For today’s look at Recovery Audit Contractors, I thought I’d start with some essential truths about the RAC’s and their function.

I see RAC’s as golems. While the four regional RAC’s are now self-sufficient, they are a creation borne out of a recognized need to control dollars paid from the Medicare and Medicaid programs. They are given their shape by Medicare’s proprietary software and as an extension of limited investigations, either by the OIG or the Medicare Administrative Carriers (MAC’s), that have already been completed. In other words, the creator starts small with a scale model of an audit, followed by their man-made, unfeeling behemoth that is the RAC bringing down their earthen fist and smiting everything in its path of a similar nature. The current appeals success rate for RAC determinations strongly indicates that the golems do so indiscriminately, without an instinctive need to be correct.

In anticipation of the inevitable RAC investigational expansion into physician services, experts in the industry have trained their attentions on the Comprehensive Error Rate Testing (CERT) Reports. These reports are a good indicator of what services require additional investigation, but there are a few stops in between a CERT and a RAC that further narrow the focus of the eventual RAC investigation. A quick review of my inbox this morning revealed one of these interim studies, the results of which illuminate the world of RAC audits like an H-bomb test on Bikini Atoll.

WPS, the current MAC for Jurisdiction 5 (covering Iowa, Kansas, Missouri and Nebraska) and the legacy MAC for Illinois, Michigan, Minnesota and Wisconsin, released the results of a Service Specific Probe of CPT code 99214 for the specialty of Family Practice this past Monday. 100 such services were randomly selected for prepayment review. Of these, 52 were allowed as billed following documentation review. Based on the utilization of this code, that number seems low. With the belief that we learn from our mistakes, I now present the results of the remaining 48 claims.

11 claims were down-coded based on the documentation provided. To review, a level 4 established patient visit requires that two of the following three elements be satisfied by the documentation:

• Detailed history
• Detailed examination
• Moderate medical decision making

While not specifically stated in the CMS E/M guidelines, with established visits, it is always a good idea to have medical decision making be one of the two elements selected. I recommend this based on medical necessity most often being defined based on treatment options selected for the condition being treated. Even in a patient with a list of comorbidities written on a 3-foot scroll, if a patient has a bit of a rash, the greatest history ever taken and an examination and auscultation of every square inch of a patient is still treating a bit of a rash, and the E/M code selection needs to reflect this.

2 of the claims in the study were determined not to support the billing of an E/M service based on the documentation forwarded for the study. If I had to venture a guess, I would say that these were related to encounters where a minor procedure was planned upon scheduling, the patient presented for the procedure, and the physicians in question billed both a procedure and an E/M service.

The remaining 35 claims represent a different kind of hurdle for physician practices. These claims were denied outright because the providers did not provide the requested documentation for the services within the alloted 45-day period. This study included only 100 claims. If we expand that number out by a few zeroes, apply it to RAC documentation requests and then extrapolate that 35% of these requests will either be mishandled or ignored, the obvious conclusion is that by virtue of their internal practices, physicians are doing the RAC’s work for them. Who knew doctors had this kind of time on their hands? 

Part A providers knew the RAC’s were coming, and any facility worth its salt set up processes long ago to respond to RAC requests. It is my personal belief that a connection can be made between provider readiness for RAC audit requests and the so-far successful appeal rate of RAC decisions by Part A providers. A physician practice, in most cases, lacks the organizational infrastructure to prepare to respond in the same way as Part A providers. A solution to response readiness is not – and in many ways, cannot be -  a one-size fits-all proposition. A good start would be educating administrative staff to be able to recognize a RAC request upon receipt.

With a 48% error rate in this limited probe, it is safe to say that high level established patient visits are now officially warming up in the RAC bullpen. A surprise occurs when you never see it coming. Like the relief pitcher who replaces the obviously tired starter, we saw them warming up. The RAC’s, in the same manner as the next pitcher jogging in from the outfield, is easy to see. In fact, based on their size, so are most golems.

