In this issue of the Ki Community newsletter, we are excited to share the Data Science Scrum Framework. This framework is a resource to understand how Ki develops actionable answers to specific global health questions in our data science rallies. We use an adapted scrum management framework and the principles of Agile development to:
- Focus on the most important questions using the right data
- Engage data scientists in the most sophisticated modeling activities
- Make sure data scientists are constantly informed by domain experts
- Capture and document all work
Learn more about the Data Science Scrum Framework at https://www.synapse.org/#!Synapse:syn17077830
We would also like to congratulate the Ki Grand Challenges community on the virtual convening that took place last year. Read more below about the 2020 Ki Grand Challenges data science convening.
INDIA COUNTRY OFFICE
We are pleased to share an update on work in the India Country Office (ICO):
Considering complex health and development challenges, their multifactorial causal pathways, and the existing data silos, ICO brings together grantees, decision-makers, and domain experts to boost the impact of investments by creating a data and analytics community with a focus on global development. ICO has a leadership role in promoting the adaptation of Ki principles and operations for regional culture, practices, and areas of development in India.
ICO and Ki initiated interactions with the Bihar Measurement, Learning, and Evaluation (MLE) partners to create a Data and Evidence ecoSystem in India (DESI). This is a pilot program to innovate on interventions and service delivery approaches by responding to learnings from data in a cyclic approach. The goal of this program is to work toward ensuring short data analysis cycles feed continuous program improvements and build an evidence base for investment and decision making.
ICO and Ki conducted a workshop, followed by bilateral interactions with the Bihar MLE partners to confirm value and generate excitement about combining partners and their data to inform policy and investments in India. The partners bring rich domain expertise and deep local experience to help fuel data analysis activities. Partner feedback helped ICO learn more about the perceived potential barriers and work towards alleviating them.
ICO, along with its partners and Ki, is co-curating a governance mechanism that will help in setting processes and safeguarding the community. Data sharing agreements and data transfers are underway. In addition, ICO has also initiated interactions with potential repository hosts and targets to setup this data community supported by a repository in 2021, demonstrate value, and expand to a second collaboration beyond Bihar MLE partners, such as Uttar Pradesh Technical Support Unit and Family Planning MLE.
Simultaneously, the ICO is working with the FAIR Working Group to pilot grant milestones tied to data sharing to leverage funding incentives at the time grants are made to increase data sharing at grant completion. These pilots have been informed by learnings from DESI and future DESI related efforts may be aided with improved data sharing.
KI INSIGHT STORY – PREGNANCY RISK STRATIFICATION
Ki Insight Stories are an opportunity for us to thank the Ki Community for your generous data contributions and keep you informed on the important insights that Ki and our BMGF partners are generating.
Our latest Insight Story presents the results of the Pregnancy Risk Stratification (PRS) data science rallies. The data for these rallies represented about 137,000 mothers and children from a variety of countries. This data made it possible to predict the risk of pre-eclampsia throughout pregnancy using longitudinal data and predict the risk of neonatal death or stillbirth.
The goal of these data science rallies was to inform the development of an algorithm that can be used by minimally trained users in low-and-middle-income countries (LMICs) and enable the prediction and prevention, or mitigation, of adverse birth outcomes.
We deeply appreciate the hard work and involvement of all our data contributors, domain experts, and partners. Without you, this work would not be possible. We would particularly like to acknowledge the data contributors for the PRS data science rallies:
If you are a Ki data contributor or partner, we invite you to download the full Insight story by joining the Ki Insight Stories space on Synapse, a collaborative research platform.
KI GRAND CHALLENEGES CONVENING
Ki Grand Challenges hosted its annual data science convening in October 2020 as a satellite event for the Grand Challenges Annual Meeting. Due to the COVID-19 pandemic, this convening was the first to be conducted entirely virtually. While we missed the chance to gather in person, the virtual setting allowed us to create a truly global convening, with attendees representing all three Ki Grand Challenges regions: India, Africa, and Brazil. The virtual setting also allowed us to invite more attendees to participate, including entire research teams.
The plenary session on the impact of data-centered projects on health policy was delivered by Cristiano Boccolini (Brazil), William Ogallo (Africa), and Shinjini Bhatnagar (India). The second part of the meeting was led by KiGC grantees who proposed topics and conducted several breakout sessions, and there were opportunities for networking and discussions of key topics related to the use of Data Science for public health such as data sharing. Some of these grantees are working on hosting follow up meetings and actions.
We thank everyone who contributed to making the convening a success and we are already looking forward to the next meeting.
UPDATE ON DATA SETS
Thanks to you, Ki’s repository is ever-expanding. Please find below a list of studies that have completed QC after the last announcement.
- BARNARDS – India, Pakistan, Bangladesh, Nigeria, Rwanda, Ethiopia, South Africa
Burden of antibiotics resistance in neonates from developing societies
- HUST-TMCHC – China
Optimal gestational weight gain for Chinese urban women
- INPer-NeuroObesity – Mexico
Obesity and maternal metabolic profile as a predictor of fetal body composition, obesity and neurodevelopment in childhood
- LSHTM-ENID – Gambia
Randomized trial to investigate the effects of pre-natal and infant nutritional supplementation on infant immune development in rural Gambia
- UBA-WTGain – Argentina
Multi-central study on weight gain and guidelines of food selection during gestation and its impact on the newborn
- UCL-LBWSAT – Nepal
A cluster-randomized controlled trial testing impact on birth weight and infant nutrition of Participatory Learning and Action through women’s groups, with and without unconditional transfers of fortified food or cash during pregnancy in Nepal
- UMan-MatHealth – Nigeria
Maternal malaria status and metabolic profiles in pregnancy and in cord blood: relationships with birth size in Nigerian infants
- USP-MatStress – Brazil
Maternal psychological stress and distress as predictors of low birth weight, prematurity and intrauterine growth retardation
If you would like to learn more about these datasets or any other datasets in the repository, use our study explorer tool. The study explorer tool allows you to see all the data in the repository, at a metadata level, including the data you contributed. Here is the link: http://www.studyexplorer.io/