Short summary

The use of mobile phones as a tool for health interventions has shown significant potential worldwide. Therefore we have decided to use mobile phone interviews as a follow-up tool in a large randomised trial involving two different strains of BCG vaccine. Provided we can document that we can get in contact with a high proportion of families, define determinants for a high success rate, and document that the information obtained is valid, then using mobile phones for follow-up and collection of outcome information could likely become a very useful and widely used tool in future randomised trials in low-income countries.

---

Rationale

The use of mobile phones as a tool for health interventions (mHealth) has shown significant potential worldwide, and is increasingly used in low-income settings. mHealth interventions can be efficient in increasing awareness and demand for health services.

An important prerequisite for mHealth projects to be successful is access to mobile phones, which has grown rapidly in many low-income countries. According to the World Development Indicators provided by The World Bank, it was estimated that in 2013 there were 74 cellular subscriptions per 100 people in Guinea-Bissau compared to 45 in 2011 and 63 in 2012.

The increasing availability of mobile phones opens a new possibility when conducting longitudinal studies, namely the collection of outcome information by means of mobile phones. In Guinea-Bissau, we have initiated a large-scale randomised trial, comparing the effect of two different strains of Bacille-Calmette Guérin (BCG) vaccine at birth on the risk of hospitalisation up to 6 weeks of age. Newborns are recruited and randomised at the maternity ward of the National Hospital Simao Mendes, where around 6,500 deliveries take place every year. As a part of the enrolment procedure, we collect as many as possible and up to six mobile phone numbers from the mother, father and relatives of the newborn. Six weeks after enrolment, the family is contacted to obtain information on whether the child has been hospitalised. Using available telephone numbers, we attempt to contact the family a total of three times to achieve this information.

Provided we can document that we can get in contact with a high proportion of families via mobile phone, define determinants for a high success rate, and document that the information obtained is valid, then using mobile phones for follow-up and collection of outcome information could likely become a very useful and widely used tool in future randomised trials in low-income countries.

---

Description of the cohort

The Bandim Health Project (BHP, www.bandim.org) runs a Health and Demographic Surveillance System (HDSS) site in urban Guinea-Bissau following women and their children through home visits. The HDSS was established in 1978. As of 2016, a population of around 100,000 individuals in six suburbs of the capital Bissau (Bandim 1 and 2, Mindara, Belem, Cuntum 1 and 2) are followed with regular home visits. All pregnancies and deliveries are registered in the study area, and a large range of background factors related to maternal age and education, socioeconomic status etc. are collected at the time of delivery.

In 2015, we initiated a randomised trial, BCGSTRAIN, randomising healthy newborns about to be discharged from the maternity ward to two different strains of BCG. Currently the strains are used interchangeably by UNICEF, but we have reason to believe that they have different effect on overall health. Around 6,500 children are born every year at the maternity ward. Of these, approx. 20% come from the BHP study area, whereas the rest come from other parts of the capital Bissau.

The plan is to enrol 12,000 newborns within 2 years, thus enrolling children from all over the capital. A study of this magnitude is only possible because we will conduct follow-up by means of mobile phones, and not as we have previously done, through home visits. However, the subgroup of participants living in the BHP study area receive additional follow-up with home visits at 2 and 6 months of age and HDSS visits as per routine every 3 months. 

Data and biological material

At enrolment after informed consent has been secured, we collect mobile telephone numbers of the mother, the father and up to two people living in the same household. We aim for getting as many and up to six numbers. There are currently two main mobile providers in Guinea-Bissau, MTN and Orange, and it is very expensive to call from one to the other; hence individuals often have two mobile numbers. We also interview the mother about background factors such as maternal age and ethnicity, and we measure arm circumference as an indicator of nutritional status.

Follow-up: At 6-8 weeks of age, at the follow-up mobile phone interviews and at home visits to infants residing in the BHP study area, the following questions are asked in a standardised, comparable manner:
a)  Has your child been to a medical consultation since the enrolment into the trial? If yes, how many times?

b)  If yes, for each episode

1. Where and which date/at what age was the consultation?
   
2. Was the child hospitalised and if yes, for how many days?
   
3. What was the diagnosis?
   
4. Is the child alive? If not, when did the child die?
   

With respect to the mobile phone interview, it is noted which numbers were called and how many times, which number that lead to contact, and which relative provided the information. With respect to the home visits, it is noted which relative provided the information. 

Collaborating researchers and departments

Faculty of Health Sciences, University of Southern Denmark

• Medical student Elise Brenno Stjernholm

OPEN Odense Patient data Explorative Network, Odense University Hospital & Statens Serum Institut

• Professor Christine Stabell Benn, MD, DMSc, PhD
• PhD-student Frederik Schaltz-Buchholzer, MD

Statens Serum Institut

• Peter Aaby, DMSc

OPEN Odense Patient data Explorative Network, Odense University Hospital

• PhD-student Emil Rossing Olsen

Department of Public Health, Hospital of Southwest Denmark, Esbjerg

• Morten Bjerregaard-Andersen, MD, PhD