Does EHR Use Lead to Lower or Higher Costs? Part 6




By Michael Zaroukian, MD, PhD, FACP, FHIMSS and Peter Basch, MD, FACP


Incorrect Assumptions Lead to Incorrect Conclusions That Misinform Policy

Please refer to Part 1 of this post for our comments on the recent Wall Street Journal op-ed questioning the policy behind the EHR incentive program; Part 2 for our comments on the extent to which EHR use was associated with higher short term charges, as published in the Center for Public Integrity report and recently summarized in the New York Times; Part 3 for our comments on where one would expect EHR use to show short-term cost increases;  Part 4 for our comments on repetitive or boilerplate verbiage; and Part 5 for our comments on billing coding software.

EHRs did NOT cause note-bloat (clutter and verbosity).  When we started training, only medical notes of 3rd year medical students consisted of multiple pages of statements of negation, such as “patient denies shortness of breath at rest, patient denies shortness of breath with exertion, patient denies shortness of breath when lying down, etc.”  Even then, this repetitive and burdensome training requirement felt more like a grade school punishment (stay after class and write on the blackboard 50 times, I WILL NOT CHEW GUM IN CLASS); it served its purpose of making new students remember the exact verbiage and when to ask each question on the next ‘Review of Systems’ item on the list.   Thankfully, the nightmare of having to produce or read such verbose monstrosities stopped with a student’s advancement to the 4th year of medical school.  Medical training thereafter included coaching on how to concisely document only what was important (e.g., the most pertinent positives and negatives to establishing a diagnosis or planning treatment), with the understanding that based on your training and expertise, you would best use your time in making better decisions rather than wasting it in superfluous documentation; likewise, you would show respect for or other providers’ time by producing for them concise and informative notes.  In fact, you could pretty much tell someone’s level of training and clinical maturity by their consistent ability to achieve the triple crown of completeness, clarity and conciseness in their clinical notes.

All of that changed abruptly and durably in 1995; this was not due to the introduction of EHRs but rather was a side effect of an effort to solve what was seen as a provider billing “fairness” problem.  Prior to 1995, providers were largely free to self-declare whether a visit was “limited, intermediate, or comprehensive” – and document as appropriate.  The provider’s bill was an attestation of the level of service, and the documentation of that service (still primarily using standardized boilerplate verbiage) was left to the provider. This new approach, known as the Evaluation and Management Documentation Rules (E/M Guidelines) clarified that the level of service and thus the basis of the bill was NOT what was done, but rather was what was documented if the element of care was also medically necessary.  The guideline explained the documentation requirements in a series of complicated tables specifying how many of which elements of history, examination and medical decision-making were required to support each level of service and how they differed for new patients vs. established patients, as well as for patients receiving primary care vs. specialty care.

While the specifications were described in detail, they were not clear and unambiguous to many providers and could not be easily done in one’s head, or even with coding sheets.    It can be confusing and difficult for providers to correctly code a visit by counting the elements in a paper note or from the printed output of an EHR note; different providers might code the same note differently.   The proliferation of entire new industries of coding software, coding consultants, and coding auditors that followed the introduction of the E&M coding guidelines is further evidence of this complexity and confusion. 

In addition to creating ongoing provider confusion as to what exactly is needed to bill for a particular service, the 1995 and 1997 E/M Guidelines catalyzed the very conditions that the Center for Public Integrity is now concerned about.  The coding rule complexity, along with major penalties for over-coding increased the probability in the paper chart world that providers would deliberately under-code their visits to minimize the financial and reputational risks of being caught in the ‘speed trap’ of inadvertent over-coding.  With the complex math required to tally and correctly code visits using what was then the new E/M guidelines, it is no wonder that EHR vendors began to use the arithmetic capacity of their EHR systems to develop structured data entry strategies that could capture and count the elements of care that could inform more accurate E/M coding.  Eliminating under-coding and facilitating appropriate coding became an important aspect of the business case for physicians purchasing an EHR system. 

