance/Hearth is an effective tool in discovering the solutions from within.

Need for the Study

Maternal and child under nutrition is highly prevalent in low and middle income countries like ours, resulting in increased mortality and overall disease burden.

It is well recognised that among the basic and underlying causes of under nutrition include environmental and economic factors with poverty having a central role. Thus for a sustainable solution to this very common problem, there needs to be an ecological and holistic approach rather than the current and commonly used approach of the government giving supplementary feeds to those who are malnourished. The latter approach not only puts a financial strain on the government but also make the community members dependent on this external aid. Therefore this study concentrates on identifying affordable and sustainable solutions within a community which can be used to prevent undernutrition. This approach is also not resource hungry unlike the traditional approaches where the focus is on finding and fixing what is wrong and missing in the community rather than identifying what is already working and build on the strengths of existing healthy practices within the community that is protecting them from undernutrition even though they are faced with same constraints of resources. This approach is known as the Positive Deviance approach.

Knowledge shared through this approach not only changes behaviour but also changes how a community perceives malnutrition and their ability to change the situation. The Positive Deviant approach has been shown in other studies and projects to quickly eliminate malnutrition and through the sustainable new behaviour; the younger siblings have also received these benefits. Positive deviance is a successful approach to decrease malnutrition and has enabled hundreds of communities the world over to reduce and prevent malnutrition. This approach is also culturally acceptable which helps bring about change in our society.

However, there are only a few studies that use this concept to reduce malnutrition in India, whereas the potential here is very high. And , although Tamil Nadu has shown gains in terms of reduction of the problem of malnutrition, it still has a high percentage of malnourished children and therefore this study was undertaken to estimate prevalence in fourteen villages within the rural field practice area of the department of community medicine of PSG Institute of Medical Science and Research and explore the feasibility, sustainability and effectiveness of combating the problem by using the Positive Deviance concept and approach.

Tamil Nadu has a high prevalence of malnutrition. Recently, there has been a paradigm shift in the primary focus from "Management of Malnutrition" to "Prevention of Malnutrition". Towards achieving this, strategies and activities have been proposed under various components for the year 2006-07 with priority for greater attention on the health and nutritional status, by the Government of Tamil Nadu. One of the guiding principles suggested to help achieve "Malnutrition free Tamil Nadu", is effective nutrition intervention, and communication to bring about behavioural change. To help achieve this goal, this study was undertaken in fourteen villages under the rural field practice area of the department of community medicine of PSG Institute of Medical Science and Research using the Positive Deviance concept.

Objectives

• To estimate the prevalence of under 3yrs undernutrition in 14 villages of Vedapatti.
• To identify Positive Deviant practices in the community.
• To rehabilitate undernourished children identified in the most affected village.
• To reassess families at their homes after 6 months and ensure sustainability of Positive Deviant practices.

Review of literature

Classification of under nutrition

Under nutrition is defined as the outcome of insufficient food intake and repeated infectious diseases. It includes being underweight for one's age, too short for one's age (stunted), dangerously thin for one's height (wasted) and deficient in vitamins and minerals (micronutrient malnutrition) according to UNICEF. Low weight for age is termed as underweight.

Weight for age classifications are widely used in assessing nutritional status of children as it gives a picture of both acute and chronic onset malnutrition. The earlier classifications include Gomez, Indian Association of Paediatrics, and Welcome. These classifications use different reference standards, but the current recommended standards are the WHO references.

This study uses the WHO references to classify underweight. In a study done by Bridget Fenn and Mary E. Penny across three countries, fewer children were classified as underweight according to the WHO classification when compared to the NCHS reference standards.

Another study done by Marc-Andre Prost et al showed contrasting results. The WHO standards gave a prevalence of underweight 3.6 times higher early in infancy,0 - 4 months (6.1% against 1.7%) and half the estimated prevalence of the NCHS reference in the second half of infancy, 11 - 15 months(6.6% against 13.6%)

Implication of New WHO Growth Standards on Identification of Risk Factors and Estimated Prevalence of Malnutrition in Rural Malawian Infants

Marc-Andre´ Prost1*, Andreas Jahn1,2, Sian Floyd1, Hazzie Mvula2, Eleneus Mwaiyeghele2, Venance Mwinuka2, Thomas Mhango2, Amelia C. Crampin1,2, Nuala McGrath1,2, Paul E. M. Fine1, Judith R. Glynn1

In a study done by P.R. Deshmukh et al , in Anji, Maharashtra, the prevalence of underweight as assessed by WHO standards was significantly lower when compared with the assessment based on NCHS reference (p<0.01). But, WHO standards gave higher prevalence of severe underweight than NCHS reference though the difference was not statistically significant (p>0.05).

