Testimony on Comparative Effectiveness Research
Harrison Spencer, Association of Schools of Public Health (ASPH)
Comparative Effectiveness Research Priorities
National Advisory Council for Healthcare Research and Quality
April 3, 2009
I am Dr. Harrison Spencer, President of the Association of Schools of Public Health (ASPH), which represents the Council on Education for Public Health (CEPH) accredited schools of public health in North America.
To help place my comments in context, I want to emphasize that the schools of public health focus on a broad of biological, psychological, social, organizational, and environmental factors that affect health. Thus, you can think of ASPH as representing entities that study the conditions and behaviours that affect the health of each and one of us and consider that achieving health is much broader than just clinically preventing and managing disease.
I thank you for allowing me to share a broader perspective beyond clinical screening, diagnostics and treatment to a view of the impact of health on entire populations. Comparative Effectiveness Research is also for ensuring that policymakers, practitioners, and other decisionmakers can make informed, evidence-choices about the most effective prevention and wellness interventions for use in their particular clinical states, communities, and worksites. The goal is to ensure the greatest health impact despite often scarce resources.
Organizations that have been developed to provide Comparative Effectiveness Research guidance in other countries, such as the National Institute for Clinical Excellence (NICE) in the UK, and similar models in France, Germany and Australia, have adopted broader foci that include disease management and prevention, vaccines for immunization programs, and health care system organization (Chaikidou et al).
I will focus on three major points:
I will elaborate briefly on these points:
My first point is the need to focus on a broad range of factors affecting health because several prominent definitions of Comparative Effectiveness Research emphasize clinical interventions. The IOM (2007), for example, has focused on the comparison of diagnostic or treatment options.
The ultimate goal of Comparative Effectiveness Research should be to inform clinical or health policy decision making to improve health. Although clinical decisions and clinical providers affect health, there are a wide range of policies and interventions that often affect health much more than clinical or pharmaceutical interventions. State and local health departments, worksite health specialists, and other policymakers and practitioners regularly indicate that they need more information on the effectiveness of a wider range of prevention and other public health interventions than the Guide to Community Preventive Services has been able to conduct to date.
Thus, we think it is important to include specific programs and interventions, organizational and system characteristics, and policies and regulations that affect health.
Examples of questions we need to answer with data from rigorous studies include:
Contrary to the perception of many, there is a strong emphasis on evidence based practice in public health and that emphasis has increased dramatically in recent years (Jenicek, Brownson et al., Glasziou and Longbottom, Satterfield et al.). Although some think that these kinds of questions are less amenable to study than clinical procedures there are excellent studies in each of these and related areas, but there should be more.
When prioritizing among the wide range of evidence-based policies and programs, decision makers with responsibility for the public health of their communities or worksites may need to consider different analyses that fit within the broad rubric of Comparative Effectiveness Research.
My second point relates to the terms of the research designs to be supported and used. For many programs or policies that have a profound impact on health, it is either infeasible to conduct randomized trials and/or trials have not yet been conducted. Furthermore, well controlled randomized trials, although obviously very informative, typically have serious problems of external validity. That is, randomized trials frequently enroll a much more restricted population than the one we would like to make inferences to. For example, women, certain racial groups, and children have frequently not been included in trials of treatments that would be applicable to them. This problem is perhaps even greater when one wishes to evaluate programs or policies focused on non-clinical
Thus, in addition to randomized trials, a new Comparative Effectiveness Research agenda, should consider designs are more common for evaluating more comprehensive, population focused, interventions, such as cross-sectional studies, quasi-experimental designs, and time series analyses.
Finally, we need to support both the generation and synthesis of the kinds of data I have alluded to. The basis of Comparative Effectiveness Research needs to be rigorous research. More research on population focused needs to be supported, including research on population programs and interventions, organizational system characteristics, and regulations.
There currently are mechanisms and sources of support for behavioral interventions, and certain types of program evaluation. However, there needs to be more support for system, program, and policy interventions.
In other countries, the organizations that make recommendations based on comparative effectiveness research do conduct the research rather they synthesize available data. Thus, as you consider designing a new Comparative Effectiveness Research organization, consider the capacity for synthesizing multiple types of data, other than just randomized trials, including observational cross-sectional designs, and studies, quasi-experimental series analyses, and time series analyses.
Chaikidou K, Tunis S, Lopert R, Rochaix L, Sawicki PT, Nasser M, Xerri B. Comparative Effectiveness Research and Evidence-Based Health Policy: experience from four countries. Ihe Milbank Quarterly, 2009; In press.
ASPH, 1101 15th Street N.W., Suite 910, Washington D.C. 20005