Determining Inaccurate Coordinates in Electronic Data Collection for Surveillance and Immunization Supportive Supervision: A Case Study of Nigeria EPI Supportive Supervision Module

"The empirical results can help us to understand the variation in geospatial data collation across devices and highlight critical criteria for choosing mobile phones for mobile surveys and data campaigns."

Programme managers and policymakers need accurate and timely data as input for decision-making processes that impact behaviour monitoring and ultimately help reduce morbidity and mortality. The use of mobile-phone-based systems and geospatial technologies for collecting such data is becoming a critical component of public health programme implementation. This study describes and evaluates a data-gathering process used by the World Health Organization (WHO-Nigeria, Expanded Programme on Immunization, or EPI) that uses phone-based in-built global positioning system (GPS) sensors to identify the position of users while they undergo supportive supervision. The aim of this study is to identify the best and most appropriate mobile device to be used for data collection during supportive supervision, thus improving performance.

As reported here, the Ebola Virus Disease outbreak in Nigeria in 2014-2015 required submission of real-time data on each case and contacts to the Emergency Operations Center (EOC) for early detection and response. Mobile data collection replaced the paper-based manual method of data collection to conduct supportive supervision and other surveys in the WHO. Specifically, the Open Data Kit (ODK) and Form Hub technology were used in combination with the Dashboard technology and ArcGIS mapping for follow-up of contacts, identification of cases, case investigation and management, and strategic planning during the response. Lessons learned were collated and brought into the routine immunisation (RI) and surveillance systems via the integrated supportive supervision (ISS) system. Spatial data are collected using the ODK GPS tool, which interfaces with the mobile phone GPS sensor to fetch geo-coordinates. At the beginning of ODK's introduction for ISS visits, no phone brands were recommended by WHO, with personnel allowed to purchase and use any brand of their choice. This study reviewed data gathered from 8 brands of smartphones (40 models) over 1 year of using the mobile data collection to predict the deviation pattern for spatial data acquisition via mobile phones by different brands.

Immunisation coverage varies across Nigeria with improvements needed in every state, as no state meets the global coverage of 90% coverage for the three doses of pentavalent vaccine. The immunisation coverage is least in the Northern zones and mitigated by having more WHO staff to support the ministry of health (MOH) EPI activities. The WHO personnel implemented the use of android-based mobile data collection using the ODK for supportive supervision with 81,505 records available on the server, corresponding to 81,505 ISS visits to health facilities. All data from the 36 states plus federal capital territory (FCT) (36 + FCT) in the country were selected and included in the study.

The study found that, from the total number of visits recorded (81,505), 1,119 (1.6%) represent records with inaccurate coordinates. The iTEL brand fared worst, with a 63.1% proportion of records with inaccurate coordinates across 3 different models; in contrast, Motorola phones had no inaccurate coordinates. The South-East had the best records at all levels, with mistakes only coming from the local government area facilitators (LGAFs), while the South-South and South-West cluster coordinators (CCs) had a greater number of records with inaccurate coordinates as compared to the other zones, which may be due to poor training or staff not being aware of the importance of the geo-coordinates.

Overall, the LGAFs recorded the highest number of inaccurate coordinates, as this group of personnel is the lowest cadre with limited or no adequate training on the use of ODK at the time of the study. The LGAFs also receive low remuneration as compared to CCs and state coordinators (SCs); thus, they may be more inclined to buy cheaper, low-end mobile phones that are not able to capture coordinates easily.

Based on the results, the researchers recommend:

• Setting a minimum requirement for any phone to be used in conducting supportive supervision visits;
• Procuring phones that have the least proportion of inaccurate coordinates, as identified in this study, halting the bring-your-own-device (BYOD) policy; and
• Providing training to all staff carrying out this activity on the basic use of phones, the taking of geo-coordinates, and scenarios that affect coordinates' capture in the field.

The researchers note that WHO-Nigeria staff members' use of ODK for ISS visits can strengthen accountability within the RI and surveillance system, improve transparency, and increase donor trust through documented and verifiable evidence. The findings from this study can be used to improve the ISS process and perhaps support increases in immunisation coverage in all parts of the country.