Newly Developed WHO Growth Standards : Implications for Demographic Surveys and Child Health Programs P.R. Deshmukh, A.R. Dongre, S.S. Gupta and B.S. Garg

Prevalence of under nutrition

The World Bank estimates that India is ranked 2nd with 47% after Bangladesh for the most number of children who suffer with malnutrition (in 1998). The prevalence of underweight children in India is among the highest in the world, and is nearly double that of Sub-Saharan Africa with dire consequences for mobility, mortality, productivity and economic growth.

According to the National family health survey -3 the percentage of under 3 children who were underweight in Tamilnadu were 33.2 % (31.3% in the urban areas and 34.8% in rural areas)

The National Nutrition Monitoring Bureau observed that in about 40% of the households, the intake of dietary energy by preschool children was inadequate. About 55% of the preschool age children were underweight, 52% were stunted and 15% were wasted.

In a study done by Bhanderi D et al, the prevalence of under weight (wt. for age below 2SD) was 43.67%.

"An epidemiological study of health and nutritional status of under five children in semi-urban community of Gujarat." Bhanderi D, Choudhary SK.

A study done to find the prevalence of malnutrition in Uttar Pradesh showed that the maximum over all prevalence of malnutrition was in the age group of 13-24 months. The maximum number of grade IV malnutrition was found in the age group 25-36 months.

Harishankar, Shraddha Dwivedi, S.B. Dadral, D.K. Walia, "Nutritional status of children under 6 years of age" Indian Journal of Preventive and Social Medicine. Vol. 35 No.3 & 4 2004

Methods of assessing nutritional status

Nutritional assessments can be done by both direct and indirect methods. The method used can result in a wide variation in the prevalence of malnutrition.

P. Mohanan et al in their study evaluated the efficacy of Body Mass Index, Mid-Arm Circumference and Weight-for-age in detecting Malnutrition in terms of sensitivity, specificity and predictive value and concluded that weight-for-age is the best indicator.

P.Mohanan, A.Kamath, B.Motha, M.Philip. Evaluation of Anthropometric Indices of Malnutrition in under-five children. Indian Journal of Public Health 1994 July-September;28 (3); 91-94.

A Comparison of Anthropometric Methods for Assessing Nutritional Status of Preschool Children in the Philippines was done to identify the most reliable anthropometricmeasurements that reflect nutritional status and the Dugdale's nutritional index, weight/height, was a reliable measurement for malnutrition. This is exceptionally useful when the age of the child is not know.

A Comparison of Anthropometric Methods for Assessing Nutritional Status of Preschool Children: The Philippines Study Marilyn D. Johnson, MS,William K. Yamanaka, PhDandCandelaria S. Formacion, MS

Factors associated with malnutrition

Gender

1. A case-controlled study conducted in a rural area in Tamil Nadu, India, on 97,000 children showed that female gender was a significant risk factor for malnutrition.

YALE JOURNAL OF BIOLOGY AND MEDICINE 70 (1997), pp. 149-160. Copyright C 1997. All rights reserved. A Case-Control Study of Maternal Knowledge of Malnutrition and Health-Care-Seeking Attitudes in Rural South India Kaori Saito, Joshua R. Korzenika, James F. Jekel and Sara Bhattacharji

2. Also another study in Tamilnadu on a total of 2954 children attending the TamilNadu Integrated Nutrition Project (TINP), showed that there was an association between female sex and malnutrition. In this study the malnourished children were referred to as negative deviants.

Use of Positive-Negative Deviant Analyses to Improve Programme Targeting and Services: Example from the TamilNadu Integrated Nutrition Project MEERA SHEKAR, JEAN-PIERRE HABICHT AND MICHAEL C LATHAM

3. Girls showed only a slightly higher level of prevalence of malnutrition in the study done by The Research and Special Studies Division of the Department of Census and Statistics, Sri Lanka
4. A study done in the slums of Chandigarh on 1286 preschool children found no significant gender difference. This study shows similar reports as our study in Vedapatti.

H.M. Swami, J.S.Thakur, S.P.S.Bhatia, Vikas Bhatia. Nutriotional status of preschool children in an ICDS block of Chandigarh. Journal og Indial Medical Association; 99(10): 554-556

5. The sex of the child was found to be significantly high in association to malnutrition. The percentage of female children that were malnurished was higher in school going children.

