Approximately a year ago I wrote a blog for Health Affairs (The Case for Meaningful Use: Strengthened, Not Weakened) which addressed an article in the mainstream medical press that appeared to call into question the value of electronic health records (EHRs). That paper correctly pointed out that older EHRs, without effective and or targeted clinical decision support, without clinical or workflow direction, and without financial incentives for doctors to reach quality and value targets – did not appear to improve quality. While one could have concluded, “why would you expect otherwise;” or “thank you for stating the obvious; this is the reason why the Meaningful Use program is the Meaningful Use program and not the ‘Dollars for Buying Computers Program’;” the study authors incorrectly concluded that their study called into question the Meaningful Use program.
A few weeks ago, several articles came out in another highly respected medical journal, which called into question the ability of EHRs to be reliably used for quality measurement and improvement. One study demonstrated that clinicians underutilized the problem list of the EHR (and thus did not identify the correct number of patients for quality reporting for particular conditions). Another study showed that extracts from the EHR were very reliable for certain measures (where lab data came into the EHR directly via an interface, for example); and unreliable when there was a requirement for manual entry into the EHR. While both of these articles were clear as to their findings, the buzz in the popular IT press incorrectly suggested that these studies called into question a basic premise of EHR use and the Meaningful Use program – the reliability of clinical data from EHRs.
I wish to remind those authors and my colleagues that while the study authors were thoughtful and deliberate, the headlines got it very wrong. The majority of American doctors are at Stage zero of Meaningful Use, and a minority at Stage 1. Stage 1 assumes that providers, staff, patients, family members, etc are beginning their journey in the use of health IT as enabling infrastructure to improve care. And the stated purpose of quality measure reporting in Stage 1 is to show that measurement, aggregation, and reporting out of EHRs is possible; and does not assume that the measurement seen is an accurate reflection of quality. And as many commentators to the Meaningful Use regulations have pointed out, we should look carefully at the experience in each stage to better inform subsequent stage requirements. So perhaps the headlines should have read, “Wisdom of Iterative Approach to Meaningful Use Again Born Out.”
And just last week, a piece was published in Health Affairs “Giving Office-Based Physicians Electronic Access to Patients’ Prior Imaging and Lab Results Did Not Deter Ordering of Tests.” While the title of the article was more or less accurate (perhaps less, as this study was not a prospective study but a review of prior experiences from which the authors drew correlations and not causation); there was a clear misunderstanding of the potential value of health IT expressed in their preamble. The authors stated that “reduced ordering of imaging and other diagnostic studies is often cited as a likely mechanism for cost savings related to health IT.” I am not sure where they are getting their information, but the idea that health IT per se can be reliably associated with anything other than the speed of a process has not been advocated in policy circles for almost a decade. Further, the actual hope was that health IT, if constructed and implemented optimally, and used with an appropriate workflow that was supported with aligned financial incentives, could decrease the ordering of unnecessary and/or redundant testing or ordering. In fact, under the best of circumstances, depending on the patient population and specialty of the physician; one could indeed hope for the opposite (for example, using health IT to help appropriate preventive and chronic care services move from approximately 50% to everyone).
Further in the article, the authors noted that (paraphrasing)“… EHRs at a few flagship institutions utilizing customized health IT with clinical and financial decision support have demonstrated decreased ordering of radiologic studies and other diagnostic tests; but that no one had looked at commercially available products used outside of those flagship institutions (using 2008 EHRs) for similar effects.
So before I state again, “thank you for once more validating the purpose of the Meaningful Use program; except for advanced health systems which had advanced EHRs with embedded clinical and financial decision support (which proved themselves effective); commercial EHRs without clinical and financial decision support did not do what no reasonable person would have expected them to do… And one of the purposes of the Meaningful Use program was to bring such advanced features to the market – so kudos to the 111th Congress, you got it right…” another observation…Sensitivity analyses demonstrated that “neither having an EHR or using CPOE was associated with the likelihood of image ordering.”
So this retrospective non-study, with a raison d’être which misstated a potential value of health IT; found that customized EHRs of 2008 were able to reduce ordering of radiologic studies; that-off-the shelf EHRs in 2008 demonstrated no increase or decrease in ordering; and that the use of web-based image or results viewing (without an EHR and without clinical decision support) was correlated with an increase in ordering. And your point? “We conclude that use of these health information technologies, whatever their other benefits, remains unproven as an effective cost-control strategy with respect to reducing the ordering of unnecessary tests.”
So for the sake of argument, let’s assume that their study was prospective and could draw causal conclusions. And also for the sake of argument, let’s assume that their study also only included orders that were deemed duplicative or clinically unnecessary. The conclusion that a web-based, no context, no clinical or financial CDS enabled, no enhanced or incentive supportive workflow – led to no cost benefits, and perhaps even increased costs. Ok. Thank you, maybe. Concluding that using health IT in a way that no one has ever suggested as useful proves what?
The Meaningful Use program is not without its flaws, and you will see many critical comments about the Stage 2 NPRM in future posts. However, as much as they make for titillating headlines, these recent studies have again strengthened and not weakened the case for Meaningful Use.