PENERAPAN REGRESI LOGISTIK BINER UNTUK MENGUKUR RESIKO ANEMIA DENGAN STATUS GIZI IBU HAMIL
The prevalence of anaemia is an important health indicator for pregnant women. This paper demonstrates the use of logistic regression modelling techniques for an ordinal binary in terms of both response and predictor variables. By using 120 cross-sectional data to measure the attitude of pregnant women observed at three Community Healthcare Center (CHC) in the Serang District-Banten Province on September 2012, we investigated a factor predicting anaemia and tested whether changes in nutritional
... nges in nutritional status were associated with changes in anaemia status. The anaemia status threshold using haemoglobin concentration (g/dl) was measured by the Cyanmethemoglobin method, and the nutritional status was categorized by measuring Upper Arm Circumference (cm). This paper discusses the roles of nutritional status determining the risks of anaemia and compares the results obtained using the contingency analytical methods. The paper also demonstrates how the logistic regression modelling approach dealing with the odds ratio statistic can better explain the risks of anaemia determined by nutritional status found in the CHC studied. We found that anaemia status is affected by nutritional status (coef. 1,07; p. 0,066), and the results of contingency analysis reveals that anaemia satus is associated with nutritional status (chi-sq. 3,60; p. 0,058). According to odds ratio statistic of the logistic regression (OR 2,92), the risks of anaemia on pregnant women with cronic energy deficiency would be 2,92times higher than they are on normal nutritions.