Comparing reported and observed use of insecticide-treated nets in Tanganyika, Democratic Republic of the Congo: a post-campaign longitudinal household study.

backgroundAccurate measurement of insecticide-treated net (ITN) use is essential for evaluating malaria control programs. Most routine indicators rely on self-reported ITN use, which may be affected by recall and social desirability biases. This study aimed to compare reported and directly observed ITN use, assess the validity of self-reported indicators, and examine behavioral patterns and net durability following a mass distribution campaign in Tanganyika Province, Democratic Republic of the Congo.

methodsA longitudinal observational study was conducted across four survey rounds between April 2021 and August 2022 in Tanganyika province. Household members were directly observed over three consecutive nights to document ITN use and nighttime behaviors, followed by household interviews. Self-reported ITN use was compared with direct observation at the individual level. Diagnostic performance metrics, including sensitivity, specificity, accuracy, and Cohen's Kappa, were calculated using observation as the reference standard. ITN access, reported use, and physical integrity were also assessed over time.

resultsA total of 72 households (456 individuals at baseline) were included. Self-reported ITN use remained high across rounds (average approximately 91%), whereas observed use ranged from 72.9% to 76.5% when expressed relative to the total surveyed population. Sensitivity of self-reported ITN use was consistently high (97.8-99.4%), while specificity was low (18.7-47.7%), indicating frequent misclassification of non-users as users. Overall accuracy ranged from 79.1% to 85.6%, but agreement between reported and observed use was limited (Cohen's Kappa: 0.25-0.56). Direct observation revealed that ITN use was dynamic, with incomplete or inconsistent use occurring throughout the night. ITN access declined over time, and physical integrity deteriorated substantially, although most nets remained serviceable. Reported ITN use among children under five years remained consistently high (97-100%), suggesting prioritization within households.

conclusionsSelf-reported ITN use is a highly sensitive but poorly specific indicator, leading to systematic overestimation of effective coverage. ITN use is dynamic and influenced by both access and net condition, which are not fully captured by standard survey indicators. Malaria control programs should interpret self-reported ITN use with caution and consider complementary approaches to better assess actual protection and sustain effective coverage over time.