SMS-based smartphone application for disease surveillance has doubled completeness and timeliness in a limited-resource setting – evaluation of a 15-week pilot program in Central African Republic (CAR)

El-Khatib, Ziad; Shah, Maya; Zallappa, Samuel N; Nabeth, Pierre; Guerra, José; Manengu, Casimir T; Yao, Michel; Philibert, Aline; Massina, Lazare; Staiger, Claes-Philip; Mbailao, Raphael; Kouli, Jean-Pierre; Mboma, Hippolyte; Duc, Geraldine; Inagbe, Dago; Barry, Alpha Boubaca; Dumont, Thierry; Cavailler, Philippe; Quere, Michel; Willett, Brian; Reaiche, Souheil; de Ribaucourt, Hervé; Reeder, Bruce

We conducted a set of analyses for completeness and timeliness of reporting, duration of transmission, and usability of the app (see definitions in Box) [1] . First, we compared these parameters for MK during the pilot period (epidemiological weeks 9-23, 2016), here termed 'MK 2016' , with the data for the same district in the same time period the year prior when standard paper-based reporting had been used (epidemiological weeks 9-23, 2015), termed 'MK 2015'. Second, we compared the data from MK 2016 to that from the district of Nana-Mambéré (NM) during the same time period (epidemiological weeks 9-23, 2016), termed 'NM 2016'. Wilcoxon rank sum test was used to compare data completeness and timeliness for MK 2016 with MK 2015 and NM 2016. As the surveillance focal points were absent from their health facilities for the booster-training workshop in Berberati during epidemiological week 14, they did not submit their Weekly Reports until returning the following week. Therefore completeness, but not timeliness of reports, was calculated for week 14. For analysis, the 15-week period was divided into five three-week periods and the distribution of transmission times for these periods compared using a Kaplan-Meier survival analysis (Stata/SE Version 12.0 stcox Efron procedure) [20] .