EHR: Great Expectations?

It’s so interesting how the same statistics can be interpreted through different prisms to arrive at opposite conclusions. In a past post, I wrote about the recent comprehensive MGMA study that explored the EHR experiences reported by over 4,500 professionals, representing 120,000 providers, focusing on the effect of EHR implementation on providers’ operating costs and productivity.

In this month’s issue of MGMA Connexion, David Gans, MGMA vice president, discusses the results of this study. He points out that EHR benefits to providers increase as they increase training and optimize their EHRs—“optimize” being defined as allocating sufficient time for physicians and staff to become familiar with the system—not a surprising finding. The tone of the article leads the reader to view as good news the fact that 37.7% of the respondents who claimed that they had optimized their EHRs reported productivity gains.

This statistic lends itself to a glass-half-full versus glass-half-empty analysis. Shouldn’t we look at this result with considerable concern since it means that 63.3% of optimized EHR users are not experiencing improved productivity? As the shaded area of this chart illustrates, if we did nothing to change these results, the vast majority of EHR users would never realize productivity benefits.

This is not a condemnation of EHRs, but rather a caution that physicians must evaluate their EHR options carefully to ensure that they purchase a system that is designed for their particular specialty and workflow—one that they can implement easily and optimize quickly, and that delivers increased productivity from the outset.

Meaningful Use Retrofit: Unintended Consequences

Question: What do you get when you try to retrofit a primary-care-focused government program to be relevant to specialists?
Answer: Unintended consequences.

In response to the outcry from specialists about the primary-care nature of meaningful use, CMS implemented some exclusions to make the program more workable for them—but these changes did little to make it more “meaningful.” In fact, there were some significant—and, I have to believe, unintended—consequences that will actually create challenges for many specialists:

  • Some physicians will have to add what they consider non-relevant vital signs to their workflow. Acknowledging that vital signs are not within the scope of practice for some specialties, CMS provided a potential exclusion for this measure. However, to take advantage of the exclusion, providers must attest that all three vital signs (height, weight, and blood pressure) do not apply to their practice. This leaves physicians who routinely record just one or two of the vital signs without a way to satisfy the measure, other than to add the non-relevant measurement, too—blood pressure for orthopaedists, height and weight for ENT specialists, etc. (For an explanation, read The All 3 Vital Signs Dilemma.)

Is it the government’s intention that orthopaedists take blood pressure on every patient? Perhaps—but if so, are these physicians then responsible for treating this condition? Some would argue that orthopaedists need to know their patients’ blood pressure if they are going to perform surgery—likely true, but why should this requirement also apply to non-surgical patients? It should not, but let’s remember: this is a primary-care-focused program, designed by primary-care physicians, and this type of analysis was never performed.

  • Some specialists will have to report on clinical quality measures (CQMs) that have no relevance to their practices—and in doing so will provide meaningless data. The accommodation for specialists in regard to reporting on CQMs was originally intended to allow them to identify non-relevant measures by reporting zero as the denominator. However, the meaningful use measure requires physicians to report CQMs using data that is directly generated by the EHR. This results in the inclusion of patients in the denominator of CQMs that are outside the scope of the reporting physicians’ specialties. Physicians will be forced to report on problems for which they did not treat the patients, reporting zeros as the numerators, but not the denominators. (For further details, see Clinical Quality Measures: Who Defines Relevance?)

This raises several issues. First, physicians are reluctant to go on record as not addressing a known problem, even when that problem is outside their specialty. In this litigious medical environment, this is not an unreasonable concern. Second, the data generated in this manner will be meaningless since numerous specialists will provide data on the same patients and on the same problems as are being treated by primary-care physicians—making it will appear that, on average, very little care is being provided for these problems. Some physicians will be trapped into providing data on measures that are not relevant to their practice, even though valid measures exist. For example, any pediatrician with even a few patients over 18 will have to report on those few older patients—using the core measures meant for adult populations—rather than report on his/her entire pediatric patient base, using the alternate core measures that were specifically added to capture valuable information regarding the pediatric population.

CMS acknowledges that the purpose of this particular requirement during Stage 1 is merely to get people comfortable with automated reporting—specifically stating that the government is not looking to gain valuable information about medical practices or the state of care being delivered. I am concerned, however, that the discussions to date about reporting on CQMs in Stage 2 do not advance us towards producing data that will be any more meaningful. To accomplish that, the entire structure would have to look more like PQRS, where physicians report only on measures directly relevant to their practices.

For meaningful use to be meaningful to specialists—beyond the financial value of the incentive money—the program needs to be restructured to fit their needs and practices. It is quite evident that trying to apply primary-care requirements to specialists is not effective.

EHRs and Productivity Loss: How Can This Be Acceptable?

The purpose of automation is to increase efficiency and productivity. Every industry that has undergone the transformation from paper to digital has realized these benefits immediately . . . every industry, that is, except the EHR industry. Why is this acceptable?

Even the AMA acknowledges this failure—and yet seems to accept it. Toward the end of its newly released, and otherwise very helpful, video on how to select an EHR is the test question: “What is the ‘best practice’ in terms of the number of patient visits to schedule during the first week of operation with your new EHR?”

