Archives for NIH

Today I attended the Institute of Medicine (IOM) Meeting titled “Next Steps: Implementation Workshop on Standards for Systematic Reviews and Clinical Practice Guidelines” to understand better the key steps in going from clinical research to practice.  The entire day was spent with experts, primarily academic, discussing two recent IOM reports about standards for systematic reviews (SR) of comparative effectiveness research (CER) and clinical practice guidelines (CPG), respectively.  I asked several questions with the intention of understanding the current linkage between the panelists’ thinking and health IT standardization and interoperability efforts driven by the Office of the National Coordinator (ONC) for Health IT.  Close to the end, I made a comment that I think they needed to hear about the role of health IT in the future of CER, SR and CPG.

The majority of the attention for the day was on implementing the standards proposed, particularly conflict of interest standards necessary to establish trust in the results of the SR or CPG committees, and study thoroughness requirements which directs SR protocols not only to review the best published material, but also to consider “grey” literature such as unpublished source data from the published authors, conference abstracts. etc.  The IOM will be publishing final conclusions from this workshop that will discuss many of the important points made at this session, and a survey of 38 attendees and committee members to determine the likelihood of standard implementation, the timing of the implementation, and the barriers to full implementation.

In asking the first of my questions probing this issue I was told that informatics was out of the scope of the charge from Congress, so they had not considered these issues in the definition of SR and CPG standards.  No one dared to think about what kind of information might be helpful for more powerful SR, nor did they venture to define how the output from the SR could make it easier for the CPG process.  After my question a response did mention the importance of this effort for clinical decision support, only strengthening my confusion about why the standards for CPG would not include a definition that would be understood by a computer for easy automation in a clinical practice setting.

After summary presentations of the reports and the survey, I asked my second question about how the efforts of various organizations would be impacted and coordinated:

• The newly created Patient-Centered Outcomes Research Institute (PCORI) which will be funding CER and the SR studies which evaluate the CER,
• The Agency for Healthcare Research and Quality (AHRQ), which was funding the IOM report and workshop and has been the primary source of funding for the majority of the SR studies,
• The National Institutes of Health (NIH) which also funds SR studies, particularly at the larger institutes like the National Heart, Blood and Lung Institute (NHLBI) and the National Cancer Institute (NCI); and
• The ONC which is defining a health IT standards and interoperability infrastructure, as well as a new exchange architecture including a universal exchange language (UEL) as proposed by the President’s Council of Advisors on Science and Technology (PCAST) which could serve as a major enhancement for CER, and thus, SR.

In asking this second question, I quickly learned that CER must be kept separate from SR which in turn must be kept separate from both CPG.  So I had to restate that what I was interested in here was the coordination of funding, in particular.  I was told that the Affordable Care Act that established PCORI had a Board of Governors with representation from both AHRQ and NIH.  So that left ONC on the sidelines, though they could have a major impact on the ability of CER to produce standardized results that would make SR and CPG more straight forward.

I sat through the entire afternoon of talks about SR wanting to say something about the changing context in which SR was being standardized by this IOM effort.  Each of four hours was painstakingly used to discuss every important detail of the steps in SR from study design to reporting.  In the last hour focused on reporting I took the microphone and said something like this, or should I say, I wish I had said it like this:

“I’m Tom Caruso from the Biomedical Informatics Think Tank ™ an organization made up of 21 leaders in biomedical informatics and related areas mostly located at major academic medical centers around the country; however, the comments that I am about to make are mine only and do not represent the opinions of BITT ™ or any of its members.  Because a chasm exists between my frame of reference and the frame of reference of all but a few people in this meeting, I am concerned that what I have to say will be dismissed as either not relevant or about a far too distant future.  With that statement, I hope I have partially breached that chasm and you can hear what I have to say.

Everyone in this room would agree that the management of health information is changing rapidly.  We can at least agree on that point.  Now consider how dramatically these changes will change healthcare, and consider how the practicing clinician will soon be using aids to help him or her make personalized clinical decisions.  CPGs will be incorporated into clinical decision support systems and they will guide the physician through every particular case.  You therefore must be thinking about SR reports that include statements of the probability that some outcome, B will result given a set of particular conditions, A.

If you doubt my resolve about the future, which is not so far away, think about how Watson beat two of the all-time winners of Jeopardy.  IBM is taking this technology into the clinic, and it is only the beginning of where healthcare information exchange opportunities will lead us.  We must be thinking about the health IT needs for CER, how CER produces information for optimum SR, and how SR generates information that can be used in these clinical decision support systems that are coming very soon.”