Obviously, I am an advocate of EHR use.  Actually, tracking EHR implementation outcomes is somewhat of an obsession.   For years, I have kept a folder on my desktop called “failed projects” that contains the woes, assessments, and post mortems of IT projects gone bad.  Since HITECH, I’ve noticed something interesting.  Whereas pre-HITECH it was fairly common to see articles in the trade media about EHR project failures, since 2009 the word “failure” is rarely seen.   I get this. No one wants to be a gloomy Gus–me included.   However, ignoring a problem doesn’t make it go away.  And being someone who really believes that HIT can improve care, increase productivity, and lower costs, I feel the best approach is to figure out how to make every project a success.

One of my earliest posts, What Constitutes a Successful EHR Implementation?,  related my concerns about equating the presence of an EHR in a practice with successful implementation or adoption.   The danger in this way of looking at HIT utilization is that it does not consider how, or to what extent, practices should benefit from implementing  a system.   Nearly 18 months later,  I think the question raised in that post remains pertinent: What is a successful EHR implementation?    A recent Health Affairs article and a survey report from  American EHR Partners (AEHR) supply new information that suggests that equating the presence of an EHR with successful implementation is ill-founded.

Let’s start by looking at the survey results from AEHR.  The survey was carried out from 2010 until 2012 and involved clinicians from 10 specialty societies.  Here are a few of the key findings:

  • The average length of time that survey respondents had been using their EHRs was more than three years at the time of completing the EHR satisfaction survey.
  • Satisfaction and usability ratings are dropping. This holds true regardless of practice size, specialty type and across multiple vendors.
  • Overall EHR user satisfaction reveals a 12% drop in satisfied users from 2010 to 2012 and a corresponding increase in very dissatisfied users of 10% for the same period.
  • In 2012, 39% of clinicians would not recommend their EHR to a colleague.
  • Satisfaction with the ability to improve patient care revealed a drop in very satisfied users from 2010–2012 of 6% and an increase of very dissatisfied users of 10% for 2010–2012.
  • The length of time to overcome initial productivity challenges typical with beginning to use an EHR — 32% had not returned to normal productivity levels in 2012 compared to 20% in 2010.
  • 34% of users in 2012 were very dissatisfied with the ability to decrease workload compared to 19% in 2010.

Clearly, something is wrong.  We have national EHR certification standards, RECs in every state, more EHR-related reports, articles, books, and blog posts than ever before, yet dissatisfaction with EHRs is increasing.

Note in particular, the last two items that verify productivity losses and workload increases.   Clinical practices have the same business imperatives as other businesses.  Most businesses implement IT solutions to lower costs, meet regulatory requirements, and increase revenue.   In health care, quality of care must be rolled into the equation.    For clinical practices, it is reasonable to assume that some combination of these goals would count as a successful EHR implementation.   Of course, the exact combinations  matter.  For example, if a practice noted significant improvements in care quality and met MU/HIPAA requirements, but suffered significant decreases in productivity and higher operating costs, I doubt that practice would consider the implementation a success.  Even so, by current standards, it  would be counted as a successful EHR adoption.

The Health Affairs  article is based on work with practices that participated in the Massachusetts eHealth Collaborative (MAeHC), and it specifically looks at the economic side of EHR usage.  A Survey Analysis Suggests That Electronic Health Records Will Yield Revenue Gains for Some Practices and Losses for Many, by Adler-Millstein, Green and Bates, looked at five-year costs for practices that implemented EHRs between March 2006 and December 2007.  The survey attempted to ascertain before/after costs and benefits for the involved practices.   The authors offered the following as their goals for the study:

We sought to answer three sets of questions: What is the average five-year return on investment from EHR adoption, and what proportion of practices realize a positive return? How do these results vary by practice type and size, and what distinguishes practices that achieve a positive return on investment from  those that have a negative one? And for what percentage of practices that lost money would the meaningful-use incentives have resulted in a positive return on investment, had practices been able to receive the incentives during the study period?

The authors’ major finding was that most practices would not benefit financially from having an EHR. They state:

Average projected five-year return on investment was negative, with the average physician losing $43,743 (Exhibit 1). Only 27 percent of practices achieved a positive five-year return on investment (Exhibit 2).

Reviewing the data that resulted in such a dismal forecast revealed few surprises.  As expected, practices that increased revenues did so by seeing more patients or through improved billing practices, which is exactly what one hopes to achieve with an IT implementation – increased productivity and efficiency.  Even so, 22% of practices reported that additional hours of physician time were required after implementation.

By far, the most troublesome finding reported by the authors related to paper records.  They report that many practices failed to realize any significant cost saving because the practices still had paper records.

…practices with a positive return on investment realized savings by eliminating paper medical records, as well as dictation and billing services and positions of, or hours worked by, staff members who were performing services no longer required after EHR adoption. However, we also observed that many practices failed to make such changes. Almost half of the practices did not realize savings in paper medical records because they continued to keep records on paper.

For me, the take-away from this point from this paper is not that practices did not uniformly fare well after EHR implementation, which is old news (see  MGMA report 2011).   No, the real lesson here is that many practices, even with the assistance available from the MAeHC, one of the best run and best-funded EHR pilot programs, were still unable to optimize the use of their EHRs.  Moreover, it proves once again that the mere presence of an EHR is a poor proxy for successful implementation/adoption.

What now?  The good news is that some practices do benefit from having an EHR.  Using these practices as models to formulate a more appropriate definition of successful implementation seems reasonable. One place to begin, would be setting specific cost, revenue, and productivity objectives based on model-practice traits at the initiation of all REC-supported EHR projects.  These factors could then act as benchmarks for implementation success for a three-year period after go-live.

There are a number of factors that affect a practice’s ability to optimally use an EHR.   Access to funds and expertise are important, but as we have seen, not necessarily sufficient to assure that a practice will benefit from having an EHR. The usability of EHR systems is also part of this equation (39% of AEHR respondents would not recommend their EHR to a colleague).  However, judging by the skyrocketing interest  in EHR design I am seeing,  I am optimistic that better systems are on the way.

HITECH has had the intended effect: EHR use is increasing. Resources should now be focused on assuring that practices predictably benefit from having an EHR rather than on convincing them to buy one.  After all, if the former is successful, the latter will take care of itself.


Leave a Reply

Your email address will not be published. Required fields are marked *