“Knowing is not enough; we must apply. Willing is not enough; we must do.” That’s one early quote attributed to Goethe from Moyses Szklo, Professor of epidemiology at Johns Hopkins and editor of The American Journal of Epidemiology, who spoke as the 157^th Cutter Lecturer at Harvard earlier this year.  His talk was entitled “Semantic and Other Challenges in Translational Epidemiology”. The Goethe quote provided an early clue to Szklo’s personal views about the proper allegiances for epidemiologists, a perennial topic of discussion in the field.

Szklo began his talk by noting that his first thought had been to title his talk “Epidemiology for Public Health” because the concept of translation is really not a new one, he said. He noted that in earlier years many epidemiologists such as Hopkins’ Mort Levin had health officer or public health backgrounds, took for granted their interest in population health, and were proud of their public health achievements such as those surrounding the polio vaccination trials in the 1950’s. He noted that the recent resurgence of interest in social epidemiology signals a renewed interest in population-based strategies.

Definition of Translational Epi

Szklo asserted that translational epidemiology as a concept is still evolving. He presented the varying definitions of translation research as proposed by the National Institutes of Health and later by the Institute of Medicine and found fault with both of them from an epidemiologist’s perspective. He offered his own definition of translational epidemiology as the “effective transfer of new knowledge resulting from epidemiologic studies (including trials) into the planning of population-wide and individual-level disease control programs and policies.” He further differentiated translational epidemiology from implementation science which he described as more focused on the logistics of translation than the transfer of knowledge per se.

The Cycle

To frame his presentation Szklo modified a diagram proposed by Tugwell in the Journal of Chronic Diseases which identified the different steps in the cycle or loop of public health problem-solving.  These are:

According to Szklo, epidemiology is relevant to virtually all the steps in the cycle.

Epi Concepts Related To Translation

An important feature of Szklo’s talk was his decision not to discuss what he called “strategies” for translation, including such activities as multidisciplinary approaches or comparative effectiveness studies, but rather to focus on a conceptualization of translation as the task of transferring knowledge.

Other conceptualizations of how best to go from knowledge to action have been proposed, and Szklo jokingly referred to the famous cartoon reprinted in the Dictionary of Epidemiology which shows two persons discussing the process at a blackboard describing the critical step from knowledge to action as “and then a miracle occurs!”

Confounding

In presenting his version of the process, Szklo first elaborated on specific epidemiologic concepts such as bias and choice of analytic model to illustrate that thinking about these concepts varies depending on one’s purpose---to elucidate causality or to prevent disease.

For example, confounding is something which epidemiologists interested in causality see as a threat to causal inference and thus try “to kill off” as soon as possible, said Szklo. But public health epidemiologists can greatly value confounded associations because they can help identify high risk markers to be used for screening or other purposes. He cited the example of alcohol and lung cancer saying that persons who drink alcohol are likely to develop lung cancer more often. “It’s true,” he said, “it’s not causal, but it’s true!” Another example involves African Americans and hypertension.  Such confounding information could be useful information to know.

The Translation Process

In focusing and elaborating on the steps in the translation of knowledge process, Szklo described the usual steps 1) to collect evidence, 2) conduct systematic reviews, 3) estimate cost-effectiveness, and 4) adopt an evidence-based policy.

He expanded the tasks under 2) above (systematic review) to include 2a) assessing the level of evidence, 2b) the selecting policy or program options, 2c) conducting decision analysis, and 2d) issuing recommendations. According to Szklo, these subtasks are the ones which should draw the attention of the translational epidemiologist.

Carrying out step 2a) he said is not an exact science because it enters the realm of inference and requires judgment. Carrying out step 2b) requires a consideration not only of economic factors, but also of public values, and could benefit from the input of ethicists as well as economists, said Szklo. He acknowledged to the audience that he was “inventing” the steps

in the translation process at this point, and that his presentation should be considered a proposal, and is open to discussion. Readers are invited to submit their thoughts in the space provided online where this article is published.

Ivory Tower vs The Field

In concluding his presentation, Szklo contrasted what he called “academic” epidemiology and translational epidemiology. Because the purposes associated with these two roles differ, some of the key concepts in epidemiology are viewed differently. (see table published separately along with this article).

Epi is Boring…

The talk ended more provocatively than it began with a quote from the late Hopkins epidemiologist George Comstock who told the speaker that “If not considering a possible application to prevention and public health, epidemiology tends to be fairly boring…” While he was quick to identify this statement as just one person’s view, it became clear in the question and answer period, as it had been suggested by the quote from Goethe at the beginning of the talk, that Szklo is fundamentally in agreement with Comstock because he told the audience that in his personal view  “it is terribly important for epidemiologists to get involved in policy.” 

Table
Contrasting Purposes and Concerns of “Academic” Epidemiology and Translational Epidemiology
 

Reader Comments:
Have a thought or comment on this story ?  Fill out the information below and we'll post it on this page once it's been reviewed by our editors.