As more and more physicians adopt electronic health records, the debate over whether electronic health records actually improve the quality of care has risen to a new crescendo. Yet the discussion is not shedding much light on the key issues.
In the latest tit for tat, a new study in the Annals of Family Medicine found that type 2 diabetes patients in practices using paper records achieved better intermediate outcomes than did patients in practices with EHRs. In contrast, a recent paper in the New England Journal of Medicine showed that EHR-based practices provided better treatment and produced better outcomes for diabetes patients than did paper-based practices. And a paper presented at the recent American Association of Clinical Endocrinology meeting found that the use of an insulin order set in a hospital EHR improved glycemic control for hospitalized patients with diabetes.
These studies are not strictly comparable. The first two were based on ambulatory care, whereas the third looked at inpatient care. Also, the AFM study used a far smaller sample than the NEJM study did and examined data from 2004-2006, while the NEJM dataset was from 2009-2010. And the EHR-based practices in the AFM study were using inadequate or no decision support tools, while those profiled in the NEJM paper had relatively good decision support.
What the latter comparison suggests is that, with the right kinds of clinical alerts, some EHRs can help physicians improve the quality of care. But these EHRs must be used in ways that promote that improvement. For example, doctors must enter structured data into the system, rather than dictating their notes, for the decision support tools to be helpful. Also, the EHR data must be available to all members of the care team and must be used in care coordination, as it was in the practices studied by the NEJM researchers.
So, at least in this case, it would be incorrect to say that the research does not clearly support the value of EHRs in quality improvement. What these two studies show instead is that, depending on the capabilities of an EHR system and on how it is used, it is possible to use an EHR to provide better care to patients with diabetes.
A similar controversy surfaced a year ago when David Blumenthal, then National Coordinator of Health IT, fired back at a Stanford University study that found that the use of EHRs did not improve patient care. Again, the researchers used old data (in this case, spanning 2005-2007). Blumenthal argued that EHRs have become better since the Meaningful Use requirements were instituted in 2010.
The damning part of the Stanford study, however, was its conclusion that "clinical decision support also was not associated with higher-quality care. Among EHR visits, CDS was associated with better performance on only 1 of 20 ambulatory care quality indicators."
This conclusion was in line with a recent meta-analysis of 148 studies that found little evidence that the use of CDS improved patient care. But, while the meta-analysis discussed physician acceptance of CDS and how often they accessed the alerts, the researchers noted that few of the studies addressed these issues. And none of them apparently discussed how physicians actually used their EHRs.
Jonathan Weiner, a professor of health policy and management at Johns Hopkins Bloomberg School of Public Health, recently told iHealthBeat, "The majority of information in EHRs is largely inaccessible [for quality measurement]," because doctors tend to dictate into their EHRs rather than using point and click templates to enter discrete data. Moreover, a study of primary care practices at Partners Healthcare in Boston noted that a large percentage of diagnoses were missing from problem lists in EHRs because clinicians didn't enter them.
All of this raises an important point that could help inform the debate over whether EHRs improve patient care: Technology is necessary, but not sufficient to ensure that patients receive appropriate care and have optimal outcomes. - Ken