A team of about 20 researchers and professional support staff, employed by the London School of Hygiene & Tropical Medicine, set off to find out “what works, why and how” when it comes to improving the health and lives of women and babies. Over the next 7 years we generated a wealth of new findings and knowledge, summarised in the Informed Decisions for Actions in Maternal and Newborn Health 2010-2017 Report, and on our brand-new website.

What’s changed since 2010? The world has moved from the era of the Millennium Development Goals with their focus on maternal and child survival, to a focus on universal health coverage with the advent of the Sustainable Development Goals. People around the world are living longer: in sub-Saharan Africa, girls born in 2015 have a life expectancy nearly 4 years more than for those born in 2010. Globally, there were 1.1 million fewer child deaths in 2015 than in 2010, and for every 100,000 live births there were 30 fewer maternal deaths in 2015 than there were in 2010. Despite this progress, maternal and newborn death rates in Nigeria, India and Ethiopia remain among the highest in the world. And just as in 2010, most of these deaths are preventable at low cost.

The IDEAS team has built strong and lasting working relationships with implementation partners. Between 2010 and 2017, we completed nearly 80 technical support activities for implementation partners, from reviews of research protocols to cross-country learning workshops. And we have produced 17 data sets, 27 reports, 19 journal articles, 10 research briefs, and 5 infographics.

Our research focused on four learning questions:

• What innovations are implemented and how are they expected to improve maternal and newborn health?

We identified 57 diverse innovations, put in place by 9 implementation partners, with whom we worked to identify anticipated effects of each innovation.

• Did innovations increase the coverage of life-saving interventions, and if so when, how and at what cost?

We did cross-sectional baseline surveys of resident women with a recent birth, of primary health facilities and of front-line health workers, in 2012 and 2015. We found important gains in access to antenatal care and care at birth and, in Ethiopia and Uttar Pradesh, some improvements in the quality of care families received.  However, indicators for immediate newborn care lagged behind, and many of the inequities in access to health care were observed to persist. We used qualitative methods to understand how practices were influenced by front-line health workers, and found that in Ethiopia newborn care practices changed through three important factors: (1) Getting the word out: ensuring that the right messages get to families with high coverage and through multiple and trusted channels. (2) A desire to be modern: harmful behaviours and practices can  change because families want to be modern, and because knowledge gives them the power to oppose contrary views and (3) Delivering in a facility: facilities provide information and are also responsible for behaviours such as wrapping the baby and early breastfeeding.

• How and why does scale-up happen?

We undertook over 200 in-depth interviews and three case studies of successfully scaled-up interventions and identified six critical actions which implementation partners adopted to catalyse innovation scale-up. These included: designing innovations with scale-up in mind; generating evidence on how to implement at scale as well as evidence of impact; harnessing the power of influential individuals who could be instrumental to scale-up; being prepared for and responsive to the policy, health systems and sociocultural context; supporting government in a transition to scale, thus ensuring continuity; and embracing aid effectiveness principles.

• To what extent did scaled-up innovations affect coverage of lifesaving interventions?

Scale-up takes time, and working at scale brings an additional degree of complexity. Early on in IDEAS we found that health decision-makers at district level in each country shared our interest in data for decision-making. This led us to develop a novel method for assessing implementation strength of scaled-up innovations – the Data Informed Platform for Health. After studying the feasibility in all three countries we carried out an 18-month prototype phase in India. And in Ethiopia, responding to a government request, we are undertaking an evaluation of the national-scale Community Based Newborn Care programme by studying changes in intervention and comparison areas using surveys and qualitative enquiry.

A second phase of IDEAS has started; in which we are working more closely with government in each of the three settings. The research includes improving coverage measurement; tracking progress in coverage of life-saving interventions; supporting the local use of data in decision-making; understanding the mechanisms underlying quality-improvement; and conducting a study of sustainability.