Validade e reprodutibilidade de medidas antropométricas, morbidade referida, consumo de alimentos e bebidas obtidos por inquérito telefônico

Abstract

The Surveillance System of Risk and protection Factors for Chronic Diseases by calling (VIGITEL) generates important information about risk factors and prevalence of chronic diseases. Through self-report information, allow us evaluate and follow representative samples, with lower cost and greater agility. However, for the planning of promotion and prevention actions is necessary that such information be tested for reliability and accuracy. The aim of this study was to evaluate the validity and reliability of the indicators of nutritional status, self-reported morbidity and consumption of food and beverages used by the system in the city of Belo Horizonte (BH), Minas Gerais. The reproducibility of the indicators of food and drink consumption and nutritional status was assessed through repeated interviews with intervals of 7-15 days (n=258). The validity of the indicators of food consumption was assessed using as gold standard the realization of three 24-hour records, 7-15 days after the original telephone interview (n=217). For the validity of the indicators of nutritional status and self-reported morbidity was used as the reference methods direct measurement of anthropometric data, blood pressure and biochemical examinations (n=67). The three samples were composed predominantly of female participants, with demographic, anthropometric and health characteristics, mostly similar to the total sample of VIGITEL 2009 in BH. The reproducibility of the indicators of consumption of foods and beverages showed results of regular to almost perfect agreement through the analysis of the kappa coefficient (0,34 a 0,86). The validity of these indicators showed a sensitivity close to or greater than 70% for the most part, except the indicator of consumption of soft drinks (47.6%) and alcohol consumption (45.2%). Specificity ranged from 29.5% (vegetable consumption) to 87.5% (alcohol consumption). For reliability and validity of indicators of nutritional status, we found high reliability between the subsequent reports of measures. By checking the agreement between self-reported measures and measured ones, there was substantial difference between them. However, when performed the classification of nutritional status with these indicators, revealed a great ability to detect overweight when it was actually present, observing high levels of sensitivity and positive predictive value. In the validity study of self-reported morbidity (hypertension, diabetes and dyslipidemia) were observed significant differences between the VIGITEL questionnaire and laboratory evaluation to identify people with the disease, resulting in low sensitivity values for all three indicators. The findings of this study support the use of surveillance strategy of risk factors system VIGITEL to identify trends and magnitude of the risk and protection factors related to food consumption and nutritional status in the population of Belo Horizonte, due to reasonable reproducibility and validity of its indicators. However, the morbidity indicators did not present results that confirm its accuracy in this population, which shows the need for further studies of validity of the system indicators throughout the country in order to provide greater accuracy of data generated, always considering its potential limitations.