Abstract
BACKGROUND: Numerous national health intervention campaigns, e.g. supplementary immunization campaigns/activities (SIAs), have been conducted in low- and middle-income countries (LMIC) in the last decades. These campaigns are rarely evaluated for overall health outcomes. Information on campaigns is critical for evaluations. We investigated; 1) quality of campaign information sources and 2) implication of quality for outcome evaluations. METHODS: We focused on three campaign types: oral polio vaccine (OPV), vitamin A supplementation (VAS) and measles vaccine (MV) campaigns in two case countries, for which "gold standard" information on campaigns collected regularly at Health and Demographic Surveillance Systems (HDSS) sites: Guinea-Bissau and Bangladesh. We compared the campaign information from HDSS with information from the World Health Organisation (WHO) and the Rotary Foundation (Rotary, only OPV campaigns). First, campaigns were matched and compared based on intervention type, date of campaign and target age group. Second, we assessed the implications of using various sources of campaign information on the estimated effect of OPV campaigns on all-cause under-3-year mortality in Cox proportional hazards regression models. RESULTS: The proportion of matched OPV campaigns was highest between HDSS and Rotary. VAS campaigns (only information from HDSS and WHO) matched poorly. The estimated effect of OPV campaigns information on child mortality in Bangladesh went from being statistically significant (HR = 0.69 (0.52-0.90)) using HDSS campaign information to not being significant (HR = 0.93 (0.71-1.21) using WHO campaign information. CONCLUSION: Compared with the HDSS, Rotary had the best campaign information on the conduct of OPV campaigns, whereas the WHO quality of campaign information was low for both OPV and VAS. A low quality of campaign information may alter conclusions of health outcome evaluations. Reliable and precise information on campaigns is essential to assess their effects. Public and private campaign stakeholders should track campaign information meticulously and support that publicly data is available for researchers.