"Epidemiology of malnutrition in a rural field practice are of Navi Mumbai" By Sumedha Joshi and Santosh S. Walgankar... Indian Journal of Preventive and Social Medicine Vol 35, 1 and 2, 2004

Mother's Literacy

1. The Research and Special Studies Division of the Department of Census and Statistics, Sri Lanka, undertook a methodological study to investigate the prevalence of malnutrition in children under five years of age, and the factors contributing to such a situation. The study found that 25% of pre-school children that were undernourished had mother's who had not gone beyond the primary level of school education. The effected proportion dropped to 11% when the mother has had at least secondary education.
2. Mother literacy was found in this study to have a positive effect on bringing down the incidence of malnutrition in socio-economically backward villages around Agra.

"Positive Deviance determinants in Young Infants in rural Uttar Pradesh" by Vani Sethi1, Sushma Kashyap1, Veenu Seth1 and Siddharth Agarwal, Department of Foods and Nutrition, Lady Irwin College, New Delhi, India. Indian Journal of Pediatrics Volume 74, June 2007.

3. Sandip Kumar et al in a study among 600 under 5 children in West Bengal showed that illiteracy of both parents was a significant risk factor for malnutrition.

Sandip Kumar Ray, Anima Halder, Biswajit Biswas, Raghunath Mishra, Satish Kumar. Epidemiology of Under Nutrition. Indian Journal of Pediatrics 2001 November: 68:1025-1030.

Socio-economic status

This study in 2003 on 4187 children showed contrasting results in Nigeria. The southeast and southwest regions had large inequalities between the poor and the rich whereas northeast and northwest regions had a considerably small gap between the rich and the poor on malnutrition.

Using extended concentration and achievement indices to study socioeconomic inequality in chronic childhood malnutrition: the case of Nigeria Olalekan A Uthmancorresponding author1,2

A study done in 1000 under 5 children in Rajasthan showed that 82% of the malnourished children belonged to the socio economic classes 4 and 5.

"Nutritional disorders in rural Rajasthan" A.L.Soni1, R.N.Singh1and B.D.Gupta. Indian Journal of Peadiatrics, May, 1980,Vol 47: 199-202.

A case-control study of maternal knowledge of malnutrition and health-care-seeking attitudes in rural Tamilnadu, showed that socio-economic status was a stronger risk factor for malnutrition than health-care availability and health-care-seeking attitudes.

A case-control study of maternal knowledge of malnutrition and health-care-seeking attitudes in rural South India, Saito K,Korzenik JR,Jekel JF,Bhattacharji S.

Birth order

The study done by Vani Sethi et al also evaluated other factors that contribute to malnutrition in the society and found that third or earlier born infants escaped malnutrition

"Positive Deviance determinants in Young Infants in rural Uttar Pradesh" by Vani Sethi1, Sushma Kashyap1, Veenu Seth1 and Siddharth Agarwal21Department of Foods and Nutrition, Lady Irwin College, New Delhi, India. Indian Journal of Pediatrics Volume 74, June 2007

Another study in Uttar Pradesh observed that grade III malnutrition was absent in the children whose birth order was one and two. A significant rise in Malnutrition was found in the children of birth order IV and above.

Health and Population - Perspectives & Issues 4(2):106-112, 1981

A community based, cross-sectional study was conducted in the Mollasimla village of Hooghly district of West Bengal, showed a significantly higher proportion of malnutrition was found to be present among female children of higher birth order and those belonging to families with lower per capita income compared to the males.

Gender inequality in nutritional status among under five children in a village in Hooghly district, West Bengal. Dey I,Chaudhuri RN.

PROTEIN ENERGY MALNUTRITION IN CHILDREN - A CASEFOR THE NEED OF A PLANNED FAMILY Deoki Nandan*, J. V. Singh** and B. C Srivastava

A study done by Harishankar et al showed that the highest percentage of malnourished children was seen in the first born children (47.2%) and the least in children with birth order 3 and above (17.8%)

Harishankar, Shraddha Dwivedi, S.B. Dadral, D.K. Walia,"Nutritional status of children under 6 years of age" Indian Journal of Preventive and Social Medicine. Vol. 35 No.3 & 4 2004

Spacing

The same study in Uttar Pradesh, by Deoki Nandan et al also studied the relationship to spacing and malnutrition showed that there was a direct association between Protein energy Malnutrition and less spacing between sibilings.

Health and Population - Perspectives & Issues 4(2):106-112, 1981. Protein Energy Malnutrition In Children - A Case For The Need Of A Planned Family Deoki Nandan*, J. V. Singh** and B. C Srivastava

Studies using Positive Deviance concepts

The successful application of the PD approach has been documented in more than 41 countries in nutrition and a variety of other sectors from public health to education to business.

Positive deviance is not specific to nutrition practices, but can be used for many other behaviours.

