P1-322 Statistical modelling in the prediction of kala-azar in Nepal

BackgroundKala-azar (Visceral Leishmaniasis) is endemic in 12 Terai districts of Nepal. More than 5.6 million people living in these districts are believed to be at risk of developing Kala-azar. The objective of this study was to predict future trends in Kala-azar using available data.MethodsWe carried out a retrospective study using data collected from the Health Ministry records of Nepal, between 2003 and 2007. Descriptive statistics and statistical modelling were used for analyses.ResultExcluding the constant term from the equation, the logarithmic model was the best fit for predicting Kala-azar cases (R2=0.85, p=0.01). The reported number of Kala-azar cases in 2007 differed from our projected number of cases by 68. Using a cubic equation it was estimated that there will be 2548 reported Kala-azar cases in Nepal by the year 2015.ConclusionThe number of Kala-azar cases in Nepal is predicted to increase. Curve fitting method could be an effective exploratory modelling technique for predicting cancer frequency and trends over the years. The Kala-azar control programmes should aim to reduce Kala-azar morbidity and prevent mortality by applying a primary healthcare approach which includes active community participation.