Congress and the President Pass NSF Authorization Act
Behavioral research accounted for over $2.1 billion worth of the NIH budget in 2001, but this success, for some reason, has not extended to training. Yet the pendulum may be swinging in the other direction now. The Committee for Monitoring the Changing Needs for Biomedical and Behavioral Research Personnel of the National Research Council has recently begun its quadrennial review of research training at NIH, and invited APS to make recommendations on how to improve this highly important field. The committee faces a heavy responsibility, as this group is required by Congress to make recommendations to the National Institutes of Health, every four years, on how to improve the quality and depth of research training. This is not the first time APS has testified before this body- you can read the most recent testimony, as well as our testimony from 1998, here.
Testimony of the American Psychological Society
Alan G. Kraut, Executive Director
Committee for Monitoring the Changing Needs for
Biomedical and Behavioral Research Personnel of the
National Research Council of the National Academy of Science
Mr. Chairman, Members of the Committee:
Thank you for inviting me to discuss NIH research training. I want to touch on a number of issues today -- the role of behavior in health; a history of behavioral recommendations from this committee; and Congressional interest in research training. First, though, I want to provide some of my own historical context.
APS and Research Training
Research training and support for young investigators has been a core issue of the American Psychological Society since our founding nearly fifteen years ago. I have testified in front of several of this committee's predecessors, and we have raised the importance of training before Congress and federal agencies in countless meetings and in our more formal communications. As one example, we were the only science organization insisting that training money accompany NIMH, NIAAA, and NIDA as those agencies transferred from the old ADAMHA to NIH in 1992, which was a controversial topic at the time. Since then, we have consistently counseled each director of the NIH Office of Behavioral and Social Sciences Research (OBSSR) that research training is an issue on which OBSSR could make its mark as it sought visibility at NIH. And we've worked with NIH over the years, in the hopes they would become more sensitive to training for behavioral scientists.
One result was that we were invited by Science Magazine (thank you Floyd!) and other outlets to write about the importance of research training, behavioral and non-behavioral. And even before APS came into being, I was a large part of psychology's effort to influence training at NIH from 1980 to 89. One personal note: I am delighted to see Georgine Pion on this committee, since she was the real brains behind what we had to say about psychology's training in the late 80's. Finally, on an even more personal note, I was an NIH NRSA trainee many years ago and know first-hand what this beginning can mean to a career.
Say It Ain't So
I wish I could say these efforts made a critical difference in NIH training for behavioral scientists over the last fifteen years. Don't get me wrong - I think APS and the behavioral science community has had a major impact on support for behavioral research at NIH. At your last meeting, Raynard Kington, then-Director of the NIH Office of Behavioral and Social Sciences Research, presented a slide on the amount of NIH research funding for behavior. It totaled over $2.1 billion in 2001: $359 million at NIMH; $343 million at NIDA; $238 million at NCI; $222 million at NICHD; etc. Those all represent dramatic increases from 15 years ago, and I am proud to say that APS played a large role. I would even argue that even OBSSR itself would not have come into being without the support of APS.
But, in preparing for today's presentation, in going over past statements before this committee, and in reviewing this committee's past recommendations for behavioral science and how they have been interpreted by NIH, I am forced to say that the community of behavioral scientists, including APS, previous committees, and yours truly have not had the measurable impact on NIH behavioral science training policies that we should have. I'll give my thoughts in a moment on why not, but I want to emphasize that I absolutely believe that we can have an impact this time.
You are a newly formed committee. For the first time, you have a Vice-Chair who is a behavioral scientist. For the first time, you have a critical mass of behavioral scientists as members, a subcommittee on behavior you could say. You have new staff. There is a new Director of NIH, who has talked openly about the importance of behavior. The new NIH Deputy Director is the former Director of OBSSR.
The Importance of Behavioral Science: Basic, Applied, Clinical
There is another reason I am optimistic. It is because of one overriding "truth we hold to be self evident" at APS, and one I believe other decision-makers in health will recognize eventually. It has been documented in official report after official report, some from this very Institute of Medicine, some from the U.S. Department of Health and Human Services, some from the Surgeon General of the United States, some from Congress. Different reports; the same conclusions:
In any realistic picture of our Nation's health, a core finding is that behavior is central to many, maybe to most of our Nation's leading health concerns: heart disease; stroke; lung disease and cancer; obesity; AIDS, suicide; teen pregnancy, drug abuse and addiction, depression and other mental illnesses; neurological disorders; alcoholism; violence; injuries and accidents -- all have large behavioral components. It was IOM that first documented that 7 of the 10 leading causes of premature death in this country are due in large part to behavior and account for over 50% of mortality.