Why does the AMA think that the correct answer “A”?:  “Reduce the number of patient visits by up to 50% for the first week to allow you and your staff to learn how to use your new EHR.”

Why isn’t it “D”?:  “Your new EHR was carefully selected to fit into your practice smoothly and seamlessly. There should be no impact on patient volume that first week.”

Why does the typical EHR implementation follow the bottom line of the graph below, when it should look like the top one? Dr. Jacqueline Fincher’s testimony at last week’s HIT Policy Committee’s hearing on “Experience from the Field” is representative of the all-too-common experience.

Dr. Fincher reported an “absolute requirement to drop patient volume by half for the first three months [due to] an exponential learning curve,” and that she and her partners “have never gone back to the previous volume of patients,” even after 5 years of EHR use.

Some argue that the medical business is different from other industries like banking and shipping. That is very true. The type of data collected is different, and the level of employee responsible for inputting much of the data is also very different. In most industries, it is the lower-level, less costly employees (such as bank tellers and UPS truck drivers) who input data, while in medical practices, it’s actually the CEOs (i.e., the physicians) who do it. This makes productivity all the more critical for an EHR. According to the recent MGMA study on EHR adoption, fear of productivity loss is the primary barrier to EHR adoption—a concern justified by reports from experienced users, as illustrated below.

For the EHR industry to evolve as necessary for widespread adoption to become a reality, choice “A” must be rejected as totally unacceptable by physicians and the professional organizations that represent their interests. Physicians should expect more from their EHRs—they should demand that vendors deliver productivity, not merely fancy features and functionality. The truth is, they can get both, but only if they do their due diligence.

Physicians: Don’t Count on EHR Support from Hospitals

Anthony Guerra, noted HIT industry blogger and editor of HealthSystemCIO, has written extensively about the pressures and stresses facing hospital Chief Information Officers (CIOs) due to the myriad government programs making demands on their skills and their time. His recent survey revealed that a mere 16% of CIOs manage to maintain a relatively normal workweek of 40–49 hours, while 35% report working over 60 hours per week. In my opinion, the current level of stress extends throughout all levels of the IT staff—a sentiment echoed at today’s HIT Policy Committee meeting as they evaluated the recommendations for Stage 2 meaningful use.

This is not surprising. In the midst of upgrading to meet meaningful use requirements—a bigger challenge for many hospitals than originally anticipated—IT departments are expected to simultaneously prepare their facilities to comply with the impending 5010 requirements and convert their systems from ICD-9 to ICD-10. Also looming in the not-too-distant future are the newly defined Accountable Care Organizations (ACOs), which will require significant and expanded data and reporting capabilities. All of this is compounded by a shortage of IT professionals in the healthcare arena.

So how does this relate to private-practice physicians—the constituency on whom I typically focus? I’ve cautioned in a previous post, One Size Does Not Fit All, about the mismatch between the EHR needs of hospitals and of physicians in private practice. Physicians should also be wary of adopting their hospital’s EHR if they are doing so with the expectation that the hospital’s IT resources will be at their disposal. They will be sadly disappointed—supporting private practice physicians, particularly specialists, will be low on the list of priorities for IT staff when their plates are already overflowing.

Usability: Can Every EHR Be Above Average?

Usability is the key differentiator between the long-term success and failure of an EHR implementation. The findings of the recent MGMA study lead to the inescapable and troubling conclusion that too many physicians do not consider their EHRs “usable.” A bad EHR choice is costly for the particular physician(s) and, while it might suffice in the short term for the purpose of earning meaningful use incentives, it will do nothing in the long run to foster sustained EHR adoption. Recognizing this, the HIT Policy Committee’s Adoption and Certification Workgroup convened an 8-hour hearing last week on the subject of how to define and measure usability. Recommendations were offered that mirrored my EHR reform proposal, and various groups/studies are already working on usability testing. One such group is CCHIT, which has introduced a usability rating tool into its commercial certification (not to be confused with government-certification) process. In her testimony, Karen Bell, M.D., Chair of CCHIT, discussed the results (Chart 1) and her recommendations.

So what’s wrong with this picture?

What’s wrong is that, to be useful to physicians, it has to look like this:

This is not meant as an indictment of CCHIT—the organization is to be commended for having taken an important first step in defining usability and creating a process for measuring it. The problem—which Karen Bell did acknowledge when challenged about it—is that if this rating scale were an accurate reflection of usability, there would be many fewer complaints about EHRs and, in my opinion, many fewer failures.

To provide physicians with the objective information that will be valuable to them in EHR purchase decisions, the ratings must be comparative and follow a normal distribution, as illustrated in Chart 2 above. Because achieving this distribution would require more aggressive usability criteria, it would distinguish those EHRs that have the greatest positive impact on productivity and cost savings from those that have a lesser, or negative, impact in these areas.