Ref:http://www.positivedeviance.org/PD_Evaluation_Report_for_DEPKES_FINAL.pdf

1. Positive Deviance was used in two Colombian hospitals to combat the spread of MRSA Infection. All hospital staff up to the security guard involved themselves by reminding the visitors to practice good hand hygiene. As a result the infection rates have dropped down my more than 75% from 1.1 infections per 1000 patient days to less than 0.2 infections per 1000 patient days.

Ref: http://www.positivedeviance.org/projects/healthcare.html?id=49

2. In West Bengal, India, ICDS has undertaken pilots projects in the use of the PD approach in Nutrition and Child Care Program (NCCP) in 4 districts to improve the nutritional status of children under three years of age. The projects substantially decreased the number of malnourished children by promoting good care practices. The PD informed project enabled families to break the dependence on donated food, by identifying cheap locally available and bringing it daily to the NCC session to prepare and feed their malnourished children. Every month the malnourished child is weighed and in most cases, mothers find their children gaining weight between 100 and 600 gm.

Ref:http://www.positivedeviance.org/projects/nutrition.html?id=77

Ref:http://www.unicef.org/india/nutrition_1557.htm

3. In 1990 Save the Children initiated a PD program in Viet Nam to enable poor villages to address the pervasive problem of childhood malnutrition. At that time 60% of children under the age of 5 suffered from malnutrition in Viet Nam. The initial pilot project was in the first 4 villages. In each of the villages, six of the poorest families with well nourished kids were chosen and caretakers were questioned and observed. In every instance where a poor family had a well-nourished child, the mother or father was collecting tiny shrimps or crabs or snails (the size of one joint of one finger) from the rice paddies and adding these to the child's diet along with the greens from sweet potato tops. Although readily available and free for the taking, the conventional wisdom held these foods to be inappropriate, or even dangerous, for young children. Along with these food and atypically strict hand hygiene in 5 of the 6 PD households, other positive deviant behaviors emerged, involving frequency and method of feeding and quality of care and health-seeking behaviors. Through the PD inquiries, community members had discovered for themselves what it took for a very poor family to have a well-nourished child. Rehabilitation started as for two weeks every month, mothers or other caretakers would bring their malnourished children to a neighbor's house for a few hours every day. Together with the health volunteer, they would prepare and feed an extra nutritious meal to their children. This showed great success by reducing malnutrition by as much as 80%. The project was then applied in large scale reaching more than 2 million people and in 250 communities and sustainabily rehabiliteted 50,000 malnourished children under the age of 5. This is probably the best known and best documented large scale application of PD.

Ref: http://www.positivedeviance.org/projects/nutrition.html?id=105

4. In Nepal a PD project to cover more than 8000 children under 3 years of age from 15 very disadvantaged communities was started with an aim of reducing child malnutrition in a sustainable manner. Positive practices regarding child feeding, caring, health seeking and maternity care are identified from the poor families having well nourished children through the PDI and then are made accessible to the families with malnourished children through a "learning by doing" process.

Ref: http://www.positivedeviance.org/projects/countries.html?id=82

5. In the year 2000 a Positive Deviance Inquiry to identify specific behaviors and strategies that contribute to healthy pregnancy outcomes amongst poor women was conducted. The inquiry determined that mothers-in-law played a central role in assisting women in obtaining medical care. Low-income women with weight gain greater than 1.5 kg per month in the second trimester of pregnancy reported multiple antenatal care contacts, increased rest during pregnancy, and more consumption of meat and vegetables. These results were incorporated into a program for 200 women that resulted in a decrease in the prevalence of low birth weight.

Ref: http://www.positivedeviance.org/projects/public_health.html?id=117

The following is illustrative of the impact of Positive Deviance over the last 15 years:

Sustained 65 to 80% reduction in childhood malnutrition in Vietnamese communities, reaching a population of 2.2 million people. Significant reduction in childhood malnutrition in communities in 41 countries around the world. Reduction in neo-natal mortality & morbidity in Pashtun communities in Pakistan and minority communities in Vietnam with near universal adoption of protective behaviors and social change. Estimated 50% increase in primary school student retention in 10 participating schools in Missiones, Argentina.

Community intervention methods to combat under nutrition

Traditional nutrition interventions include growth monitoring, counselling and the provision of supplemental foods and micronutrients But over decades the weighing of children undertaken by several National health ministries has brought little or no change to the nutritional status. In fact, in a study done by Sridhar Seetharaman, in Uttar Pradesh and Rajasthan, it was found that the Mid Day Meal did not make any appreciable and significant impact on improving the nutritional status of the children.