This is a point I want to make today: It is the enormous role that behavior plays in health that should lead this committee and NIH to promote training in behavioral science. If we offer our best young scientific minds the opportunity to train in behavioral science, with all the sub- and cross- and inter-disciplinary perspectives that implies, we will be addressing many of the daunting health concerns facing our nation.
I know this committee often uses an economic employment model when deciding what recommendations to send on to NIH. Will there be jobs for the PhDs that NIH is producing? I have to say that I do not understand this approach. First, health research is based in large part on its importance to public health. The National Cancer Institute is a $4 billion-plus agency not because there are thousands of cancer researchers to fill jobs. NCI funds $4 billion in research because cancer is properly recognized as an enormous public health problem. The same should go for all the disorders that have behavior at their core. These conditions can't be addressed fully without behavioral research, in all its basic, applied, and clinical forms.
Let me pause here to emphasize a point about the range of behavioral science. When I talk about behavioral research related to health, I not only mean applied or clinical research. For all the health concerns I have cited, there is a basic science of behavior that needs to be in place before application. Before you address how to change attitudes and behaviors around AIDS, you need to know how attitudes develop and change in the first place. Or, before you can change decisions about any risky behavior, you need to know how judgments and decisions are made on a range of topics. Similarly, before you address memory decline in the elderly, you need to know the basics of learning and memory and how that changes with age. And before you address the complexity of the interactions among genetics, the brain, and, say, schizophrenia, you need to know the basics of cognition, emotion, behavioral aspects of neuroscience, and behavioral genetics.
I believe this Committee would agree with me here, but I cannot tell you how hard it has been to convey the importance of basic behavioral science in any sophisticated sense to high ranking, non-behavioral administrators at NIH. And I hear this complaint from behavioral scientists at NIH. Not to simplify this too much, but my sense is that when the NIH leadership think of behavioral research, with few exceptions, they think of applied research. A vaccine is ready and now the behavioral question is how to convince parents to bring their children in for inoculation. But, just as the development of that vaccine was built on basic research in virology and molecular and cellular biology, a persuasive message to parents builds on basic research in cognitive, developmental, and social psychology.
The same is true for questions such as: How do you get TB patients to stay on their course of treatment? What are the behavioral aspects of controlling blood pressure? How are minorities disadvantaged in access to health care? Or, even more cross disciplinary questions like, what about combined behavioral and pharmacological treatments for depression or anxiety disorders? Or, more generally what do we need to know about basic behavior, biology, and their interaction to better understand Alzheimer's, or diabetes, or deafness, or SIDS, or learning disabilities, or any of the hundreds of disorders NIH studies?
Let me make this NIH disconnect between basic behavioral science and health even more concrete. Cognitive psychologist Danny Kahneman of Princeton University was awarded the 2002 Nobel Prize in Economics. His research on decision-making has enormous implications for health. Some of his and Amos Tversky's early work was focused on judgments of pain in colonoscopies, but their work also has straightforward implications for issues of patient compliance, deciding when to see a physician, how physicians make diagnoses, etc. But Kahneman and Tversky never were funded by NIH. Such an oversight simply would not have occurred in any other area of research so directly related to health. (P.S. Now, Kahneman is being very courted by several NIH institutes)
(And if you like irony, here is another twist on basic and applied research. The National Science Foundation supports only basic research. For years, they would not accept an application for an NSF graduate fellowship submitted from a clinical psychology student. NSF does not support clinical science, they told me. But I convinced them that there was also a basic science of clinical psychology. It might involve studying the basics of emotion, or the statistical processes of intervention or prevention. They agreed and removed the ban on clinical psychologists applying for NSF fellowships. The irony is that the Director of the NSF understands that basic behavioral science underlies clinical science. I cannot name you many non-behavioral administrators at NIH who feel the same way.)
