AVADAR (Auto-Visual AFP Detection and Reporting): Demonstration of a Novel SMS-based Smartphone Application to Improve Acute Flaccid Paralysis (AFP) Surveillance in Nigeria

"AVADAR, in conjunction with health education and sensitization, are not only useful for health workers but also invaluable in engaging community members to report cases of AFP through a community-based surveillance approach."

Eradication of polio requires that the acute flaccid paralysis (AFP) surveillance system is sensitive enough to detect all cases of AFP, and that such cases are promptly reported and investigated by disease surveillance personnel. The Nigerian AFP surveillance network is faced with issues of poor detection and reporting of cases in certain areas (e.g., remote and/or security-compromised settlements), therefore increasing the likelihood of hidden reservoirs of the virus. AFP surveillance is affected by the ability of health workers (HWs) and community health informants (CHIs) to correctly recognise and report AFP cases. This paper looks at one information and communication technology (ICT)-based approach - the Auto-Visual AFP Detection and Reporting (AVADAR) smartphone application (App) - as it was applied in a 2-stage study to determine a proof of concept and the feasibility of AVADAR in enhancing polio detection and eradication.

AVADAR was conceptualised by the paper's first author and developed by collaborative efforts of multiple developmental partners, including BMGF, WHO, and Novel-T (a Swiss-based software company). The App supports: (i) weekly reporting by HWs and CHIs on the presence or absence of AFP cases in their respective areas; and (ii) instant notification of an AFP case with a minimal set of information being collected and sent directly to the Disease Surveillance Notification Officers (DSNO) responsible for the area (district or Local Government Area - LGA). The App creates a scenario where the AFP surveillance network is not dependent on a limited number of priority reporting sites. It is installed on the phones of a large number of HWs (midwives, nurses, and surveillance focal persons) and CHIs (traditional birth attendants, traditional spiritual healers, traditional bone setters, influential youths, village heads, patent medicine vendors, and traditional barbers), potentially making every health facility, health worker, and community-based AFP volunteer part of the detection and reporting system.

Conducted across a wide range of communities between June and August 2016, the 8-week pilot study included 198 participants in 2 LGAs in Kwara state and Kuje area council in the federal capital territory of Nigeria. The researchers subsequently scaled up the study in 2017 to cover 8 LGAs in Borno state (in northeastern Nigeria), with 527 HWs and CHIs added to the AVADAR network.

HWs and CHIs were trained in small groups of no more than 20 people by facilitators at the LGA level. A mixture of practical demonstrations and role play sessions were conducted. At the end of the training, relevant HW and CI information was collected (e.g., name, phone number, and health facility) and uploaded to the server that connects each HW and CHI to their respective DSNO.

All participants were provided with mobile phones with the AVADAR App pre-installed. As a learning tool, the AVADAR App contains an embedded 30-second video of a child with symptoms of AFP, as well as a case investigation form that HWs and CHIs use to record case details of each child suspected of having AFP. At a designated day and time each week, the video appeared on the phone of study participants. Informants in the study reported suspected AFP cases on a weekly basis to a cloud-based server using the AVADAR App. Once a HW or CHI sends a suspected AFP case report, the server automatically sends a short messaging service (SMS, or text) alert to the respective DSNO's mobile phone with a summary of the suspected AFP case information. The DSNOs use a customised open data kit (ODK) application form on their mobile phones to investigate all reported AFP cases and send confirmed case details to the AVADAR server. Data are collated from the AVADAR servers and used for data analysis.

A HelpDesk phone line was provided for informants to call if they needed troubleshooting assistance. In addition, joint partner field monitoring and supervision visits were conducted on a weekly basis by WHO and eHA.

The performance of HWs and CHIs was measured on 2 primary indicators: (i) "completeness" refers to when HWs or CHIs send in at least one report (AFP or "zero" report) within a week; and (ii) "timeliness" refers to when HWs or CHIs send in at least one report (AFP or "zero" report) within 48 hours of receiving the video notification reminder on their mobile devices. Four out of the 10 wards in the Kuje area council met the 80% target for completeness, while none of the wards met the target for timeliness of reporting. (Kuje had consistently poor network connectivity, which may have caused some SMS reports from informants to be delayed or not get through to the server.) All wards in the Oyun LGA in Kwara state met the set target of 80% for timeliness and completeness of reports.

Overall, in both LGAs, 83% reporting by wards of true AFP cases was documented during the AVADAR study, compared with 36% using the traditional AFP surveillance system during the pre-pilot period. This is an improvement of over 130%.

Data shared in the paper also show the App's potential to detect cases in security-compromised and silent reporting areas. "In this final push to eradicate polio from Nigeria, all resources are needed to ensure that the surveillance network is sensitive enough to detect hidden reservoirs of the virus."

In short, the App demonstrated an improvement in AFP reporting by tackling issues with the surveillance system on a small scale. When planning scaled-up efforts beyond Borno state, there are aspects of AVADAR that the researchers note should be considered, including: the status of surveillance performance indicators, proportion of displaced and nomadic populations, security accessibility, network availability, ample equipment buffer stock supply for replacement purposes, geographic spread of proposed LGAs/districts, and the cost-effectiveness and sustainability of the utilisation of smartphones. They suggest that the inclusion of a focus group discussion (FGD) with the study participants in the study protocol would shed more light on the challenges and areas to focus on during expansion efforts.

Furthermore, "While the results clearly show that the use of AVADAR led to more cases being reported, it is important to identify the differences in the detection and reporting of cases by health workers compared with community members. This would indicate specifically to whom the App could be more tailored and whether it would be an optimal and cost-effective intervention for further use and expansion to additional geographic locations in the polio eradication campaign."