Impact Of Mid Day Meal On The Nutritional Status Of School Going Children, Sridhar Seetharaman, NIRD, Hyderabad

Zulfiqar A Bhutta et al used a cohort model to study the interventions that affect maternal and child undernutrition and nutrition-related outcomes. The interventions included promotion of breastfeeding; strategies to promote complementary feeding, with or without provision of food supplements; micronutrient interventions; general supportive strategies to improve family and community nutrition. They found that these interventions could reduce stunting at 36 months by 36%; mortality between birth and 36 months by about 25%; and disability-adjusted life-years associated with stunting, severe wasting, intrauterine growth restriction, and micronutrient deficiencies by about 25%.

Prof Zulfiqar A Bhutta PhDa, Tahmeed Ahmed PhDb, Prof Robert E Black MDc, Prof Simon Cousens PhDd, Prof Kathryn Dewey PhDe, Elsa Giuglianif, Batool A Haider MDa, Prof Betty Kirkwood PhDd, Saul S Morris PhDd, Prof HPS Sachdevg, Meera Shekar PhDhand for the Maternal and Child Undernutrition Study Group, Lancet Volume 371, 8 February 2008, Pages 417-440

Reasons for age selection

Care is an important determinant of nutritional status. It determines the delivery of food and health care resources to the child by optimizing the existing resources to promote good health and nutrition in children.

Ramakrishnan U. UNICEF-Cornell colloquium on care and nutrition of the young child-planning. F Nutr Bull 1995; 16: 286-92.

The first two years of life are the "window of opportunity" to prevent early childhood undernutrition that causes largely irreversible damage. This is proved by the following studies across the world.

Study shows more benefit from reaching all at-risk children under age 2 with nutrition, versus starting assistance among the malnourished up to age 5

HAITI: Child Malnutrition Study Shows Early Prevention is Best Source:World Vision - USA. World Vision. Website:http://www.visionmundial.org/index_in.php

1. A study on psychosocial care among patrilineal and matrilineal households who lived in Jakarta, Indonesia was aimed at investigating psychosocial care practices of mothers and the relation of psychosocial care to the nutritional status of children aged 6-36 months. The study found that the proportions of underweight and wasting were significantly lower in patrilileal children compared to those of matrilineal children, since the father worked and the mother spent more time in care giving.

Psychosocial care and nutritional status of children aged 6-36 months among patrilineal and matrilineal households in Jakarta. Judhiastuty Februhartanty, Avita A Usfar, Ermita Dianawati, Duma O Fransisca, Airin Roshita and Umi Fahmida

2. A study in Guatemala about the economic effect of nutritional intervention in early childhood showed that villages that were earlier assigned a nutritional supplement called Atole for children between the age of 0 and 2 years now had adults earning upto 46% higher hourly wages. This suggests that investments in early childhood nutrition can be long-term drivers of economic growth.

Effect of a nutrition intervention during early childhood on economic productivity in Guatemalan adults. John Hoddinott, John A Maluccio, Jere R Behrman, Rafael Flores, Reynaldo Martorell

Supplementary meals

According to the National Programme of Nutritional Support to Primary Education, 2006 Mid-Day Meal Scheme Guidelines, by the Central Government, it has been also decided to raise the nutritional supplementation from the existing 300 calories and 8-12 grams of protein to minimum 450 calories and 12 grams of protein.

The Mid-Day Meal programme does not merely aim to provide a cooked meal, but one satisfying prescribed nutritional norms. At the same time, this has to be done within certain cost norms. Some suggestions given by the National Programme which would help in achieving the twin-objectives in India include preparation of 'Single Dish Meals' using broken wheat or rice and incorporating some amount of a pulse or soyabeans, a seasonal vegetable/green leafy vegetable, and some amount of edible oil will save both time and fuel besides being nutritious. Broken wheat pulao, leafy khicheri, upma, dal-vegetable bhaat are some examples of single dish meals.

National Programme of Nutritional, Support to Primary Education, 2006, (Mid-Day Meal Scheme), Ministry of Human Resource Development, Government of India.

Materials and methods

Study Area

The study was undertaken in fourteen villages under the rural health centre of Vedapatti, Coimbatore. The village with a high proportion of under 3 malntrition was taken as the intervention village.

Study Population

There were 797 children between the age of 6 to 36 months in the 14 villages.

Inclusion Criteria

Children between 6 to 36 months living in these 14 villages

Exclusion Criteria

Children who are not residents, but visiting these villages.

Children of families who have moved into the villages within a period of 1 month.

The age of the children in the study was obtained from birth records maintained by the PSG Rural Health Centre, Vedapatti which was later confirmed with the mothers during the inquiry.