Along with the confusion and wasted time and effort mentioned above, the E/M guidelines struck a death blow to brief and cogent notes.  It actually takes much more time to think through everything that should be considered and then create a concise, 250-word visit note from what would otherwise be an undisciplined 1000 word stream of consciousness that regurgitates superfluous or irrelevant data along with the key findings of a clinical encounter.  However, the E/M guidelines now represented an actual financial penalty for the brevity that contributed to conciseness and clarity.  Where what you document according to arcane rules now becomes more important than what you actually do during a visit; it is no surprise that not only do we have providers who complain that they feel like they spend more time and energy creating  documentation than caring for the patient, we also have a payment environment that encourages ‘verbosity PLUS’ – with the downside of insufficiently documenting a service being not just that a bill will be rejected but the added specter of accusations of “billing fraud.” 

It gets worse.  The E/M documentation guidelines do indeed predispose some doctors to over-document (which depending on the patient and provider, could include findings that may not have actually been procured), driven by the fear of failing a coding audit.  The guidelines are clear that since documentation is the ultimate defender of one’s billing practices, one should make sure that the documentation supports the billing level.  By the way, CMS officials also made it clear that they consider under-coding to be billing fraud, even if it is never prosecuted.  

Yes, it is true that EHRs make it easier to populate notes with templated text or previous findings, or even many such findings, with a single click.  However, the reason why physicians feel anything but embarrassment about having their electronic signatures affixed to bloated and inelegant notes is not the fault of the EHR, and certainly not the fault of the EHR incentive program, it is because of payment policy.  In this case, the payment policy is not just rewarding volume; it is rewarding verbosity using standard terms that expands the volume documented (even if one stays within the bounds of medical necessity).

To the point of EHRs enabling over-coding and increasing healthcare costs – when the provider payment system requires that a certain number of fields be completed to justify payment; there is valid cause for fear in the provider community that they are all likely to be suspected of “billing fraud.”  There is enough ‘wiggle room’ in the documentation rules that providers may never really know if a given note meets a particular coding level. As a result, there is a constant pressure to more extensively document one’s findings (over-document) and err on the side of under-coding. 

Now take this monster spawned from the perverse payment system – the bloated and inelegant medical note – and run it through an EHR.  As most EHRs now have coding software that analyzes structured notes and indicates the billing level that fits the documentation, providers now have a real-time, high-reliability coding expert (coding software) that can correctly suggest the correct code supported by their current documentation, for example, a level 4 visit rather than the level 3 that their intuition and greater fear of overestimating the correct code would have caused them to code previously.  The provider might have been over-documenting and/or under-billing for years but  now, armed with this real-time coding advice, they can confidently and systematically “right-code” each visit.  

Our conclusions:

(1)    True billing fraud is a crime that hurts us all.  We support the call by Attorney General Holder and Secretary Sebelius for vigorous prosecution of fraudulent billing.  While our defective payment system might seemingly allow for verbose notes to facilitate higher billing, billing guidelines also reference appropriateness and medical necessity.  We thus remind our colleagues that over-documentation of what was not performed or what was clearly not necessary is unethical and unprofessional, particularly when it is done with the intent to inflate a bill.

(2)    We believe the Wall Street Journal op-ed misrepresented the conclusions of the recent report from McMaster University investigators on the economic analysis of EHRs used for medication management.  We agree that too few reliable studies have been conducted; and even fewer that present an end-to-end analysis of modern EHR systems with embedded value-based decision support, used in care settings where the business driver was not solely volume and complexity of services.  We do not believe that the McMaster study says anything negative about existing policy regarding incentives for EHRs and the Meaningful Use program.

(3)    The report by the Center for Public Integrity presents troubling findings about a very small number of physicians with aberrant billing profiles; as well as presenting inferences and comments about EHRs, documentation, and fraud; and presents conclusions based on conflated logic; and not the conclusions of the OIG.  We strongly support the call from Attorney General Holder and Secretary Sebelius to seek out and prosecute intentional and criminal billing fraud.  That said, even the OIG report did not suggest that the small number of physicians found with aberrant billing profiles committed billing fraud.

