Clinical research can most broadly be divided into sponsored trials, investigator-initiated protocols, and public health/epidemiology. Sponsored trials nearly always test drugs or devices and are meant to provide data to the FDA. They are generally designed and managed by corporations and are typically spread among five-to-forty centers since the FDA requires geographic diversity. Data obtained in sponsored trials is processed internally by the initiating companies and they usually write the resulting manuscript themselves. When accumulating data fail to support the tested product, sponsored trials are often stopped. Since no one center has access to sufficient data, publications rarely result from trials that are cancelled mid-course.
Sponsored research nonetheless brings two benefits to universities. The first is a training in data-acquisition — although not in protocol design, data analysis, or manuscript preparation. The second is that they are usually well funded, and the money they provide can be used to support preliminary investigator-initiated studies that can then serve as a basis for grant applications.
Investigator-initiated research tends to involve a single center and is often innovative. Usually, the principal investigator designs the protocol, manages data acquisition, does the statistical analysis, and writes the manuscript. Unlike clinical trials, this sort of work often leads to peer-reviewed funding. Because investigator-initiated projects involve all aspects of research and tend to be more academic, they are especially beneficial to universities. A secondary benefit is that companies usual distribute sponsored trials to faculty who have already published investigator-initiated studies relevant to their products.
Public health and epidemiology research usually focuses on populations as a whole. Typically, they evaluate the influence of public policy and behavior on broad outcomes. These are usually “natural experiments” or cohort trials, rather than prospective, randomized clinical studies. Controlling bias and confounding factors is thus often the most difficult aspect of these projects. As might be expected from the nature of their studies, faculty in public health and epidemiology are expert in complex statistical techniques.
The academic missions of most university hospitals is supported by a small fraction of their faculty who have advanced research training. These faculty are perfectly capable of doing laboratory or clinical research on their own. Additional research is performed by public health and epidemiology experts who have specialized training in these areas.
In most academic medical centers, however, 80% of the faculty are more than 80% clinical. Typically, many of these clinicians are unsatisfied because they fail to reap traditional academic benefits. They thus eventually feel that their efforts are unappreciated by the institution. This is disruptive to universities, whether they remain as unhappy employees or leave to private practice positions.
It is expensive when clinicians leave the University because costs associated with their recruitment and training is lost. An additional hidden cost is that departing clinicians often take patients with them. As importantly, the private groups these faculty join directly compete with University for patients and procedures. A primary reason for leaving is lack of academic recognition and a feeling of having “failed academics.” Universities will never be able to compete with private-practice salaries. Furthermore, hours are usually worse in academic than private environments. If universities are to retain clinicians, it is thus critical that they be given the opportunity to fully participate in the academic mission of the institution — and that includes the opportunity to do research.
Faculty who have ideas for research projects should find adequate support within the institution. Too often, junior faculty put considerable effort into a first study — which subsequently fails. They are thus denied the personal pride and academic recognition that results from publication. It is usually the last study they attempt. The question, then, is what factors prevent a clinician with an idea from obtaining grant support, successfully completing a project, and eventually publishing a paper?
Most clinicians make observations and ask questions during the course of their practice; many of these questions can be the basis for studies. Furthermore, most of these faculty are enthusiastic and interested in pursuing their ideas. So what prevents ideas from becoming published papers? At least five factors conspire to limit translation of ideas into publications.
Among these, deficient protocol design is the most serious. Protocol design errors are by far the most common cause of failed studies. Furthermore, this is an insidious problem because investigators do not recognize the errors until the study is complete and the resulting manuscript is rejected for publication. At that point, the effort (often a year or more) is completely lost. Few clinicians subsequently attempt another study.
The second most important factor is lack of assistance with manuscript preparation. This is a difficult issue since writing manuscripts — and teaching people to write — is an arduous process. Furthermore, it seems unlikely that clinicians who only occasionally publish will ever become facile writers. We can thus assume that substantial assistance with manuscript preparation will be required whenever research is performed by investigators who are primarily clinical.