Unintentional Fraud: The Slippery Slope of EMR Up-Coding

Electronic medical records were bad enough when they slowed you down. Now they might also be facilitating billing fraud.

In talking with physicians from all over the country, I have found that most say that charting via an electronic medical record system is the invention of the devil. Only rarely will an emergency physician say that they go faster with an EMR than with what they were charting with previously. And whenever I hear a physician assert this, I generally wonder how efficient they were in the first place. Some (many) EDs have embarrassingly slow turnaround times – discharged patients may average four or more hours in the ED. In such dysfunctional EDs perhaps an EMR can be shown to improve turn-around times from miserable to just plain bad.

But in the EDs that are moving (2-3 patients per hour per provider with an average admit rate (about 20%), EMRs have dramatically slowed the process in most cases. The only exceptions I hear about are those that allow free text dictation in addition to checking the boxes – but I really don’t count these as true EMRs.

ADVERTISEMENT

Although charting of histories, physicals, interval progress notes and medical decision making via an EMR can be miserably tedious and slow, I’ve found that many physicians like computerized provider order entry (CPOE). But I think they like it for the wrong reasons – order sets. CPOE gives the doc the ability to push one order and blow out about $2,000 worth of tests for chest pain or belly pain. Oh sure, the creators of the order sets (often the physicians at the particular ED) say that providers can remove orders from the sets that they don’t want. But realistically, what is the likelihood that providers will do so? Nil.

And by the way, I know that physicians resent being called “providers,” but give me a different word when there are also PAs and NPs seeing patients in the ED. My belief is that in “good” EDs (where physicians take seriously their responsibility to order appropriately) a “before” and “after” study will demonstrate that after the introduction of CPOE the average patient gets more tests and higher bills than previously. I would love to see this study – shouldn’t be hard to do at all. But it can’t be done at a “teaching hospital” where the emphasis is teaching residents how to order every test in the book (a practice that I believe is rampant).

I think charting is largely a huge waste of time and is extraordinarily costly. The author of a story in the “WhiteCoat’s Call Room” blog chronicled how much time was spent on various tasks in a 12 hour shift. It was a single covered ED in which the author said they saw 1.9-2.5 patients per hour when all the rooms were full (pretty modest). In any case, here are the woeful details.

ADVERTISEMENT

Out of a total of 720 minutes in the shift:

• Time spent being with patients=247 minutes
• Time on the computer:

• Charting and CPOE=219 minutes
• Looking up medical records=42 minutes
• Entering discharge instruction/prescriptions = 41 minutes
• Entering admitting orders/transfer forms= 63 minutes (yes, writing admitting orders is often perfectly reasonable)

• Discussions with other physicians=69 minutes
• Researching medical issues=13 minutes
• Eating lunch=5 minutes
• Personal phone calls=4 minutes
• Miscellaneous down time=12 minutes
• Signing out to oncoming physician=5 minutes

The bottom line – during each hour 30 minutes was spent on a computer and about 20 minutes was spent with patients. The author, as an aside, notes that some nurses have said he spends too much time with patients!

Even if the computer time was only 20 minutes per hour it is still extraordinarily costly if seeing one new patient is assumed to generate the conservative figures of $100 for the physician and $500 for the hospital. Unfortunately the physician did not indicate the number of patients seen during the shift, but if we assume for the sake of discussion it was a measly two patients per hour, the ability to see just a half a patient more per hour would generate a net of $300 an hour ($7,200 per day $2,628,000 a year). The point: slow, computer-generated charting is extremely costly.

ADVERTISEMENT

Why do we chart?
First, to tell other providers what we thought was wrong and what we did (our colleagues just want to know what our diagnosis was, not the 12 item ROS) and the chart can tell what tests and treatments were done and prescribed (they want to know that as well).

Second, the chart justifies the bill (more on that later).

Third, we chart to defend ourselves if malpractice is claimed. But given the infrequency of malpractice suits (there is only about one suit for every 30,000 ED visits) most of this time is unnecessary.

And, finally, we chart because a small subset (generally very, very small) are checked in quality assurance reviews (an endeavor that is highly subjective and full of holes). In short, the only person who is likely to read your chart is the coders who are generating your and the hospital’s bill. Most charts are never read by anyone else.

ADVERTISEMENT

Like most things in medicine, EMRs carry with them unintended consequences. Most EMRs placed into formerly efficient EDs slow the providers down significantly. If the providers are hospital employees, then this is the hospital’s problem (the staff are paid by the hour and not by the patient). If the providers are independent contractors it becomes the providers problem – less patients per hour directly hits the bottom line of the providers. In this scenario the hospital is also hurt due to less patients being seen, but they are often under the delusion that EMRs and CPOE are good for patient care or safety. Where they got this belief is unclear – it is virtually impossible to show they improve the quality of care in this setting. Therefore what we have is slower care and no better care (or care that is hard to prove to be better).