(4)    Even if a report suggested a clear association with EHR use and short term cost increases (and these recent reports did NOT); that is not necessarily an indication of a problem.  In fact, the underpinning of the value of decision support for many fields, primary care in particular, is to redress the fact that necessary and appropriate preventive and chronic care services are severely underperformed in the US.  Any such study would need to look at where these excess costs occurred, and if the excess costs were actually a reflection of EHRs doing their job in improving processes of care, thus leading to long-term cost savings from disease prevention, early detection and better chronic disease management.

(5)    Repetitive and/or boilerplate documentation across patients and for the same patient over time is not prima facie evidence of inappropriate documentation; it is prima facie evidence that the person authoring the note graduated medical school. Repetitive and/or boilerplate documentation was not created by EHRs; it is present in handwritten and dictated notes, and is typically the source of drop-down list choices for EHRs, based on specialty and scope of practice.  Physicians should not be threatened for consistent use of terminology just because common findings are indeed commonly found and documented in a consistent manner.

(6)    Coding software is the electronic expression of open and transparent coding rules. These coding rules are complex, and because physicians are paid for documentation of the complexity of medically necessary services and not for an attestation of such services, physicians may be down-coded, not paid, and subject to billing fraud audits and penalties, for forgetting to document even a single element of a multi-page document.  We acknowledge and are troubled that there are individuals who misuse a tool that helps others comply with complex regulations in a professionally appropriate manner.  However, we do not believe it appropriate to penalize those who would use the software wisely and well to prevent those who would misuse it from doing so.  Those who seek to deceive and defraud will simply find another way.

(7)    The existing Evaluation and Management (E/M) billing guidelines of 1995 and 1997 have not served their intended purpose, which was to make medical evaluation and management service billing transparent and fairer.  Instead, the guidelines have created confusion that continues to this day, and makes many doctors feel like they spend more time documenting or worrying about what they document.  Note-bloat is wasteful of the documenter’s time and it often distracts providers from using the EHR for what it was intended – to make care better, safer, and more affordable.  It also creates verbose monstrosities for the next provider of care, where the ideal of a concise clear note appears instead as a text-based “Where’s Waldo” puzzle in which important information needed to ensure quality care is hidden from view.  It has also necessitated an entire industry – coders, coding instructors, coding auditors, and coding software.  We contend that a different payment system that does not rely on rules that encourages “note-bloat” and with specifications that only professional coders can understand, would free up time and resources that could be better devoted to direct patient care. Evolving from our current payment system to one that pays for appropriate services and outcomes and NOT bloated documentation will help to reduce the burden on the producer and reader of medical documentation, and should reduce cost in the system by allowing practices and hospitals to spend more of their time and efforts on care, rather than documentation and coding.  It should also reduce concerns of EHR use as a cause of billing fraud.

(8)    We do not need any more research that concludes the obvious, that ‘EHRs in and of themselves do not make care better, safer or less costly.’  We and most others already accept this, and current EHR incentive program policies reflect a similar conclusion through their incenting of meaningful use rather than EHR implementation and adoption alone.   Studies of health IT effectiveness should look at system goals and the inherent financial incentives for performing or avoiding the performance of specific services. In that context, we should then examine whether health IT has an effect on adherence to goals, costs to achieve these goals, and other objectives.  It is foolish to study health IT infrastructure divorced from how it is deployed and used.

(9)    The Meaningful Use incentive program is far from perfect; we are regular commentators for proposed rule enhancements and have no shortage of opinions as to what could be improved.    However, on balance we believe that the Meaningful Use program is a reasoned one that has succeeded in accelerating movement of the health IT industry in a positive direction, and moving providers toward use of steadily improving health IT that will make care better, safer, and less costly. 

Tell us what you think – please add your comments and suggestions to this discussion.

About Michael H. Zaroukian, MD

Chief Medical Information Officer, Professor of Medicine, Michigan State University
Medical Director, Clinical Informatics and Care Transformation, Sparrow Health System

Lansing, MI

Speak Your Mind