Society for
Epidemiologic Research (SER) Presidential Addresses
Outgoing SER President
Critical of NIH
Makes Recommendations to
Improve Funding Procedures
“T here’s good news and bad
news” that’s how outgoing SER president Elizabeth
Barrett-Connor described the current status of funding for
epidemiologic research. Speaking at a plenary session of the 17th
annual meeting of the Society for Epidemiologic Research held in
Houston in mid-June, Dr. Barrett-Connor noted that there has been a
1.5 fold increase in real funds over recent years, and that “things
are not as bad as they seem, although they could be better.”
Among the other encouraging
developments, she pointed to the growth of non-student membership in
the SER now standing at approximately 1400, and to the increased
recognition of the value of epidemiology and the consequent increased
use of epidemiologists as consultants on research projects. She also
noted that recent developments relating to diabetes epidemiology and
to organizational changes at NHLBI and at NCI were favorable to
epidemiology.
The Bad News
The majority of Dr.
Barrett-Connor’s remarks were intended to stimulate changes in the
current funding procedures, and she focused on several existing
problems. First, she pointed out the costliness of multi-center
epidemiologic studies designed usually to test one question. Since
private organizations rarely fund epidemiologic research, she observed
that basically the NIH “is the only game in town,” and she criticized
its unwillingness to approve add-on projects to large studies which
would help individual investigators. She was also critical of the NIH
for being too big, not sufficiently innovative, and for lacking
inter-institute coordination.
Perhaps her harshest criticisms
were leveled at the extra-mural advisors. Although she described some
of these as excellent, she noted that most are not epidemiologists
themselves, and they are too powerful. They were accused of rarely
providing positive reinforcement and of acting as if the funds in
question were their own money.
Recommendations
Observing that it is much easier
to say when something is broke than to tell how to fix it, Dr.
Barrett-Connor nevertheless went on to make specific recommendations
to improve the current funding procedures for epidemiologic research.
Among her recommendations were the following: 1) Create a new separate
institute for epidemiology, public health, and preventive medicine; 2)
Require at least an MPH or equivalent for NIH administrators; 3)
Increase the proportion of investigator-initiated research and
decrease the proportion initiated by the agencies; 4) Decrease NIH’s
responsiveness to Congress and increase the agency’s awareness of
criticism from scientists; 5) Decent- ralize the sources from which
funds can be obtained, including perhaps certain state and local
agencies; 6) Allow CDC to sponsor more investigator-initiated
research; and 7) Sponsor more senior scientists to free them from the
burdens of peer review and pressure to publish.
Published June 1984
Postscript 2000
Fifteen years later,
there is still good news and bad news with regard to NIH funding of
epidemiologic research.
Complaints:
The cost of the increasing
number of large multi-center contract studies is still a problem and
both the cost and the number are increasing. But the ability to answer
more than one question has made this approach more cost effective. The
incredibly expensive Women’s Health Initiative is likely to be well
worth the money in terms of the variety of novel hypotheses that can
be addressed using these data.
Although I no longer believe
that really big studies are bad, they do sponge up limited NIH
resources and R01 funding. Moreover, very big studies tend to be less
innovative--the sheer numbers of known covariates that must be
transmitted to the dataset discourages add-ons and innovations.
Innovative research is risky
(meaning risk of failure, not harm to patient). “Not sufficiently
innovative” remains a problem for the NIH as recognized by the last
Director and recent mandates to study sections to look for novel
ideas. (Of course, the external reviewers don’t like novel ideas very
much...more about that below).
One way to get both big and
novel is add-ons to test novel ideas. Add-ons to existing or new large
studies two obvious advantages: new information can be obtained at
relatively low cost using the already established study cohort, and
substudies by junior investigators can provide experience and
publication opportunities--essential for the academic survival of
these unsung heroes of studies.
Inter-institute co-ordination
has certainly improved, in part because the large grants need more
support to exist and thrive.
Remedies:
1) The suggestion that a new
separate institute for epidemiology, public health and preventive
medicine be created has not been implemented, or even considered as
far as I know. Fortunately, the Centers for Disease Control has
assumed some of this role, with somewhat more (but still inadequate)
extramural research funding and affiliation with the Emory School of
Public Health.
A different kind or
epidemiology, perhaps less applied and more etiologic, is now thriving
in some Institutes at the NIH--this requires someone with a senior
leadership role who insists on its importance and support (e.g.,
Maureen Harris at the NIDDK). Some other institutes
have leaders who believe that the only important research is at the
bench, preferably the molecular or genetic-level, and some use their
own research priorities to deny funding for epidemiologic studies with
highly favorable extramural review and high priority scores.
There is more work to do here,
both at the NIH and in the general scientific community. We need to
speak for the value of genetic epidemiologists who can improve
awareness that we will miss or misunderstand the role of common genes
if we fail to consider gene-environment interactions. A recent issue
of the American Journal of Epidemiology was devoted to environmental
hazards and health, another area where epidemiology, not genetics or
molecular biology, is likely to lead the way.
2) An MPH for NIH administrators
still seems like a good idea...but one whose time has not yet come.
Preferably an MPH in epidemiology, which is, I believe, the basic
science of good clinical research.
3) The proportion of
investigator-initiated research at the NIH has improved. Two caveats
which may limit future progress. The NIH can fund numerous RO1 “bench
studies” for the same dollar cost as a single case-control study. And
Collaborative Agreements count as investigator-initiated research,
although considerable “advice” provided about the question and the
cohort are provided in the program announcement and by the post study
section intramural reviewers.
4) It was naive to suggest that
the NIH decrease its responsiveness to Congress, from whence its
funding flows.
5) Research money has been
decentralized and the NIH is no longer the only game in town for
epidemiologic studies, especially when the study is a large clinical
trial. Such studies are large enough to provide both efficacy data and
a placebo group--the latter can be used to study the risk factors for
and natural history of the outcomes. Concerns about the excessive
influence of industry money are real, but industry has adopted and
supported clinical trials and foundations have supported epidemiology
and preventive medicine, with largely positive results despite
concerns about conflict of interest. In some states, tobacco money has
supported excellent investigator-initiated research on medical
conditions related to smoking.
6) As noted above, the CDC is
sponsoring more investigator-initiated research, but needs more
funding and freedom.
7) Freeing senior scientists
from the burden of peer review has been accomplished in part with the
MERIT awards. These awards allow established investigators to obtain
funds for research using shorter grant applications, and are a step in
the right direction. MacArthur “genius” awards would be an even better
idea--no strings attached and time to think as well as produce data.
8) Recent funding increments
have been the most positive change in the NIH-extramural research
association. Nevertheless, the percent of grants being funded in the
next few years will not increase.
|
