Last week a new article from The American Journal of Medicine entitled, “Hospital Computing and the Costs and Quality of Care: A National Study” by Himmelstein, et al. appeared in my Twitter stream. In fact, Brian Ahier (@ahier), whom I and about 3300 other tweeps like me follow, sent me a DM asking for thoughts. In that article the authors sort of breathlessly conclude that current hospital computing has minimal impact on quality and no impact on cost. Shocking. Actually, it’s the kind of gotcha article that really grates—the kind that isn’t particularly helpful to anyone as the authors seem intent on drawing sweeping conclusions from pretty limited data.
For starters, how can we draw any conclusions about the impact of widely adopted, meaningfully used electronic records until they’re, well, widely adopted? As research by Ashish Jha et al. highlighted in Chapter One of the recently released 2009 RWJF HIT Adoption report (results from that research also published in April NEJM) show that only 1.5% of hospitals have a comprehensive EHR system—and only another 8% have a so-called basic system. I’m not sure how one can draw important conclusions about national hospital computing given such an unbelievably low national rate of adoption.
More importantly, though, most do not think that simply adopting, even widely, a technology would ever magically on its own improve quality or lower costs. I’m not sure why these authors seemed to say otherwise. The point as many have noted over and over again is for health professionals to adopt and then USE (remember our year long discussion regarding meaningful use?) the technology FOR improved quality, including improved efficiency. As I discuss extensively in Chapter Five of the 2009 HIT Adoption Report, one important use of the technology will be, for instance, automation of performance measurement and public reporting. The automation enabled by widely adopted, meaningfully used EHRs will hopefully accelerate the creation of results oriented information—information that will facilitate payment reforms as well as improvement. We absolutely cannot do the kind of payment reforms that the nation needs without creating measures from the automated collection and aggregation of clinical data. Bundled payment reforms, like the Prometheus model for example, will not work without this kind of automation—and to get to that automation we need widespread adoption and meaningful use of the technology. But the adopted technology is only an important step.
We are still nowhere near that kind of use in an environment of ubiquitous electronic records. Given that small fact, to conclude that there is “evidence” that hospital EHRs “don’t” improve quality or lower costs seems pretty silly. That’s like saying a stethoscope should, but shockingly doesn’t, improve the quality and cost of care just because an intern buys one and hangs it around her neck.
The authors end their article asserting that predictions about cost and efficiency improvements from widely adopted EHRs “are premature, at best”. To me that statement is pretty disingenuous, at best. What’s premature is expecting magic transformation when folks are just getting the tools out of the box—and trumpeting the lack of that magic transformation as if it’s evidence.
Michael W. Painter, J.D., M.D., is a physician, attorney, health care policy advocate, and 2003-2004 Robert Wood Johnson Health Policy Fellow. He is currently senior program officer and a senior member of the RWJF Quality/Equality Team.