Snowball Vs. House-to-House Technique for Measuring Annual Incidence of Kala-azar in the Higher Endemic Blocks of Bihar, India: A Comparison

BACKGROUND Visceral Leishmaniasis, commonly known as kala-azar, is widely prevalent in Bihar. The National Kala-azar Control Program has applied house-to-house survey approach several times for estimating Kala-azar incidence in the past. However, this approach includes huge logistics and operational cost, as occurrence of kala-azar is clustered in nature. The present study aims to compare efficiency, cost and feasibility of snowball sampling approach to house-to-house survey approach in capturing kala-azar cases in two endemic districts of Bihar, India. METHODOLOGY/PRINCIPAL FINDINGS A community based cross-sectional study was conducted in two highly endemic Primary Health Centre (PHC) areas, each from two endemic districts of Bihar, India. Snowball technique (used to locate potential subjects with help of key informants where subjects are hard to locate) and house-to-house survey technique were applied to detect all the new cases of Kala-azar during a defined reference period of one year i.e. June, 2010 to May, 2011. The study covered a total of 105,035 households with 537,153 populations. Out of total 561 cases and 17 deaths probably due to kala-azar, identified by the study, snowball sampling approach captured only 221 cases and 13 deaths, whereas 489 cases and 17 deaths were detected by house-to-house survey approach. Higher value of McNemar's χ² statistics (64; p<0.0001) for house-to-house survey approach than snowball sampling and relative difference (>1) indicates that most of the kala-azar cases missed by snowball sampling were captured by house-to-house approach with 13% of omission. CONCLUSION/SIGNIFICANCE Snowball sampling was not found sensitive enough as it captured only about 50% of VL cases. However, it captured about 77% of the deaths probably due to kala-azar and was found more cost-effective than house-to-house approach. Standardization of snowball approach with improved procedure, training and logistics may enhance the sensitivity of snowball sampling and its application in national Kala-azar elimination programme as cost-effective approach for estimation of kala-azar burden.