Estimating household healthcare expenses for treatment of locally important diseases by sample size constrained network sampling

Arijit Chaudhuri, Indian Statistical Institute
Shankar Dihidar, Indian Statistical Institute

We present a theory plus empirical report in estimating in a cost-effective way the household healthcare expenses for treatment of locally important diseases requiring institutional care within a compact area where getting access to remote households is difficult. Approaching traditionally we may encounter households containing no sufficient relevant data on healthcare expenses. So, we think it prudent to apply ‘Network Sampling Technique’ to get sampled households rich in relevant information. For this we start selecting various health centres. On gathering addresses of in-patients treated there within a specified time period we may locate households within the geographical area specified for the enquiry relatively easily. But an inherent hazard in this procedure is that total sample size may be extraordinarily large. So, to keep within our budget, with a self-inflicted bound on total sample size, we work out here necessary modifications on available literature on Network Sampling.