Assessing the validity of health facility-based data on insecticide-treated bednet possession and use: comparison of data collected via health facility and household surveys--Lindi region and Rufiji district, Tanzania, 2005.

Objective To assess the validity of health facility (HF)-based data on bednet and insecticide-treated bednet possession and use by children <5 years old. Methods We compared estimates based on data collected via HF surveys of under-5s attending well-child visits (e.g. immunizations) and sick-child visits vs. representative household surveys (a 'gold standard' method for measuring insecticide-treated net coverage). In Lindi region, Tanzania, we collected contemporaneous data on 637 under-5s via a HF survey (444 well-child visits and 193 sick-child visits), and on 305 households with an under-5 (including 354 children) via a household survey. In Rufiji district, Tanzania, we collected contemporaneous data on 1433 under-5s via a HF survey (911 well-child visits and 522 sick-child visits), and on 328 households with an under-5 (including 455 children) via a household survey. Results Possession of bednets by households with an under-5 was similar using HF data and household data in both Lindi region and Rufiji district. However, reported use of bednets was significantly higher in HF data than household data in both Lindi and Rufiji, as was reported use of insecticide-treated bednets. HF-based data accurately estimated community-level bednet possession in households with an under-5, but overestimated community-level bednet use by 9-35% and insecticide-treated bednet use by 15-21%. Conclusions Information bias rather than selection bias appears to be a key cause for the overestimation of bednet and insecticide-treated bednet use (e.g. social desirability bias: caretakers of under-5s attending health facilities might be more likely to report using bednets and insecticide-treated bednets). Additional studies of the validity, cost and utility of HF-based data to monitor insecticide-treated bednet use are needed before recommending this monitoring strategy for widespread use. Overestimating insecticide-treated bednet use could lead to inappropriate public health actions and missed opportunities for achieving local and global public health goals.