ABSTRACT
We believe that global health practice and evaluation operate with misleading assumptions about lack of reliability of small population-based health surveys (district level and below), leading managers and decision-makers to under-use this valuable information and programmatic tool and to rely on health information from large national surveys when neither timing nor available data meet their needs. This paper uses a unique opportunity for comparison between a knowledge, practice, and coverage (KPC) household survey and Rwanda Demographic and Health Survey (RDHS) carried out in overlapping timeframes to disprove these enduring suspicions. Our analysis shows that the KPC provides coverage estimates consistent with the RDHS estimates for the same geographic areas. We discuss cases of divergence between estimates. Application of the Lives Saved Tool to the KPC results also yields child mortality estimates comparable with DHS-measured mortality. We draw three main lessons from the study and conclude with recommendations for challenging unfounded assumptions against the value of small household coverage surveys, which can be a key resource in the arsenal of local health programmers.