Even more important, this more challenging evaluation will create a market in which vendors are forced to compete on usability and how to better meet the needs of physicians. Physician satisfaction levels will increase. It will elevate quality across the board and raise the level of the entire EHR industry. Perhaps, as Dr. Ross Koppel testified at the Usability Hearing, if health IT were more usable, we wouldn’t even need incentives to spur EHR adoption!

Stage 2 Meaningful Use: What Do You Think?

At this week’s HIT Policy Committee meeting, the Meaningful Use Workgroup presented its Stage 2 thinking to date, based on the 422 comments they received on their initial proposal. As discussed in a previous EMR Straight Talk post, the issue at the forefront is timing—with providers and vendors expressing significant practical concerns, and consumer groups pushing for rapid advancement.

The workgroup presented the following options for consideration by the Policy Committee. (I invite you to voice your opinion by responding to the poll below.)

  1. Maintain current timeline. Stage 2 would begin in 2013 for providers who demonstrate meaningful use in 2011. Providers who first demonstrate meaningful use in 2012 would have until 2014 to meet the Stage 2 requirements.
  2. Maintain the current timeline (as above), but allow a 90-day reporting period, instead of a full year, when providers are first governed by Stage 2 requirements. This would give providers until October 1 to begin their first year at Stage 2, instead of January 1—a nine-month delay.
  3. Delay Stage 2 by one year, allowing providers 3 years instead of 2 years at Stage 1. This means that the earliest any provider would have to meet Stage 2 expectations would be 2014.
  4. Phased-in approach separating existing from new functionalities:
    - Stage 2a (2013) would increase thresholds for measures for which the functionality already exists, (required to meet Stage 1), adding only new clinical quality measures.
    - Stage 2b (2014) would add new measures that require new EHR functionalities .

The responses from various HIT Policy Committee members covered the gamut.

  • Some were in favor of moving aggressively at all costs, presenting various arguments such as: (a) If we don’t pressure providers now, we will face the exact same issues at the next stage; (b) More extensive data capture does nothing to move us towards Stage 3 goals; and (c) We cannot just address the physicians’ workflow problems and ignore the challenges patients face in dealing with the current, difficult-to-navigate healthcare system.
  • Other Committee members, like Gayle Harrell, cautioned against trying to do too much too quickly—as she has from the outset—and stressed the long-term value to the program of setting providers up for success. Pushing them too hard could cause them to drop out after they earn the bulk of the incentives associated with Stage 1.
  • The phased-in approach was perceived as creative, but I was surprised that there was not much discussion about the administrative complexities this plan creates—to say nothing of the challenge of conveying it to providers.

MGMA Study Reveals #1 Reason Physicians Fear EHRs

The evidence is indisputable: the fear of lost productivity associated with EHR implementation is uppermost in the minds of physicians, and their fears are justified by the actual experience of the majority of EHR adopters to date. The titles of two articles about the recently released MGMA EHR survey say it all: “Survey: EHRs Often Don’t Increase Doc Productivity” (Health Data Management) and “HITECH Drives Docs to EHRs, but Cost, Productivity Issues Remain” (Healthcare IT News).

MGMA is to be commended for the size and scope of this important survey (4,588 practices representing 120,000 physicians), for the multiple ways it segmented the survey population, and for the detailed analysis of the results. One important segmentation was missing, however—that of physician specialty, or, at a minimum, of primary care versus specialist. The EHR experience of orthopaedists or ophthalmologists, who may see as many as 60 patients a day, is dramatically different from that of a family practice physician who sees 20.

Productivity was the pervasive issue. The only group that reported some productivity gains was the 16.3% self-proclaimed “optimized users” of EHRs—those who have had sufficient time following implementation to master the EHR. (The report did not define “sufficient time.”) Among this group, 41% reported that physician productivity has increased. What is disturbing about this statistic, however, is the implication of the converse—that even among these most accomplished EHR users, the majority of physicians (59%) are seeing a decrease, or at best no increase, in productivity. For the total population studied, 43% have just worked their way back up to where they were before implementation, and 31% of respondents are experiencing an actual productivity decrease.

Productivity was the major factor accounting for why 8% of survey participants are in the process of replacing their EHR with another, while anticipated productivity loss was reported as the most significant barrier to EHR implementation for physicians still using paper charts. Among these paper users, 78% fear productivity loss during implementation and 67% worry about the effect even after the transition to an EHR.

This data confirms past experience regarding productivity loss and raises these critical questions:

  • Why do only 16.3% of EHR owners categorize themselves as “optimizing their use of an EHR”?
  • While government incentives will certainly address the financing concerns expressed by small practices, how will this money address the productivity obstacle for all adopters?
  • What accounts for the loss of productivity?
  • When technology has replaced an antiquated paper process in other industries, it has always brought increases in productivity. How do we deliver the same results in healthcare?

The MGMA report did not tie satisfaction and productivity to the particular EHR being used, but there were clearly some successes, so it is important to understand what differentiates these implementations. It all comes down to usability. According to a recent HIMSS Task Force Report on why adoption has been so slow, “A key reason, aside from initial costs and lost productivity during EMR implementation, is lack of efficiency and usability of EMRs currently available.” I maintain that lost productivity and lack of usability are one and the same.