Applicants urged to find diverse partners for new NIH DS-I Africa program

September / October 2020 | Volume 19, Number 5

September / October 2020 Global Health Matters newsletter [PDF <1M]

Medical worker performs ultrasound connected to a table on a pregnant woman lying on cot.
Photo courtesy of Philips

Philips, which developed this portable ultrasound device, is
one of many private companies hoping to partner with
academic researchers on data science activities.

By Susan Scutti

Multidisciplinary collaborations will be key to the success of NIH’s new Harnessing Data Science for Health Discovery and Innovation in Africa (DS-I Africa) program. Applicants are encouraged to form partnerships that reach beyond academia to include governments, the private sector, NGOs and other stakeholders. An additional requirement is to bridge disciplinary divides, where health scientists collaborate with engineers, computer scientists, data experts and others. The goal is to build a sustainable and robust ecosystem for health discovery in Africa.

Multidimensional connections serve pragmatic purposes, according to the partnership panel at the recent virtual DS-I Africa conference. Teams of diverse colleagues raise and address a wider range of issues than siloed collaborations and this greater scope offers a higher probability of success, according to Dr. Beatrice Murage, who develops partnerships in Africa and Asia for Philips. She said she is hoping to work with academic researchers who are customer-focused, able to develop products tailored to “pain points” - what customers ask for and need - and who take the long view. “We want to build solutions that last beyond, let’s say, one pilot or one study or one thesis, solutions that we can scale over time and over market and that have actual business value.”

Her advice for potential partners includes, first and foremost, understanding the customer. “You need to get out there and just have a feel of the problem,” said Murage. Second, partners need to have a business mindset; those hoping to build a solution for a particular problem must first ask, Is this something a client would buy or invest in? Third, Murage suggests that potential partners also “be willing to share what you learn with the ecosystem…share best practices, share what did not work, find a way to communicate your results to different types of audiences and see how their feedback then actually helps us build better.”

Partnerships with study participants are also of great importance, emphasized Dr. Julie Makani of Tanzania’s Muhimbili University of Health and Allied Sciences. She said she looks forward to the development of “well-coordinated prospective and longitudinal cohorts in Africa” so that data scientists might better understand the physiological response to illness and so “identify novel, therapeutic interventions to reduce the burdens of disease.” Makani spoke of one study that established a cohort of 5,000 sickle cell patients and followed them over 15 years. The enormous amount of data generated by this research raised questions far beyond the project itself: How can these data be used to inform policy? How can these data improve both basic science and clinical science research? “Data science is one of the ways that we can ensure that we integrate health, education, advocacy and research so that we improve health,” said Makani.

The private sector also seeks to get the most from data culled from its participant partners. IBM has a long history of focusing on methods to improve health outcomes in the maternal, newborn and child health spaces, while also exploring disease intervention planning, explained Dr. Aisha Walcott-Bryant, research manager at IBM Research Africa. Going forward, she hopes to not only create new datasets but also leverage existing ones. Data science allows her to do this, she said. “A good example here is doing stratification, a very common method to really try to tease out and understand: Who are the vulnerable populations and who is having these poor outcomes?” By using such methods to study child mortality in Nigeria, the technology giant has made unique discoveries. “For instance, mothers with three births in the past five years were more vulnerable than those who only had one birth in the last five years,” said Walcott-Bryant.

To bring about real change, partnerships with governments and other policymakers are necessary, said Brian Gitta, a tech entrepreneur based in Uganda. “We need to build awareness of the importance of data for medicines, for disease tracking and community health monitoring such that we work with a variety of stakeholders, taking ownership of this data in the process,” he said. Gitta said harnessing health data will require more than innovative technologies - it will also take creative management and bold action from governments to ensure that Africans have the resources, tools and skills to lead this transformation.

That opinion is shared by Sierra Leone’s chief innovation officer, Dr. David Moinina Sengeh. “We want to be in a place to ensure that the economy, the identity and governance becomes digital,” said Sengeh. To achieve this goal of “digitization for all,” Sierra Leone administrators rely on three principles when choosing partners. First, potential collaborators need to prioritize mobile solutions, the lifeblood of his nation’s computing power. Second, they must believe the goal of evidence-based policies is achieved through the use of artificial intelligence and other tools. Third, partners must design hybrid technology systems, Sengeh said. “Everything we do has to work online and offline, has to work on web, mobile and paper, and it has to work when there's power and when there's no power,”

Alignment in views and methodology was a key concern for all panelists but they agreed soft skills also matter. “You want that excitement, that energy on the ideas that you all are putting forth, and you want folks that you can talk with about difficult things,” observed IBM's Walcott-Bryant. Simply wanting to be a team member goes a long way as well, she added.

Humility is the most important feature of any collaboration, Makani suggested. Each partner must understand the value of all the partners involved, from governments to patient communities. “Data science is a fantastic platform that will bring disparate people and disparate communities together. No one body or institutional sector can do this alone,” said Makani. “We really need to work together.”