Okay, back to an employment model. There is one other reason I fail to understand the use of an employment model by this committee when formulating its recommendations. NIH strives to produce and support the elite of health researchers, not all health PhDs. A competition takes place in every NIH study section several times per year on who gets funded. Study section members do not and should not worry that marginally acceptable or unacceptable applications go unfunded. What we all should worry about is maintaining a continuing line of excellent applications. And it is the quality of NIH training that will result in those applications. Bottom line is that this committee should not be consumed with overall PhD production, and where those PhD's go. It should be concerned with how to produce the best of health PhDs with NIH training money. Hence, the importance of the National Research Service Awards (NRSAs).
NRSAs, NIH, and the Committee
That is what sets NRSAs apart. NRSAs may be small in number but they are powerful influences on setting priorities in training. They set trends that are followed by the entire health research training enterprise. I hear often that when a department has an NRSA training grant, it is not only a sign of that department's quality, but of where the future of that particular sub-discipline of psychology is heading. A NRSA award is a golden ticket. You already heard Raynard Kington tell you that behavioral scientists who receive NRSA awards obtain their PhDs in fewer years. They are more likely to engage in post doctoral study. They stay in research longer. They produce more publications in peer reviewed journals. They are more likely to apply for NIH grants.
Still, these differences that I might have with you over employment model cannot fully explain why NIH has not responded to your recommendations regarding behavioral research over the years. Even when you haven't followed an employment model, and when you have acknowledged the importance of behavioral factors in health, your recommendations have not had much of an impact on NIH. To be blunt, the past behavioral portions of this committee's recommendations on training priorities have been ignored by NIH going back several reports. (Of course, the fault does not solely rest with this Committee. Recall that I included myself in this failed effort.) But now more than ever, with NIH's doubled budget and research funding at an all time high, we need to succeed. This Committee's recommendations are an essential element in making sure we have those high-quality behavioral science researchers we need to address the nation's health needs.
One change from past committee procedures that would make for a change in impact would be to have your recommendations focus on individual institutes. By way of contrast, consider what happened the last time you encouraged NIH to increase its behavioral commitment to training, even putting specific numbers in your recommendations.
The committee's recommendations to NIH in 1994 called for an increase in awards from 1,069 to 1,450 to be in place by 1996. You recommended a 151 award increase in postdocs; a 228 increase in predocs. Talk about specific!
Yet, in its response, NIH selectively implemented recommendations -- ignoring the behavioral recommendations, but implementing the other recommendations -- by what I would characterize as "convenience." I could go into specifics, but the bottom line was that anything that could be done easily through "Building 1," (i.e., some central NIH structure) - such as increasing stipends - was done. Anything that required more a complicated strategy - such as negotiating something specific out of, say, NIMH - was ignored.
In fairness, NIH does not have a system for implementing training recommendations at individual institutes. The 1994 recommendation to increase stipends was something that could be done at the broader NIH level. Frankly, it was the only part that they could handle easily under the decentralized training structure that has 27 separate NIH Institutes and Centers. Your recommendation to increase the number of trainees would have taken a coordinated effort that both doesn't exist and is not an NIH tradition. That is, I believe that organizational inertia -- a kind of bureaucratic laziness -- was at the heart of NIH's selective implementation.
In order to avoid history repeating itself, this committee must structure a way to make it easier for NIH to implement your recommendations. You know that many NIH Institutes care about behavioral science. Look again at Kington's slide on where behavioral science is funded. Now, think about who you asked at NIH to give you information. In your last meeting, you only had Building 1 talk to you. (NIGMS is the exception, which I will take up in a minute.) Where were NIMH, NIDA, NIAAA, NICHD, NIDCD, and the other institutes that fund the bulk of behavioral research?
I can tell you some of what you'd find if you did explore behavioral training at NIMH, for example. You would find that their take on the basic-clinical disconnect I described earlier led them to a several year effort in which they studied how to encourage basic behavioral scientists to explore more clinical issues, and how clinical investigations might be more linked to basic behavioral findings. They now certainly could use your recommendations on training in this area they call translational behavioral science.
And you would similarly find NIDA exploring how to bring basic behavioral concepts such as cognition, and motivation and craving into drug abuse research. Again, there would now be a clear place for you to work with them on behavioral training.
The story would be similar at NIAAA and NICHD. Child Health, for example, has become a lead federal agency on reading research. But now what they need is to encourage those studying basic processes in visual perception and attention, and in socialization and other basic areas to cross-train their graduate students in to ask applied questions.
But this interaction with individual institutes is not what I see you doing. Instead, it looks as if you are again starting down a road more likely not to result in having any behavioral science impact at the level of the individual NIH institute.