As a strict adherent to the world of science, I’m often frustrated by human imperfection. I try to do my level best to surround myself in my personal life with people who demonstrate that they are continuing on a journey towards a higher level of human achievement. In the end, due to the acute level of imperfection that exists around me, this makes me something of a lousy friend, while at the same time making me a razor-sharp critic. This is my imperfection, and I attempt to overcome this with equal parts good humor and beer. Sometimes, one works better than the other.

Yet the undisputable fact of human imperfection doesn’t necessarily mean that under the right conditions, a human can’t achieve something perfect every once in a while. Baseball pitchers throw no-hitters, musicians record the perfect song, or the rare student will get a perfect score on their SAT’s. These achievements are rare, but behind each one lies hours upon hours of repetition, self-examination and exploration.

From this idea, we look at the documentation for initial inpatient encounters, represented by CPT codes 99221 thru 99223.

As stated in last week’s edition of The RAConteur, CPT code 99223 has revealed itself to be paid incorrectly 21.5 percent of the time according to the results of Medicare’s Comprehensive Error Rate Testing (CERT) program, which places this code at the top of the list. The Recovery Audit Contractors have recently added medical necessity to their stable of audit issues, which virtually guarantees that CPT code 99223 and how it is documented will be a focus when the RAC’s begin to expand their focus in the physician arena.  

As if we didn’t have enough to worry about with this particular set of codes, 2010 marked a major change in the usage of these codes with Medicare’s elimination of reimbursement for consultations. High-level inpatient consultations are now reported with the CPT codes for initial inpatient encounters, driving utilization of this code set through the roof. What two years ago appeared to only affect services by the managing physician during a patient stay has now rapidly expanded to include initial specialist encounters in the inpatient setting. Prior to the elimination of reimbursement of consultation codes, the CERT program determined that up to 50% of consultations billed in some jurisdictions were billed incorrectly. While some of this error rate could be chalked up to poor documentation of the request and report, quite a bit more had to do with the level of service not being supported by documentation. My suspicion is that this problem is still pervasive with the change to initial inpatient encounter codes, as the underlying documentation issues may still not be remedied.

With all of this as a backdrop, let’s take a look at the documentation for these services, particularly 99223.

In order to bill a 99223, the documentation requires a comprehensive history, a comprehensive examination and medical decision making of high complexity. For the first two elements, this means describing the presenting problem at least four ways, reviewing at least 10 of the 14 body systems, documenting a past, family and social history and performing an exam on at least 8 body systems. This is an opportune time to point out that if your history and exam does not include everything catalogued above, you cannot bill a level higher than a 99221. Given the stark differences in reimbursement between 99221 and 99223, I can’t begin t0 stress this point enough. Anything short of a comprehensive history and examination means you cannot bill higher than a 99221. Additionally, if your history and exam fail to reach a detailed level, Medicare regulations instruct providers to bill with the code for unlisted E/M services (99499) along with documentation, with reimbursement determined based on the veracity of the documentation (translation: not much).

 Now that you have a service with a comprehensive history and examination documented, the determination of the level of service to be billed now depends on the medical necessity and decision making for the service. If testing of any kind is scheduled for the patient, you are now halfway to attaining an audit-proof 99223. It is the other half of this equation where the problem lies, which is determination of patient risk and treatment options. This is where a philosophical discussion ensues.

A patient admitted to a hospital has been discovered to be in a physical state that is beyond self-care, necessitating a hospital stay. The burden of proof for the hospital stay falls squarely on the shoulders of the physician managing the patient upon admission and beyond. When it comes to documentation of treatment course, a lack of detail can do much more than decrease your reimbursement for the initial encounter. Poor documentation of treatment course, or a deviation from the documented plan of treatment, has the potential to put the hospital’s reimbursement in jeopardy if medical necessity isn’t established for the length of stay.

With this set of facts established, being honest about the patient’s treatment course is the best path to accurate billing and “audit-proof” reimbursement of the initial encounter. In simple terms, tell me where the patient is medically right now, and what you plan to do about it.

Of course, no discussion of E/M is complete until we discuss coding by time. The average total time of a level 3 initial inpatient visit is 70 minutes. Unlike office encounters, hospital floor time dedicated to the admission and treatment course of the patient can count towards the total time of the encounter if billing by time, and if more than 50% of the encounter was spent in counseling and coordination of care.

