Use of positive-negative deviant analyses to improve programme targeting and services: example from the TamilNadu Integrated Nutrition Project.

This paper describes the characteristics of southern rural Indian children who grow best (positive deviants) and worst (negative deviants) as compared to median growers. A 100 each of positive and negative deviants and 120 median growers were selected after analysing the 12-month growth patterns (weight-for-age) of 2954 children enrolled in the TamilNadu Integrated Nutrition Project (TINP), a major health and nutrition intervention project covering nearly one million children in rural south India. The determinants of poor growth that have been left unaddressed by 6 years of TINP exposure are delineated to address the question of what more needs to be done. Further, the rationale for differential targeting of services to negative deviants and to median growers is discussed, as are the implications for programme evaluation. Data indicate that the next generation of projects targeted at the most needy (negative deviants) should address the issues of gender discrimination in childcare, of breastfeeding, of diarrhoeal disease, and of maternal empowerment. Such interventions will, however, not improve the growth of median growers in the direction of positive deviance. Instead, programmes targeted at the median growers need to support the hygienic use of nonbreast milk supplements. Improving family wealth will also improve the nutritional status of the median growers, but less so than for the negative deviants.