Virtual Lab Project Web Conference
Remarks by Carolyn M. Clancy, M.D., Director, Agency for Healthcare Research and Quality
Partners Meeting, Los Angeles, CA
April 14, 2010
Good morning, everyone. I cannot begin to tell you how happy and excited we are to see this partners meeting of the Virtual Lab Project taking place.
Health care is about much more than biology and genetics, and your work to address gender inequities and other disparities that exist in health care is work that we are proud to support. I only wish that I could be there with you to hear the conversations in the different work groups, and learn first-hand about the ideas all of you have on creating a unique, federated data system for women's health and health disparities research and quality improvement work in California. Unfortunately, my schedule did not permit that, so I very much appreciate this opportunity to "beam in," so to speak, and spend a few minutes with you before you begin your work.
To provide a little historical context, medical education and research were based on the premise that men and women were pretty much the same, with some obvious differences. The consensus was that any other differences could be explained by environmental and/or societal factors. Medical students were taught about the famous "70 kg man," which was the model for all human physiology—male and female—except for organs found only in women. When studying diseases that affected both men and women, researchers generally studied only men. In fact, the first randomized trial of estrogen to prevent heart disease was done in 1963.
The study group? Men.
When men randomized to estrogen were found to have more heart attacks, the medical community abandoned giving estrogen to men to prevent heart disease. There was no study on estrogen's effect on heart disease in women in a randomized trial until The Women's Health Initiative was conducted 35 years later.
By the mid-1980s, experts had begun to question this paradigm. By then, it was known that females were nine times more likely to develop systemic lupus erythematosis (lupus) than males, that females recover language ability better than males after a stroke, and that females are much more likely to develop life-threatening ventricular arrhythmias in response to certain drugs.
In 1986, the National Institutes of Health issued a policy that encouraged the inclusion of women in clinical research, but audits by the General Accounting Office in 1990 and 2000 found that analysis of data by gender was still limited.
In 2001, the Institute of Medicine published "Exploring the Biological Contributions to Human Health: Does Sex Matter?" This report found that the scientific study of gender differences was essential to improving human health. However, the IOM also found that "scientists conducting research on sex differences are confronted with an array of barriers to progress, including ethical, financial, sociological, and scientific factors."
As noted in a commentary by Kazanjian and Hankivsky in Women's Health Issues in September of 2008, there is a need to integrate all the factors that influence health among women in research, and this should be a key priority. Integrating these factors may be difficult or impossible in standard clinical research, because standard clinical trials are limited by a variety of factors.
And last June, Kaiser released a study entitled, Putting Women's Health Care Disparities on the Map: Examining Racial and Ethnic Disparities at the State Level. It was a powerful reminder of the gender inequities that exist in health care. At the State level, it said, quality and disparities are related but distinct concepts. High quality does not guarantee low disparities and vice versa. Therefore, both dimensions need to be measured and tracked.
States vary widely in health care quality and disparities. Some States perform well or poorly on both dimensions. However, some have high quality and high disparities, and some have low quality and low disparities. For example, the Kaiser study found that:
- Massachusetts was best among States involving the share of women who did not get a mammogram (16.3 percent), but the percentage of women of color without a dental checkup was about 80 percent higher than that of white women.
- In Oklahoma—generally a worse-than-average State—white women and women of color both experienced significant problems with access to care. Hence, disparities tended to be smaller than at the national level.
As this study points out, disparities exist in every State on most measures. And, in States where disparities appear to be similar, the reason often has nothing to do with progress. It's because all of the women—white women and women of color—are doing poorly.
So, what do we do about all of this?
Integration of large health databases for the purpose of research offers a cost-effective method of playing "catch up" for women and minorities in areas of comparative effectiveness research where prior studies, mainly clinical trials, were done primarily involving white men. Integration of large health databases for the purposes of quality of care and comparative effectiveness research will allow investigators to study health across entire populations. Evaluating the impact of comparative effectiveness research is a key priority for the Obama administration and, as the lead Federal Agency in this discipline, AHRQ is playing a major role in expanding the use of this research nationwide.
Comparative effectiveness research is all about identifying which treatments work best for which patients and how we make the right thing to do the easy thing to do. These two elements have got to be linked. To do this in a meaningful way, it will be critical to link data that is being repurposed for use in medical research, such as administrative claims and EMR data to systems that contain gender, race, and ethnicity information. And, to link at the patient level to data in rich population-based surveys such as the California Health Interview Survey and the comprehensive hospital discharge data that is collected by the State of California. California, because of its diverse populations, varied health care delivery systems, active and experienced health services researchers, and sophisticated information technology, is an ideal laboratory.
The Virtual Lab Project can test the feasibility of integrating large health databases for the purpose of building an ongoing, multifaceted research effort in women's and minority health. This is the type of effort that can go beyond the siloed "single disease, single project" model that has characterized previous integrated large database efforts.
I applaud your efforts to leverage this planning process with stakeholders from the Right Care Initiative. At AHRQ, we continually emphasize the fact that collaboration on all fronts is key to making the most of the opportunity that we now have to affect needed change in the system.
Although the focus of the Right Care Initiative is improving the delivery of cardiovascular and diabetes care in your State, both it and the Virtual Lab Project will need to use the same types of data systems and collaborators to be successful. Your decision to pool your efforts and to think in terms of creating one data system that can be used for multiple purposes, with a uniform set of regulations and data access agreements, is completely on point.
In your effort to link people's health insurance, benefits structures, and claims paper trails to their population-based data, you have an opportunity to create a unique system that can help us better understand women's health, health disparities and minority health, and the determinants of outcomes for persons with chronic conditions such as cardiovascular disease at a much higher level than we've reached previously.
Of course, data collection is a costly process. Historically and generally speaking, data sets are created and used by academia, providers, and State organizations to answer a limited range of research questions. This Virtual Lab model will help create a prototype for connecting major databases across academia, provider organizations and State-level health organizations across the State of California. This will lead to the creation of a rich and robust database that will be open to many investigators for answering a broad range of research questions related to women's health and gender differences in health services.
Let me emphasize the importance of having a dynamic, "living" data set in which the lag between acquisition and exploration can be measured in weeks or days, and not months or years, as is currently often the case. This will be especially true if one is looking to evaluate the impact of comparative effectiveness research. This is not about us building a better library. It's about shortening the lag in translation and focusing on effectiveness.
My own personal ambition would be that we can be as transparent about the use of this information for practice and policy as we are trying to be at AHRQ about developing this information. I do not think that is going to obliterate people's fears any time soon. I understand the fears all too clearly. The fear that I think resonates with most of us is, "I might be denied access to something that would literally save my life or make a huge difference based on studies that do not include people that look like me."
We have an opportunity and a responsibility to make sure that the studies include people who look like all of us. And, this is something we fully intend to live up to. But at the end of the day, this all comes back to the fact that right now clinicians and patients are making decisions under conditions of enormous uncertainty. That is not the only reason undergirding our challenges in health care, but it is a big, big part of it. And if your model proves to be successful in California, AHRQ would look to promote its adoption in other States.
The ultimate goal is to create databases that are relevant and useful at the local and State levels and broad enough to be connected at regional and national levels to answer questions about women's health services and gender research.
This is very important. We need to continue to push the message about eliminating gender inequities and other disparities from our health care system, every step of the way. Because any attempt to enhance the U.S. health care system will be much less than acceptable if we fail to reach the people who need us the most.
Thank you. I very much look forward to hearing about the results of this meeting.