Study Period

The initial baseline nutritional assessment was done in October 2008. In February 2009 the Positive Deviance Inquiry was conducted in Daliyur village and the Hearth session, the following month. The reassessment of the malnourished children and the mothers was done after a period of six months in September 2009.

Sample Size

The sample for the cross-sectional study was calculated using the formula

Sample size (n) = 4pq/d2

p = estimated prevalence of malnutrition (NFHS-III) = 35%)

q = (1-p) = 65%

d = allowable error = 5%

Using this formula sample size calculated was 375. The number of samples per village was divided proportional to the number of children in that village (Probability proportionate to size). The children were then selected by systematic random sampling method.

Study Design

The study was done in two phases.

The first phase was a cross sectional study in which 797 children who met the criteria were included in the study to find the prevalence of malnutrition.

The second phase of the study was an interventional study with a 6month follow up period in one village with a high rate of malnutrition.

Entry criteria for the intervention village was:

1. Prevalence of malnutrition among young children of at least 30%
2. Availability of affordable local foods
3. Availability of mothers as potential volunteers in the community

All the 18 children in this village attended the hearth sessions. Among them 7 were malnourished and one severely malnourished. A positive deviant inquiry conducted revealed that of these 18 children, 3 were positive deviant children based on the following criteria.

• Mother has more than one child
• Mother does not have a child malnourished 0-59m of age
• Mother does not have a severe or atypical social or health situation.
• Child is between 6 and 36m of age.
• Child is not sick.
• Child is not losing weight currently for more than two months consecutively.
• Mother is not well to do.

The questionnaire used for this enquiry was developed by Food for the Hungry International and tested by them is several countries. Some of the questions have been changed to suit local context. The enquiry is based on the following conceptual framework:

Steps in Data Collection

Step 1: Make a baseline nutritional survey

The team that accompanied the investigator was given a day's training on measuring weight using a Salter's hanging scale, to minimize inter observer variation. They were given the following checklist:

• Verbal consent taken.
• Made sure child was wearing minimal clothing at the time of measurement.
• Zeroed the scale after the cradle is attached.
• Weighing scale at eye level of the measurer.
• Cradle stopped shaking before weight was measured.
• No one was holding the cradle.
• Weight was measured to the nearest lower 100grams.
• Measurer immediately filled the questionnaire after the measurement was taken.
• Checked the birth record to confirm DOB.
• Thanked the Mother.

All children under the Vedapatti Centre's Field Practice area, between the age group of 6 - 36 months were weighted and their weight for age was analysed using the WHO Anthro version 2.

The nutrition baseline assessment not only identifies malnourished children but also serves as a community mobilization tool. Dhaliyur village with a high prevalence of malnutrition (38.89%) was identified as our intervention village

The guidelines were adapted from How to Weigh and Measure Children: Assessing the Nutritional Status of Young Children in Household Surveys, UN Department of Technical CoOperation for Development and Statistical Office, 1996 (2) and Athropometric Indicators Measurement guide, Bruce Cogill, Food and Nutrition Technical Assistance Project, 2001 (3).

Step 2: Conducting the Positive Deviant Inquiry

Initially a pilot study was done in 2 households in separate villages using the modified Food for the Hungry Questionnaire and few modifications were made. It was inferred that morning time was ideal to conduct the inquiry as it was time consuming and evenings were mostly spent with the family and on other leisure activities. The questionnaire was translated in to the local language and the inquiry was conducted in Tamil. It was 86 questions in all and took slightly over 2 hours per household.

The positive deviant family should meet the following criteria,

• Poor family (low income)
• Normal nutritional status of child
• Minimum of two children (must be close to average family size)
• Family should be representative of geographical and social groups living in the village
• No severe health problems
• PD family must belong to "mainstream community"
• Head of household should have same occupation as the majority of villagers
• Must have access to same resources as others in the community
• Family is found in the identified minority (if the program targets minority communities only)
• Gender of PD child can be a criterion in gender biased cultures

After obtaining informed consent from all the 18 Mothers in Daliyur village, the semi structured interview observed three basic categories of behaviours:

1. Feeding Practices
2. Care Giving Practices
3. Health Seeking Practices

The exhaustive questionnaire also probed in to areas like socio economic factors, marital status of the caregiver, environmental factors and hygiene practices. As this inquiry was conducted in the morning hours during which time the mother was engaged in household chores, the interviewer also observed several aspects of hygiene, food preparation, food storage and household environment.

Once the home visits were conducted the behaviours that might positively impact a child's nutritional status were identified and grouped as follows.