But not so fast. There have to be some benefits with EMRs and CPOE, right? EMR purveyors claim that lost charges will be less with their systems, which is likely true if you take the position that every item used on a patient warrants a unique charge. However, this practice of “microbilling” has led to exorbitant up-charges (IV fluids cost about $1 a bag and are often billed at well over $100 a bag). Sure, it’s a way to make big bucks, but is it fair given that EDs also have hefty “room charges” that seem to cover nothing except having a gurney (often in the hallway) for patients to lie on?

But wait, there’s more. EMRs allow the use of macros and “cloning” (cutting and pasting from prior charts or nurses notes) to create records that appear to defend higher patient professional charges. We all know the lie called “all other systems reviewed and negative.” What a joke. But through the use of macros and check boxes it is easy to create a Medicare Level 5 chart, even when the “medical decision making” clearly does not support Level 5 reimbursement. So, it may not matter that those using EMRs are moving like snails – payments for the few patients seen are going up.

Don’t believe me? Check out the story “Medicare Bills Rise as Records Turn Electronic” from the New York Times. The investigative report claimed that hospitals received $1 billion more in Medicare reimbursement in 2010 compared to 2005 “at least in part by changing the billing codes they assign to patients in emergency rooms.”

For example, at Faxton St. Luke’s Healthcare in Utica, NY, the portion of patients in the ED in which the highest level of care was claimed rose 43% in 2009 – the same year the hospital began using electronic health records. What a coincidence! And Baptist Hospital in Nashville had its share of highest paying claims increase 82% in 2010 – the year after it began using a software system for ED medical records.

As anticipated, spokesmen from the two institutions said the increases reflected more accurate billing for services. I’m getting nauseous.

And is this “coding creep” limited to a few outliers? No way. According to the Times story, hospitals that received government incentives to adopt electronic records showed a 47% rise in Medicare payments at the higher levels for 2006-2010 compared with a 32% rise in hospitals that have not received any government incentives. To be fair, a 32% increase is pretty impressive as well; seems you don’t need computers to check the “all other systems reviewed and negative” box. And if you look at the two highest paying codes, they have increased for Medicare from 40% in 2006 to 54% in 2010.

So the government gets a double whammy – they pay a substantial chunk of the costs of hospital computerization and then get to pay for significantly higher bills for the same services as rendered in the past.

Here’s where things get really interesting. The Office of Inspector General for the Department of Health and Human Services is studying the link between electronic records and billing, and they have no sense of humor. And Aetna and Cigna and state regulators are on to the problem as well. Some contractors who handle Medicare claims have recently warned doctors about their concerns regarding billing practices. One contractor, National Government Services, warned doctors that it would refuse to pay them if they submitted “cloned documentation.” Another contractor, TrailBlazer Health Enterprises found that 45 out of 100 claims from Texas and Oklahoma emergency physicians were paid in error and noted that patterns of overcoding ED services were found with template-generated records.

The Times article tells of a whistleblower suit initiated by Dr. Alan Gravett, a former emergency physician at Methodist Medical Center of Illinois in Peoria. The suit basically contends that the EMR installed in the ED facilitated pulling exam findings “from thin air” and prompted physicians to click the famous “all systems” box even when the exams were rarely performed. The article indicated that the Justice Department was weighing whether to join an amended suit. To put Methodist’s Medicare bills into perspective, billings for the highest level of care jumped from 50% in 2006 to 80% in 2010 (making the hospital one of the most frequent users of high paying evaluation codes).

About 20 years ago there was a series of major fraud investigations regarding upcoding by emergency physician billing companies. The upcoding was blatant and the billing companies that were convicted were given huge fines in addition to being required to return the excess payments. The problem – although the billing companies committed the fraud (and increased their profits as a result of getting a percentage of collection), the money was paid to the emergency physicians for whom it billed. So the government went after these physician groups and some serious nastiness ensued.

The same scenario can occur again and, like in the past, the physician who provides the service is at grave risk, even if there is a contractual agreement to turn the money over to an ED contracting group. So be extremely careful with those macros and check boxes that make all of your chest pain patients’ charts look identical, and as if the famed cardiologist, Michael DeBakey, did the work-up. A day of reckoning is likely coming that will not be pleasant. So don’t lie. Don’t create charts reflecting work you didn’t do. Don’t try to make a simple ankle sprain a level 5 (even a 4 is a stretch) because there is concomitant diabetes and hypertension and hypercholesterolemia.

Bottom line – physician and nurse work in the ED must be facilitated and not encumbered. Slowing physicians and nurses by requiring them to be computer data entry staff is truly nutty. They need to be able to see more patients per hour, not less. They need scribes and techs and PAs and NPs. And thinking that EMRs improve collections is a delusion. If you want to microbill for every band-aid there are much more efficient ways to do so and if you want to upcode “nothing” visits and make them level 4s and 5s, there’ll be hell to pay.