Annals of Internal Medicine. However, the study also found that the total cost of care continued to rise in both groups of practices.
In 2006, the practices in the three communities received EHRs and training funded by Blue Cross and Blue Shield of Massachusetts. Of 167 eligible practices, 86% participated in the pilot program, which was conducted by the Massachusetts eHealth Collaborative (MaeHC).
The Annals study compared these practices to "matched control" practices in six other communities in which providers had applied for the grants but had been turned down. In the latter towns, EHR adoption was not widespread at the outset of the study, which analyzed claims data from 2005 to 2009. Julia Adler-Milstein, one of the study's coauthors, told InformationWeek Healthcare that EHR penetration in those areas was less than 10% in 2005 and probably didn't grow much by 2009.
[ What factors are still impeding EHR interoperability? Read Sharing Electronic Medical Records Still Too Hard. ]
The study had a lot of statistical power. It included nearly 48,000 intervention patients and 130,600 control patients who received most of their care from about 800 intervention providers or nearly 1600 matched control providers. The researchers used the claims data to compare changes in the cost of care in the intervention and control communities during the study period.
The researchers found that, relative to average per member per month cost of $151.13 across the three intervention communities in the period before EHR implementation, the practices saved an average of $5.14 per member per month (PMPM) in projected costs over the 18 months after implementation -- a drop of 3.4%. Most of the savings came in ambulatory costs, which were $4.69 PMPM, or 3.1%, less than expected.
Lower radiology costs represented about a third of the ambulatory savings. "[This] was the only outcome with an absolute decrease in monthly cost growth in the pre- to post-implementation periods for intervention patients, whereas the cost trajectory increased among control patients," the study noted. Easier access to test results may have been prompted fewer ordering of both lab and imaging studies, the researchers suggested. This finding stands in contrast to that of another recent study showing that EHR usage led to an increase in imaging test orders.
The study did not find any significant reduction in inpatient costs for patients in the intervention groups. While noting that the study was focusing on ambulatory costs, Adler-Milstein noted that physicians must be able to use their EHRs to manage population health pretty well before it has an impact on hospital utilization.
Most of the providers in the intervention communities, the researchers pointed out, were using their EHRs for the functions they must perform in order to attest to Meaningful Use stage 1. These included use of electronic problem, medication and allergy lists, electronic transmission of prescriptions to pharmacies, computerized physician order entry, and lab and radiology test results. Consequently, the study authors said, "The MAeHC experiment offers timely data on how the federal meaningful use program may impact costs in the short-term."
At the rate of savings observed in the study, the researchers said, it would take seven years to recoup the five-year estimated cost of implementing the EHR. However, this was one of the rare cases where a payer -- Blue Cross Blue Shield of Massachusetts -- absorbed those costs. Even the Meaningful Use program, which is funded by the Centers for Medicare and Medicaid Services (CMS), provides only a fraction of the five-year costs of the technology (calculated in the study as $130,822 per provider in the pilot communities). The rest must be borne by providers.
Adler-Milstein stressed that the cost savings measured in this study represent the economic benefit to society, not to healthcare providers. An earlier study of the same communities by her team showed that the average provider would lose nearly $44,000 per provider by investing in an EHR. Only 27% of the practices achieved a positive return on investment.
Among the new study's conclusions are that "the lack of a statistically significant reduction in total cost may be explained by providers not using EHRs in more advanced ways that would improve patient health status." Adler-Milstein doesn't attribute that to a lack of training or implementation support; in fact, she doubts that many of the providers in the Massachusetts pilot are using their EHRs in more sophisticated ways today than they did in 2009.
For that to happen, she suggests, the healthcare system will have to be transformed to help providers "see why they should go through what is ultimately a difficult, complex process of changing the way they deliver care." At the same time, she added, the bar for Meaningful Use will have to be raised to induce physicians to use EHRs more effectively.