Where's the Beef?

There is some good news about EHRs: a new study, by Adler-Milstein and Jha and reported in Health Affairs, found that there is no evidence that hospitals are using their EHRs to increase their Medicare reimbursements.  Such up-coding had been feared as a potential pitfall of the introduction of EHRs.  True, the study showed Medicare billings did increase in the hospitals with EHRs, but at comparable levels to hospitals who did not adopt EHRs.

This is what passes for good news in the EHR world.

That doesn't seem like quite enough, does it?  After all, EHR adoption has skyrocketed since HITECH spurred the launch of the CMS incentive program -- and over 408,000 health care professionals have received some $26b in such incentives over the past three years.  "Not making things worse!" is a pretty low bar for success.

There's more not-so-good news.  A recent JAMA study, by Samal et. alia, looked at the impact of EHR meaningful use on quality of care, and found very mixed results -- marginally better quality results on 2 measures, worse on another 2 measures, and no impact on the remaining 3 measures.  The study raised questions about how "meaningful" meaningful use is, and how we best evaluate its impact and the impact of EHRs generally.

Another new report, by D'Amore and colleagues, in The Journal of Medical Informatics, found major problems with even MU Stage-2 certified EHRs being able to exchange and read so-called C-CDA (consolidated clinical document architecture), as required in Stage 2.   The study participants were pleased to find that they were able to identify common problems and offer specific solutions, but noted that "without timely policy to move these elements forward, semantically robust document exchange will not happen anytime soon."

That doesn't sound good.

On the other hand, in a piece in Healthcare IT News, Jack Beaudoin says that the debate about EHR effectiveness should be over.  He dismisses the Samal study for several technical reasons, and cites preliminary results from a study conducted by HIMSS Analytics and Healthgrades, covering all U.S. hospitals and with a more sophisticated measure of the level of EHR adoption in each hospital.

I've not been able to find any published version of the study, just a tweet from HIMSS, but Beaudoin reports that the study found that EHRs do have measurable, positive impact on the outcomes of care, as measured by mortality.  Both he and the tweet cite a heart attack mortality rate half as high for hospitals with high adoption rates compared to those with low EHR adoption rates.

Although not every patient cohort, service line, or diagnoses/procedures had positive results, "overall, all five service lines studied show statistically significant positive relationships to EMRAM scores for at least one group of diagnoses and procedures."  (EMRAM is their adoption model.)

I would like to believe Mr. Beaudoin is right, but I think the jury remains out.  For example, in a recent report, Rand was rather reserved about the impact of EHRs, citing mixed or modest results on clinical quality.  They specifically pointed out the issues of usability and lack of interoperability as major problems, noting that "HITECH promotes adoption of existing systems, not better ones," which I think hits the nail on the head.

Rand specifically called out market leader Epic for how its closed architecture makes interoperability a challenge.  They're far from alone in this criticism.  KLAS recently focused on Epic's ability (or reluctance) to share data.  I would summarize both reports as finding that Epic is fine with connecting with other Epic installations -- of which there are no shortage -- but, otherwise, good luck connecting with them.

A good illustration of this may be that, as I reported previously, Epic recently announced an interoperability consortium (Carequality), which pointedly does not include major competitors like Cerner, McKesson, or athenahealth, who have their own consortium, CommonWell Health Alliance.    It'd be funny if it wasn't so important.

As FierceEMR asked in their editorial on the D'Amore study, when it comes to interoperability, do we "fix the potholes, or find a new path?"

Congress may be reaching even its limits for incompetence with the Hatfield-McCoy feuds going on between EHR vendors.  Members of both parties are calling for an investigation into what they deem the "information blocking problem."  The Senate Appropriations Committee has asked ONC to report how many organizations are involved in the practice, and warned that "ONC should use its authority to certify only those products that…do not block health information exchange."

Rep. Phil Gingrey (R-GA) went a big step further at a House Energy & Commerce Committee subcommittee hearing, charging that:

We’ve spent tens of billions on non-interoperable products. It may be time for us to look closer at the activities of vendors in the space, given the possibility that fraud is being perpetrated on the American people.”

He also specifically singled out Epic, which, in response, claimed they actually have a "great reputation" for interoperability.  They must talk to different people.

Look, we know that in the future that health records will be digital, just like most other kinds of records already are.  We know that siloed records, even if digital, are not of much value; it is the sharing, the analytics, and the real time response capabilities that will drive the real value.  There's so much potential:

• A study of Medicare ACOs, with particular focus on those offered by academic medical centers, cited health IT as crucial for their success.  The authors note: "There has to be an electronic medical record system robust enough to analyze and assess quality and safety issues.  It’s important to quickly identify areas where changes are needed."  
• Another study predicts that analysis of EHR records is faster and more cost-effective than the current system of clinical trials.  The lead author concluded: "The use of electronic health records in simplifying clinical trials means that we no longer need to remain uncertain about which medicine offers the best health benefits for patients."
• Finally, Health Affairs focused a recent issue on how "big data" can be used to improve our health care system. For example, one study outlined six use cases for how big data can use EHR and other data to identify and manage high risk/high cost patients.

But, like FierceEMR, it is not clear to me if going a new path might not be better than trying to fix all the potholes in the existing path.

As the importance of "patient-generated data" gains in importance, with lots more options for such data, I think we're going to find that the traditional provider-oriented EHR is going to become even more problematic.  As John Halama, CIO of Beth Israel Deaconness Medical Center told MobiHealthNews, "what I, as a CIO, need to do is gather the data on your medical record, inpatient, outpatient, devices in your home, and understand what I can do to keep you well.”  Not everything important about a patient's health happens in a provider's office or facility.

As I've discussed before, I think the problem is that we approach electronic health records with the perspective that they are a provider's record about patients -- rather than patient records that various providers have specified levels of access to and input into.  That is a huge difference.

Yes, I know -- Goggle Health failed and Microsoft Health Vault is, well, doing whatever it is doing, but that doesn't mean the approach is inherently flawed (as Robert Szczerba recently wrote about in Forbes).

New studies indicate that, although they expect EHRs to be used, as many as 53% of consumers worry about the safety of their data in EHRs (Morning Consult), and as many as 13% have withheld personal data from providers due to concerns about privacy/security (Campos-Costillo/Anthony).  I wonder if people would trust them more if the records weren't "owned" and controlled by their providers instead of themselves.

I've taken my shots at EHRs before (They Shoot EHRs, Don't They?), but it's not because I don't support the concept -- it's because I am so disappointed we're not achieving more with them.