The RAC’s are teaching us through their first medical necessity audits that no service performed for a patient in the inpatient hospital setting will be overlooked if necessity isn’t properly established. Human imperfections can and will continue on the road of human evolution. Increasingly, less-than-perfect initial inpatient visit documentation will not be welcome along the road. With repetition and internalization of what is needed, perfection need not be an idea far, far away.

I’ve been around long enough in a society so chock full of people equally split between those with neuroses and those in denial to learn that there are varying definitions to the term “seriously ill”, depending on who you happen to encounter in a given day. My Aunt Edith, a lovable lady to her core, so enjoyed talking about her various maladies that she could describe her latest paper cut in terms usually reserved for near-amputations. Yet in the end, a band-aid and a few days was usually the best treatment option.

Having audited inpatient documentation, I can draw a parallel between poor Aunt Edith’s fingers and a physician selecting a level of service for a subsequent hospital visit. The preponderance of evidence suggests that in both cases, there’s quite a bit of self-defeating exaggeration to be found.

When someone asks me what I believe to be the most over-reported E/M service, it takes me very little time to say “99233″. I can count on one hand the number of times where I have seen this code reported where the documentation actually justified the billing of a level 3 subsequent hospital visit.

The Medicare Comprehensive Error Rate Testing (CERT) reports bear this belief out to some extent. CPT code 99223 (level 3 initial encounter) was shown to be overpaid 21.5 percent of the time, but coming in a close second was 99233, with an error rate of over 18 percent. I have seen poor documentation for both of these codes numerous times, and I can tell you that 99223, from a documentation perspective, can be fixed with provider education in a fairly short time frame. For 99233, the documentation in most cases doesn’t even come close to substantiating the reporting of the code. The nature of subsequent hospital visits being handwritten makes correcting the problem a laborious task from start to finish.

When the topic of subsequent hospital visits comes up, from an educational standpoint, I like to go back to the very beginning, which is the CPT definition for each of the three codes. In the CPT, at the bottom of each code description, is a small paragraph which gives an overview of a typical patient’s condition for the code. As the level of service goes higher, there is an acuity level that must be met to satisfy the usage of the code. To start, 99231 states “Usually, the patient is stable, recovering or improving”. Moving forward, 99232 states “usually, the patient is responding inadequately to therapy or has developed a minor complication”. Finally, 99233 states “usually, the patient is unstable or has developed a significant complication or a significant new problem”.

If you are currently a provider billing a large volume of 99233’s, take a good look at these descriptions. In a philosophical sense, I am in agreement with the idea that a hospital stay is usually reserved for those whose illness has exceeded the person’s capacity for self-care. Yet it should not be a common occurrence for a patient whose illness has been identified upon admission and whose course of treatment has been defined to be unstable throughout a large portion of their hospital stay. Utilizing the definitions above, 99231 and 99232 would describe the large percentage of subsequent hospital visits. To put it in the context of a recent example, Aunt Edith seems like a 99233, but is more than likely a 99231.

Additionally, if you’re billing a 99233 a day prior to discharge, I would hope that your discharge documentation includes the words “magic wand”, as you are reporting that the patient is unstable and/or has a complication of care. A discharge the following day would be unlikely.

Based on the direction that CMS has been taking with hospital-acquired conditions (HAC’s), I can see a day when 99233’s are only paid within 1 to 3 days of admission. Currently, taking HAC’s into account is in a reporting-only stage, but I believe we are headed towards a time when hospitals will face little (or perhaps, no) reimbursement for the treatment of HAC’s. This is a fact that, unfortunately, cannot be remedied by improved documentation on the part of the managing physician.  

We now have three of the four recovery audit contractors looking at certain hospital DRG’s for medical necessity, with the current list expected to expand over time. As RAC’s begin to widen their efforts into the physician arena, it would make sense for 99233’s to be on their immediate radar. This is particularly true if they have determined that a hospital stay did not fall under the parameters of medical necessity based on a complex review of the hospital records for an inpatient stay.

