[Nutritional surveillance in Chilean schoolchildren: validity of the information]

Anthropometry is used to survey health and nutritional situation of the population. Therefore the quality of the information that is being used must be evaluated.

To estimate the agreement in weight, height and nutritional status in schoolchildren, comparing measurements made by teachers in schools and a standardized and supervised team of professionals.

Cross sectional study including 927 schoolchildren in 31 schools from 7 counties of Santiago. Schools were randomly chosen and the universe of children attending to first grade was measured. Weight, height and nutritional status collected by teachers and researches, were compared.

Total agreement for nutritional status reached 0.67, random-weighted Kappa was 0.40 and weighted Kappa, 0.42. Teachers tended to over diagnose under-nutrition and under diagnose overweight and obesity measuring 270 grams less than the qualified team (p <0.001) and 1.7 cm more in height (p <0.001), what is reflected in a difference of less than one point in body mass index (p <0.001). Discrepancies in height and body mass index were higher in extreme values.

There is a low agreement between the measurements taken by the research team and teachers. Even though there are discrepancies between measurements, high levels of overweight and low prevalences of stunting and underweight are kept, reflecting problems with exactitude, but not bias. Corrective actions to improve the quality of information, which should include training programs for teachers, instrument maintenance, supervision and verification system for data entry, are needed.