Abstract
Insecurity limits the access to the population. In North Cameroon we used cell phone survey to estimate health access and mortality rate. Results were satisfactory even if some limitations were present. Health
facilities had been heavily impacted by the conflict, population mortality rates were not alarming although a substantial proportion of deaths were directly related to the conflict.
BACKGROUND
Field access has been considerably limited in the Far North region of Cameroon due to the ongoing conflict. Médecins Sans Frontieres (MSF) wanted to estimate difficulties in accessing health care, access to
food and mortality rates of the populations in the Logone-et-Chari and Mayo-Sava departments.
METHODS
We estimated access to health care and mortality rates through cell phone interviews. We selected 30 villages (clusters) in each department. Local Community Health Workers (CHWs) collected all household phone
numbers in the selected villages from which we randomly selected nineteen. To compare telephone interviews to face-to-face interviews for estimation of access to health care and mortality rates we applied the
two methods in parallel in the town of Mora. To evaluate access to food, push messages were sent by the 3 main mobile network operators in Cameroon. We interviewed by phone all identified legal health care
facilities in the area to estimate their attendances and the services offered before the conflict and at the date of the survey.
RESULTS
We reached 43% of the 3,423 numbers called. Over 600,000 push messages were sent and only 2,255 were returned. We called 43 health facilities and reached 34. In The town of Mora, telephone interviews showed
a Crude Mortality Rate (CMR) at 0.30 (CI 95%: 0.16-0.43) and home visits reported a CMR at 0.16 (0.05-0.27), most other indicators showed comparable results except household composition (more IDPs by
telephone). Phone interviews showed a CMR at 0.63 (0.29 – 0.97) per 10,000-person per day in Logone-et-Chari, and 0.30 (0.07 – 0.50) per 10,000-person per day in Mayo-Sava. Among 86 deaths, 13 were attributed to violence (15%), with terrorist attacks being explicitly mentioned for seven deaths. Among 29 health centres 5 reported to have been attacked and vandalized; 3 remained temporally closed; Only 4 reported not being affected.
CONCLUSION
Telephone interviews are feasible in areas with limited access, although a particular attention should be put on the initial collection of phone numbers. The use of text messages to collect data was not satisfactory
and for this purpose is not recommendable. Mortality in Logone-et-Chari and Mayo-Sava was not alarming although a considerable number of deaths were directly related to the conflict.
