Evidence Summary: How HealthLearn Works

The problem

Low-quality health care causes many preventable deaths

There are millions of newborn deaths every year; some of these deaths are easily prevented when health workers implement up-to-date clinical practices. For example, studies show that waiting at least 30 seconds to clamp and cut the umbilical cord at birth reduces preterm newborn deaths by 32%. If delayed cord clamping were universally implemented, it would avert over 140,000 newborn deaths annually. In many cases, these recommended evidence-based practices are simpler or cost less than current practices. 

There are many straightforward evidence-based newborn care practices that have a large impact on health outcomes, for example:

• initiating breastfeeding quickly - within 1 hour of birth (vs. later): 25% reduction in newborn deaths

• delaying umbilical cord clamping (>30 seconds after birth): 32% reduction in deaths of preterm babies

• providing “Kangaroo Care” for babies who need it: 40% reduction in deaths of low birthweight babies

• giving the mother oxytocin immediately after birth: 49% reduction in severe bleeding

Unfortunately, life-saving clinical recommendations are often “trapped” in dense guidance documents, so we still live in a world where many health workers don’t provide up-to-date care simply because they are not aware of current best practices. This problem is not restricted to newborn care. Poor-quality health care leads to 5 million deaths every year.

The solution

Training can save lives

According to a recent report by GiveWell, training facility-based health workers to provide better care at birth reduces newborn deaths by 33%; such training may be among the most cost-effective health programs known. Though in-person training can drive huge improvements in health outcomes, it is hard to scale up while maintaining quality.

Our approach

We have developed a new way to train health workers with simple, mobile-optimized, case-based online courses. We leverage user-centered design along with insights from the science of learning and behavioral sciences to maximize engagement and learning. 

Some key features include:

• a user experience that is simple and organized, which reduces frustration and sustains learners’ sense of self-efficacy.

• continuous engagement in low-stakes questions with feedback, which is a powerful way to activate learning. 

• brief explanations that follow every question, which improves learners’ subjective experience. 

• learning driven by clinical cases and content curated to be directly related to learners’ work, which increases HWs’ interest and motivation.

• modules designed to iteratively repeat and build upon key concepts, which leverages the power of spaced repetition to activate learning.

• learning focused on only the most essential content, which improves knowledge retention by making the best use of limited working memory.

• short, modular learning experiences, which increase course completion rates. 

• a focus on basic knowledge and skills, rather than advanced clinical practices. 

An early pilot of an infection control course had excellent outcomes (a high percentage completion, very positive learner feedback, and substantial knowledge gains), and we have continued to improve the approach in subsequent pilots.

Theory of change

Our theory of change delineates the key steps that link our activities to impact. While there are uncertainties in the model, there is some evidence to suggest that we can overcome each of these uncertainties.

Our plan

We're currently piloting a short course on newborn care, and we aim to rapidly expand to offer several courses. We aim to create a cost-effective, scalable way to upskill health workers and lower the burden of readily preventable deaths. As we grow, we're committed to evaluating the key uncertainties in our theory of change and making continuous improvements to improve our efficacy. We aim to focus on a handful of topics that are major drivers of morbidity and mortality. See our FAQs for more information.