Step 3: Conducting Hearth sessions:

The Hearth sessions are designed following the Positive Deviance inquiry, as it guides the decisions on which foods to prepare and which Positive Deviance behaviour to promote within the community. These Hearth Sessions are conducted for two weeks in a row and last for usually last for two hours in the mornings. Each session consists of:

• Deworming (on first day)
• Weighing the children on first and fourteenth day
• Setting up the cooking station
• Hand washing
• Educational session and appetizer distribution
• Feeding of energy rich, calorie dense food.

The supplemental meal required to rehabilitate the children should provide a minimum of 300 calories and 8 to 12 grams of protein per day as per the Government of India recommendation. This requirement is slightly higher according to WHO with 600 to 800 calories each day with 25 to 27 grams of protein. This study considers a minimum requirement of 450 calories and 12 grams of protein per day per child.

The menus include a nutritious snack during the educational session which could be either a fruit or nuts. The foods are made from locally available and affordable ingredients. It is ensured that all food groups are present in each meal so that the child will receive a balanced meal.

The meal schedule for the two week hearth session is included in Annexure 2

The educational session consists of messages that lead to better health and nutrition, focusing on feeding, hygiene, child care practices and important health seeking practices reflected in the Positive Deviant inquiry.

Statistical Analysis

In this study undernutrition is classified based on the WHO Child Growth Standards collected during the WHO Multicentre Growth Reference Study (MGRS) in 2004. A key characteristic of the new standard is that it makes breastfeeding the biological "norm" and establishes the breastfed infant as the normative growth model.

The data was analysed using the WHO Anthro version 2 developed to facilitate application of the WHO Child Growth Standards in monitoring growth and motor development in individuals and populations of children up to 5 years of age.

In this database, weight-for-height is interpreted by using the Z-score classification system. The Z-score system expresses the anthropometric value as a number of standard deviations or Z-scores below or above the reference mean or median value. WHO recommends expressing anthropometric indices in terms of z scores of Standard deviation units as this would allow cross comparison between indices. Children who fall under the 2SD below the reference median is considered undernourished and below 3SD as severely undernourished.

The Chi square test was used to measure the strength of association between the variables under study. The t-test was used to compare means Repeat measure Analysis was used to measure the change in positive deviant practices over a period of six months.

Ethical Considerations

Ethical clearance was obtained from the institutional ethics committee before the study was started. (Annexture5)

Parental agreement for investigation and intervention was obtained by informed consent from parents of all the children who attended the Hearth sessions.. (Annexture7)

Operational Definitions

Care Giver

The person who is most directly involved in the care of the child. The caregiver may be a mother, grandmother, father, or older sibling. While this manual sometimes refers to the mother instead of the caregiver, it is important to realize that the caregiver can be anyone in the child's life, and it is this primary caregiver who should be invited to the Hearth sessions.

Hearth Session

A 12-day series of sessions designed to rehabilitate malnourished children and teach Positive Deviant practices and behaviours. Located in a home setting, caregivers and volunteers prepare an extra energy-rich calorie-dense supplemental meal or snack to feed the malnourished children. Caregivers prepare Positive Deviant foods and practice other positive childcare behaviours.

Positive

Something that is working or something that people are doing right. A positive behavior is utilizing locally available resources, instead of "special" resources that are unavailable to all in the community. Finding positive behaviours focuses on identifying success instead of failure.

Positive Deviant Behavior

An uncommon and demonstrably successful practice.

Positive Deviant Food

A specific, nutritious food that is used by the positive deviants in the community. This food is affordable and available to all.

Positive Deviant Inquiry

A survey tool used to discover the positive deviant person's successful or desired practices. A community's self-discovery process in which they witness practices of neighbours with healthy and well nourished children. An observation of those children that thrive under common and ordinary conditions. Includes observation of these children's families and their positive coping mechanisms that can be replicated within the community.

Positive deviant person

A person whose special practices or behaviours enable him/her to overcome a problem more successfully than his/her neighbours who have access to the same resources and share the same risk factors. In the context of malnutrition, a PD child is a well-nourished child who is part of a poor family (according to village standards).

Z score

Z-Score, also called a standard deviation score. A measurement of how far and in what direction a child's nutritional status deviates from the mean on the internationally recommended reference population of other children with the same age or height. Weights that are two standard deviations (<-2 Z scores) are considered underweight and three standard deviations (<-3 Z scores) are severely underweight.

Results

All the 14 villages were mapped and the prevalence was as shown below. Poochiyur, Onappalayam and Daliyur recorded the high rates of Malnutrition, whereas Kalikkanaickenpalayam and Onnampalayam had the highest rates of severe Malnutrition.

