(Introduction by Ramya Ambikapathi with guest post from Ye Shen from Harvard University)
I have been really fortunate to have met and had a conversation with Dr.Namukolo Covic, when she visited Purdue University for the Ending Global Hunger Colloquium held in April 2019. One of the profound things she said was that “so much research is focused on describing the problem, but not about the solutions” [paraphrasing]. My mentor, Dr.Nilupa Gunaratna, and I often circle back to this conversation as we discuss new and old ideas for research. These two types of research approaches (describing the problem vs research on solutions) to global nutrition research have been labelled as “conventional” and “frontiers” approaches by Pelletier and colleagues. Conventional refers to research efforts on “generalizable or fundamental knowledge that answers scientific questions” (Pelletier et al, 2013). While frontiers research is “to create actionable knowledge of issues and problems of concern to stakeholders, organizations, communities, or publics at various scales” (Pelletier et al, 2013). See table 1 below from Pelletier et al (2013) for further breakdown of the research dimensions.
While both approaches are complementary, there is a need to expand more into frontier research approach as we face large threatening challenges like the COVID-19 Pandemic and Climate Change (Pelletier et al, 2013). Focusing on implementation research and decision science is one use of frontier approach to tackle complex challenges. Ye Shen, (Twitter: @Yeah_Shen), a Health Policy doctoral student at Harvard University is pursuing a career in decision science for global nutrition and explains her career trajectory.
Why I am interested in decision science for global nutrition?
By Ye Shen
Over the past couple of years, I have sensed that there remain substantial missing links between applied nutrition research and real-world actions. These areas are not yet commonly nor adequately addressed by nutrition researchers whose primary focus is to produce one piece of new nutrition knowledge at a time. I began to recognize that research uptake would not happen as soon as a research team’s new findings could be presented to the policy world (although that alone is already an achievement on its own!). I started to accumulate unanswered questions and was particularly drawn to the following two categories:
- Determining research questions and methods that can directly inform policy decisions around nutrition actions:
- What types of policy decisions need to be made in order to act?
Here are some examples:
- Decisions on the optimal choice (or the optimal mix of options) of nutrition actions in a given context
- Decisions on the optimal design for a given nutrition action in a given context?
- Decisions on the funding of future research and data collection/ infrastructure needed
- How to appropriately define and quantify the value of a nutrition action? This gets especially complicated when we think about nutrition-sensitive / food system/ multi-sectoral actions.
- Nutrition action at what cost? What and whose resources are available to allocate for a nutrition action in question? What are the tradeoffs (e.g. opportunity costs) inherent in action (and inaction) to different stakeholders?
- How to appropriately incorporate all available information (often from multiple data sources and across disciplines/domains) into one analysis that can directly inform a policy decision?
- How much uncertainty remains given available information? What are the major drivers of uncertainty?
- Which areas of future research and data collection are worth prioritizing in order to improve decision-making?
-
-
3. Understanding how to implement nutrition actions:
-
- Given what’s known from research, what are key aspects of implementation for replication in a similar setting in the real world?
- What about scaling up?
- What about transferability to a different setting?
I became increasingly interested in the decision science framework once I realized that a lot of my unanswered questions (e.g. in category #1 above) center around decision as the unit of analysis. Decision science toolbox utilizes theory and methods from many disciplines such as statistics, economics, psychology, operations research, and epidemiology. Simulation and mathematical modeling are frequently used to model decisions under uncertainty. Cost-effectiveness analysis and cost-benefit analysis are two of the more commonly used decision analytic methods in applied nutrition. I will just highlight one more interesting decision science approach here called “Value of Information” which aims “to assess the value of acquiring additional evidence to inform a clinical or public health decision” (Mineli and Baio, 2015).
While decision science is also useful to inform some aspects of category #2 above, implementation science should play a key role here. This webpage from UW has a video that explains what implementation science is, especially in the context of global health. Decision science and implementation science are both interdisciplinary in nature with some overlap in methods and disciplines. In my current understanding, they can be complementary to each other, especially in informing decisions around optimal design of a nutrition action for a given context. Implementation research collects costs which serves as important data sources for decision analytics. Decision modeling will also benefit from implementation data on intermediate processes of nutrition actions.
I have just started my PhD program in Health Policy (Decision Sciences track) at Harvard University. I am curious about how my thinking around this topic might evolve from here after two years of coursework and after more years of doctoral research!