Here is another way to describe the NIH relationship with this committee. On the one hand, there are institutes who care a great deal about behavioral science and who could be guided on behavioral training based on your recommendations. But those institutes have not been particularly connected to this study - either its development or its implementation. On the other hand, there are NIH offices who don't particularly involve themselves in institute-specific behavioral science. There is not much they can or want to do with your behavioral recommendations. Yet, they are the ones most connected to this study.
You still are in the study's early stages. My advice is to continue your fact-finding and connect with the institutes most interested in your task. And within those institutes, don't just go to training officers. Go to the program people most involved with the substance of behavioral science. That's where you are going to learn about cutting edge research and where the next generation of cutting edge researchers ought to focus.
General Medical Sciences
Let me turn to one more issue before I leave. At your last meeting you heard from the National Institute of General Medical Sciences (NIGMS). NIGMS is known as the basic science institute of NIH. You heard Acting Director Judith Greenberg say that 40 percent of all NIH predocs on training grants were supported by NIGMS. 40 percent! That is an enormous amount of NIH's basic science training, and I have been making the case here that an enormous amount of important behavioral science related to health is basic.
Look again at Kington's slide of where behavioral science is supported. NIGMS supports none -- $0.0! I cannot tell you why a $1.865 billion institute that supports only basic science supports no basic behavioral science. I am more perplexed when I read the legislative mandate that created NIGMS:
"The general purpose of the National Institute of General Medical Sciences is the conduct and support of research, training, and, as appropriate, health information dissemination, and other programs with respect to general or basic medical sciences and related natural or behavioral sciences which have significance for two or more other national research institutes or are outside the general area of responsibility of any other national research institute." (U.S. Code TITLE 42 CHAPTER 6A SUBCHAPTER III Part C subpart 11 Sec. 285k.)
In fact, NIGMS once did have a tradition of supporting behavioral training. Now, they claim that behavioral research falls outside their mission, and ignore all those who recommend it. Former NIGMS director Ruth Kirschstein spoke before you at your last meeting. She specifically said that there weren't enough knowledgeable staff at NIGMS to oversee behavioral training. I don't know about you, but that rang hollow to me. Don't you think a nearly $2 billion agency can acquire that knowledge?
The issue of NIGMS and its snubbing of behavioral science research training has been a repeated subject in both chambers of Congress over the past five years,. When the Senate Appropriations Committee gave NIH its budget for 1999, they encouraged NIGMS "to support behavioral research training as part of its mandate to support basic research training in all areas of health and related research." In 2000, the House Appropriations Committee specifically linked NIGMS and behavioral research training, encouraging the institute to support behavior, and to consult with the science community and other institutes to identify priority areas. The Senate Appropriations Committee spoke again, expressing "concern" that NIGMS did not support behavioral science research training. They urged the institute to develop a plan for behavioral research in consultation with OBSSR.
The Senate repeated these suggestions in their Appropriations reports for 2001 and 2002. In the 2003 report, the Senate went one step further, encouraging "the NIGMS to develop collaborations with other Institutes, such as the NCI and NIMH, and the Office of Behavioral and Social Sciences Research to fund basic research to integrate physiological knowledge of pre-disease pathways with behavioral studies."
More recently, a series of letters were sent to NIH by Rep. Patrick Kennedy and Senator Daniel Inouye. They again cited the lack of behavioral science research training and funding for behavior at NIGMS. While the NIH and NIGMS responses made little references to behavior that might one day be at NIGMS, the main message was still "behavioral studies fall largely outside of the Institute's research mission..."
Over this period (1999-2003), NIGMS saw its budget increase by over $700 million. Twenty-two different NIH institutes and centers fund behavior, from NIMH and NIDA, all the way to Deafness and Communication Disorders, and Dental and Craniofacial Research. Yet, NIGMS, which is mandated by law to address behavior, dedicates exactly $0 to this effort.
NIGMS needs to obey the law and develop a research and training agenda for the behavioral sciences. There is a well of expertise to draw from, including resources and scientists inside and outside NIH. With its sizeable budget and resources, NIGMS should be on the cutting edge of basic behavioral research and training. I hope that in your final recommendations, you instruct NIH to oversee the implementation of a research training program in behavior at NIGMS.
This concludes my statement. Thank you again for allowing me to present these views. I would be pleased to answer questions or to provide additional information on the issues I've raised here today.