Cracking Down on Overcoding and Upcoding in the ED

In recent weeks, numerous hospitals and physicians in Texas and across the country have found themselves thrust into the media spotlight over suspected overcoding and upcoding in emergency departments. Between 2001 and 2008, hospitals nationwide dramatically increased their Medicare billing for emergency room care, adding an estimated $1 billion or more to the cost of the program.

Among the potential causes for the increased billing is hospitals’ widespread adoption of electronic health records (EHRs). Computerized charting—as opposed to the handwritten and dictation charting methods of the past—captures more of the work that providers perform, so that visits are documented and coded more thoroughly and accurately than ever before.

It has been long assumed that handwritten medical records can often be illegible and incomplete, making it difficult to code. Hence, coders have historically “played it safe” and routinely downcoded paper charts. As well, a material number of paper charts tend to get lost in the shuffle and do not ever get billed. The current push to digitize medical records was partly aimed at improving this situation. As a result, charts are now legible, complete, and don’t get lost.

So, why is the Centers for Medicare and Medicaid Services (CMS) surprised to be paying more in this digitized world? 

While a portion of the cost increase may be explainable with the above argument, there is validity to some of CMS’ concerns. Certain EHR functionality may cause upcoding of visits that could be considered inconsistent with accepted coding practices and potentially fraudulent. Two areas are of particular concern: charting by exception and automatic population of information.

Charting by exception. Many EHRs use charting by exception or macros for documentation. With a single click, a provider can create visit documentation. For example, a pre-programmed macro applied for a visit would mark all physical examination elements as normal. The provider must then review the documentation for accuracy, changing elements that were abnormal for that patient and deleting areas that were not examined. No matter how unintentional, this method of documentation can lead to the inclusion of erroneous information. For example, a computer-generated document might claim that the doctor checked reflexes by tapping on the knee; but that patient is a bilateral amputee, so the doctor certainly didn’t perform that function.

Automatic population of information. Some EHR systems automatically populate parts of the patient record with information documented by other care providers for that visit or from information stored in the system from previous visits. This practice is also known as chart cloning. Some providers argue that they must use macros to reduce the documentation burdens of the system. However, this information automatically populates the record without provider review to ensure accuracy or appropriateness. Indeed, the system may pull social history information into the physician record without reconciliation or review by the physician. As a result, the document will likely meet the minimum number of social history elements for a high-acuity code. There are two problems with this practice. First, it could expose the physician to suspicion of fraud. Second, it adds unreviewed and/or unreconciled information to the medical record, posing a potential risk to patient safety and quality of care.

Fortunately, providers can put sensible safeguards in place. A good first step is employing superior coders and coding services that pay detailed attention to quality assurance, accuracy, and appropriateness of codes. In addition, providers must be vigilant about the methods used by their hospitals to document care. Electronic documentation has countless advantages, but it should not come at the expense of accurate or ethical coding.

EHR vendors also need to accept some level of responsibility when they incorporate tools that induce clinicians to indulge in these types of documentation practices. In fact, it is often a part of the vendors’ sales pitch. A good EHR does not need to include macros or automatic population functionality to be efficient, effective, safe and compliant. These features expose hospitals and physicians to reimbursement audit and medicolegal risk. Information collected in documentation should be specific to a patient’s individual visit. Most importantly, EHRs should require that information on each visit be documented or reconciled by providers. Some vendors do spend the R&D efforts to accomplish this, with due consideration to usability and efficiency.

Overcoding and the associated risks are a real problem for physicians and hospitals, especially in this time when the government is looking for ways to remove cost from Medicare. By taking precautionary steps, healthcare organizations can ensure appropriate coding and prevent unwanted publicity, regulatory scrutiny and financial penalties.

— Sunny Sanyal is CEO of Dallas-based T-System Inc.