While I was off in Idaho, Bob DiLaura of Cleveland Clinic left a very insightful response to an earlier post of mine, Clinical Care Informatics vs. Research Informatics. Comments often go overlooked by blog readers, especially on older posts, and this one deserves a better fate. I'm copying it here along with my response to his insights.
Bob wrote:
Hospital HIT organizations have a fundamentally different mission from a research informatics group. The former must first perform like the power company, providing consistent, standard and highly available IT services that support patient care. They can push the curve be offering features beyond commercial installations with highly customized configurations, but beyond that they run the risk of a single organizational group losing focus. Their progress should be slower and wih greater emphasis on detail and contingency planning (and incurring the additional associated costs).
Research informatics groups often look (for better or worse) at each new clinical study as a one-off coding challenge. Each investigator is unique, and wants something different that should be accommodated as much as possible. Our challenge is to be more sensitive to the ROI for programming efforts (a stress often felt by the patient care HIT folks, but not by research informaticians), with better reuse of code and data models that already exist.
The inherent competitiveness between the institutional IT group and a research informatics group may be necessary and healthy if understood and managed at the organizational level for net benefit. Having the central IT group manage research data (for example in an EHR/EMR) may be impossible due to current system design constraints and given the way current research studies are designed seeking uniqueness. Likewise, having a research informatics group design systems that support healthcare indirectly but may be useful for future research needs is a grey area to read that likewise conflicts with the core competency and mission. Both groups add value and need to co-exist in some degree of harmony.
As efforts continue to build a National Health Information Network (the backbone that will enable exchange of healthcare information) and Personal Health Records (the ability for individuals to have granular control over their own protected health information (and thus shift the question of "who owns the data" into the hands of the individual that the information represents), the lines between institutional IT and research IT will blur even further. And this might be a good thing, but will call on the centralized resources to become more innovative and agile, while the research-focused resources will need to operate more professionally and efficiently under the competitive pressure to use limited funding to meet dramatically increasing demands on all technical staff time.
My response:
Hi Bob,
Thanks for adding your insights to this post - you have covered many of the nuances I said I didn't have time to go into, more eloquently than I would have.
It's good to hear others exploring the dilemmas and paradoxes inherent in the intermingling of healthcare and research IT. I agree the central hospital IT infrastructure would be put at risk by attempting to support researchers at the same level they support the clinical care mission. We at UMich are moving toward making available a centralized IT infrastructure as an option for researchers who are ready and willing to move in that direction, or who are compelled to do so by reason of increasingly stringent security requirements being imposed by sponsors. Our NIH Roadmap contract requires an IT security plan that is at least as stringent as what we now have in place for our clinical care IT infrastructure. While this has been a burdensome requirement to meet, it lays the groundwork for an infrastructure capable of supporting both missions equally well.
The issue of researchers' unique requirements is another with which we have struggled overe the past few years during my tenure here. We had a great homegrown solution developed in one of the more far-seeing researchers' labs that in many ways was like the Microsoft Access of clinical trial management systems - unfortunately it turned out not to be scalable due to our inability to devote resources required to match the immense amount of user experience analysis and refinement Microsoft put into its product. We also lacked the support resources that were needed to add the ancillary features such as training and documentation required to create what Geoffrey Moore would call a <a href="http://en.wikipedia.org/wiki/Whole_product"><i>whole product</i></a> that could inculcate a "compelling reason to buy" in the minds of the vast majority of potential clients, most of whom are uninterested in information technology per se. We are now trying to work with <a href="http://www.veloseresearch.com/">Velos</a> to morph their eResearch CTMS into ther kind of product we need. We'll see how that goes; I'm optimistic due to the Velos team's commitment to the academic health center market.
I have mixed feelings about NHIN and EHR. I agree that if we can make them work, they will make integration of research and clinical care IT more seamless. They will definitely give individuals more control over their healthcare data, but from the perspective of research they will not eliminate the need for informed consent - nor should that be a goal, given the spotty record of biomedical research with respect to research ethics.
Ideally a fully implemented EHR/NHIN would give us a lot better tools for the management of informed consent; one of the most common headaches in clinical research (as I'm sure you are well aware) is knowing who has consented (or withdrawn consent) for what, using which IRB-approved version of the consent forms, and with what limitations imposed (the last when using cafeteria-style consent forms). However, from what I've seen and the discussions in which I've participated, the EHR world is dominated by the needs and whims of the clinical care enterprise, who we both agree is not and should not be interested in research IT. I'm afraid the research benefits of EHR will arrive with a second- or third-generation EHR - unless we can bind our research and clinical care missions tightly together in the very near future.
--Dale
Comments