A Timely Question--Can Epidemiology Improve Public Health?
SER Symposium Organizers Seek To Find Out
Presumably, epidemiologists know the answer to the question—can
epidemiology improve public health and would answer in the
affirmative. However, Lydia Feinstein and Jessie Edwards
of the University of North Carolina, the organizers of a symposium on
this topic, were aware that epidemiologic data can get ignored in
policy and program implementation, i.e., lost in translation.
They
convened speakers on such varied topics as antibacterial soaps, HIV
and TB, and asbestos to explore the triumphs and failures of
epidemiology. We interviewed Dr. Feinstein about the symposium to
learn more about what insights were gained.
Interview
EM:
A question you asked in your abstract was—can epidemiology improve
public health? Why was that question timely in your opinion?
Feinstein:
This question was timely because, in recent years, the SER meetings
have placed a strong emphasis on developing causal inference methods
that directly estimate intervention effects. These emerging 'consequentalist'
approaches, so to speak, have expanded our ability to conduct
epidemiologic studies with clear interpretations for public health
policy and clinical practice. However, the promise of novel
epidemiolgic methods is often unrealized in the implementation of
interventions in the field or in the development of new policy. The
goal of our symposium was to showcase examples of epidemiologic
studies that have been successfully translated into clinical practice
or policy and to examine the different ways this process can unfold.
Many of us tend to get stuck in this bubble of conducting secondary or
simulated data analyses, with an end goal of getting our work
published in a respectable scientific journal. I think it's important
to take a step back sometimes and think about what it would take to
actually get our results translated into public health action.
EM:
Epidemiologic data are often not translated into policy or action.
What are some main reasons for that in your opinion?
Feinstein:
There are many reasons, one being that many epidemiologists have never
done public health field work and another reason being that many field
workers have never done epidemiology. I think it's important for us to
try to close that gap in both directions.
EM:
Did your symposium discuss any of these reasons or identify new ones?
Feinstein:
We did. We also discussed political barriers, funding priorities, and
feasibility. It was also noted that sometimes epidemiologists study
effects of interventions that could not be conducted in some settings,
or aren't generalizable to particular target populations. As
epidemiologists, we can certainly do more to ensure that our work is
relevant and timely.
EM:
How would you describe the relationship between epidemiology and
implementation science? What is the latter doing that the former is
not, if anything?
Feinstein:
First of all, we should note that people and groups have their own
definitions of "epidemiology" and "implementation science." We sought
to point out how epidemiology can be more useful by addressing issues
directly related to implementation of interventions. Usually,
epidemiology estimates the effects of interventions assuming that
those interventions can be replicated exactly and that all effect
modifiers can be measured (and modeled).
Implementation science highlights reasons that interventions may not
be implemented with high fidelity to the original design (for better
or worse). When we, as epidemiologists, consider implementation of our
findings (in the forms of interventions or policies), we are forced to
think about the assumptions underlying the generalizability of our
results to specific settings, the feasibility of implementing that
intervention or policy in the same way in other settings, and whether
the proposed intervention optimizes some utility function (e.g.,
overall population health) over other candidate uses of the resources
required for implementation.
EM:
Did your symposium identify concrete ways in which epidemiology
improved health?
Feinstein:
Yes, our symposium provided a few examples. Michael Herce
discussed how epidemiology can be used to evaluate interventions and
refine them as needed to suite a specific setting. He applied this
idea to enhance his field work in HIV clinics in Malawi and Zambia.
Similarly, Allison Aiello discussed how her dissertation work
on hand hygiene eventually led to significant policy changes,
including the banning of antimicrobial soaps.
EM:
Do you have any other comments you would like to make about your
session that are not mentioned above.
Feinstein:
My co-chair Jessie Edwards and I were really glad to have the
opportunity to organize this session. We had a fantastic turn out and
a strong showing from a wide range of individuals, including
methodologists, those whose work focuses on addressing more
substantive research questions, and public health field workers.
Bringing together this sort of integrative audience is an important
first step to improve our ability as epidemiologists to have a strong
positive impact on public health. ■
|