Prevalence of malnutrition in study area

Demographic Profile of the Intervention Village (Dhaliyur)

Dhaliyur was chosen as the intervention village based on the high proportion of malnourished in the village

Severity of Malnutrition in Dhaliyur Village

The total Malnutrition was 38.89%

Gender and malnutrition

More female children were malnourished than males (22.35% vs. 20.39%)

However there was no statistically significant difference in malnutrition in gender (p = 0.64317)

Socio-Economic Class and malnutrition

Malnutrition was higher among the lower socio economic status Grade I to Grade V (33.33%, 18.75%, 22.58%, 18.37% and 4.55% respectively)

A chi square for trend revealed that there is a statistically significant correlation between socio economic status and malnutrition (X2 = 4.457; p< 0.05)

Mother's literacy and malnutrition

Proportion of malnourished were highest among illiterate mothers (26.67%)

However no statistically significant correlation was found between mother's literacy status and malnutrition (X2 for trend= 1.410 p = 0.23502)

Spacing and Malnutrition

Malnutrition was less in children born after the recommended period of spacing of two years.

There was no significant correlation between birth spacing and malnutrition established in this study. (X2 = 2.70 p = 0.10037 (>0.05)

Birth order and Malnutrition

Although malnutrition was least observed in the first born children, there was no significant correlation between birth order and malnutrition X2 = 1.885 (for trend) and p= 0.16973 (>0.05)

Faulty Practices observed in Dhaliyur

Feeding Practices

Children not breastfed immediately after birth. Colostrum discarded

Children not fed during periods of illness as it was believed that it could not be digested during those times.

High protein foods (egg, meat, etc.) were considered at heat producing in the body.

Children shared their plates with their older siblings and therefore the amount of food consumed by them could not be analyzed.

As they had many children, the families could not afford to buy health foods like fruits and vegetables.

Care Giving Practices

Many caregivers and therefore no continuous monitoring of food and hygiene

Both parents do not spend much time with the child even when sick

Health Seeking Practices

Over the counter medication administered to the children.

Poor personal hygiene observed.

No record of deworming regularly.

Gain in weight after Hearth Session

All the Children in the intervention gained weight.

There is a statistically significant difference in the mean weights between Day 1-Day 14(Hearth Session) (p<0.001) and so also a week after Hearth Session [Day 21] (p<0.001)

Assessment after 6 months after Hearth Session

There is a statistically significant difference in the mean weights between Day 1-Day 120 (p<0.001)

REPEAT MEASURE ANALYSIS TO SHOW INCREASE IN WEIGHT OVER A PERIOD OF SIX MONTHS

Initially a repeated measures of ANOVA was done, where we found that sphericity assumption was violated at 0.05 level. Hence the unstructured repeated measure analysis was done to evaluate the gain in weight over time. In this method the summary statistics was the slope of the regression line for each subject. Statistical methods were then used to test whether the mean of the derived measure differs from zero. This approach was recommended as the repeated measurements were irregularly spaced. Assuming that the estimated slopes are approximately normally distributed the sample t-test yields 14 degrees of freedom as 15 subjects were studied. The two-sided p value (p<0.001) indicates that the mean slope is significantly different from zero, hence concluded that the weight gain was significantly increased after the intervention.

Mean slope = 0.45133

SD = 0.0521781

T value = ?15 x (0.45133) / 0.0521781 = 33.50

P < 0.001

Charles S. Davis "Statistical methods for the analysis of repeated measurements" Springer-Verlag New York.

Analysis of the change in behavior after Hearth Sessions:

The concept used to bring about change in the Positive Deviance method is to change behavior, which in turn would change attitude and therefore increase knowledge. This is contrary to the conventional methods where the focus is to increase knowledge which later may bring about a change in behavior.

The practices that were observed were categorized under Good feeding, Good care giving and Good Health Seeking practices. The following table shows the measured change in practices over the study period.

Discussion

While an experience like this on a small sample of eighteen for a short duration does not represent a statistically significant evidence of results, however, it does help us to understand the psychosocial environment that effects behaviour change and the valuable role of self efficacious PD mothers/family members as counsellors.

Summary

Recommendations

More research initiatives with a larger sample size for longer duration would be of utmost value to reveal the determinants behind adoption, maintenance, and dropout of the Positive Deviance behaviours once tried.

Limitations of the study

For practices and behavior to set in, longer periods of observation and repeated reinforcements are essential. Given the constraints of the study period, this could not be done.

The gain in weight recorded after a period of 6 months cannot fully be attributed to the Hearth sessions. Ideally a control group would have helped in comparison and establishing a cause-effect relationship.

In this study the mothers of malnourished children knew that their children were nutritionally inferior to the rest of the children in the community. This could have been avoided as the objective was not to find superior mothers but to find identify their superior behaviors.