For the tens of thousands of physicians who happen to come across this particular posting and say, “…but I spend a lot of time with my patients, and….”, I offer this. While there are average times for each of the three codes for subsequent hospital visits (to save you time, 15 minutes for 99231, 25 minutes for 99232 and 35 minutes for 99233), it is important to remember a few points. First, the documentation must state “I spent XX minutes with this patient, more than 50% of which was spent counseling and/or coordinating care”. Second, there must be documentation of what topics were covered during the counseling (weather and other local and national news updates do not count for purposes of coding), or what specific steps were taken to coordinate the care for the patient. Simply writing something like “spent time counseling” is insufficient.

Finally, know that RAC’s are enforcing CMS’ recently implemented and heightened signature requirements when auditing your documentation. If it does not include a recognizable signature, your service is not authenticated and did not happen from the RAC’s point of view.

To complete our cycle, next week’s edition of The RAConteur will take a look at the reporting and documentation of  initial inpatient encounters (CPT codes 99221 thru 99223).

Having lived in and around Philadelphia for most of my younger days prior to living in Milwaukee, three things still hold true about me: I prefer cheesesteaks and hoagies to bratwurst, I drive extremely fast compared to others in my new adopted home and I’ll forever be a rabid sports fan.

My hero growing up was Steve Carlton of the Phillies, a pitcher so inscrutable and so focused on getting people out that the only nickname teammates could come up with for him was “Lefty”. With each passing day, in a world of pitch counts and fanatical devotion to quantifying every at-bat with statistics, I appreciate Carlton that much more. He was the classic power pitcher, going in for at least 7 innings, throwing a baseball as if his life depended on it, then strolling off the field with his eyes looking down to the unexpected applause of a sports city known for such class acts as booing the Easter Bunny and pelting Santa Claus with snowballs.

Every one of his 4,136 strikeouts was a testament to mental discipline and a workout regimen that included, among other circus-like feats, driving his fist to the bottom of a bucket of sand. This devotion to his workouts yielded 329 wins and a ticket to the Hall of Fame.

So what does Steve Carlton have to do with the Recovery Audit Contractor program? Bluntly, it’s about time you asked.

Most physicians, based on the years of training involved with becoming a member of their profession, possess a certain level of discipline. Fortunately for their patients, and unfortunately for themselves, this discipline often includes only the practice of medicine and does not extend to the documentation of the services rendered to their patients. The RAC’s have already taught us many lessons, but one painful lesson on the horizon is how much of a disservice physicians do to themselves by not capturing all that they do in a clinical setting in the patient’s medical record.

As RAC’s now expand their focus to include medical necessity, with more attention being paid to physician services, there is a school of thinking that the best beginning course of action for physicians to identify their RAC vulnerability is a review of CMS’ Comprehensive Error Rate Testing (CERT) reports and specialty peer comparison of billed services. Unequivocally, I agree that this is indeed the best place to start, but once this first step is completed, the hard work begins. Specifically, it’s time to take a long and critical look at the documentation for your services. This is the point, in today’s and subsequent postings, where I’d like to insert myself into the larger conversation.

Creating “audit-proof” documentation is far from an easy task. For one thing, “medical necessity”, as any person at all connected to medical delivery will tell you, is not a one-size-fits-all proposition. Similar symptomology can have different effects depending on the presentation of the patient. This inconvenient fact places the burden of proof for medical necessity squarely in the hands of the treating physician and his or her documentation of the services rendered. With CERT reports indicating that the biggest risk area for physician audit being evaluation and management services, it is long since past the time to look at documentation of these services.

It has been 15 years since the first E/M guidelines were released by Medicare. With the volume of writings dedicated to documentation of E/M services, you would think that physicians would be at least halfway to resolving documentation inconsistencies, but this simply isn’t the case. It is my belief that the problem lies not in wanton and deliberate physician non-compliance, but rather a lack of an attempt by CMS to translate the E/M guidelines into usable clinical language that physicians can understand and utilize.

With mandatory electronic health records on the horizon, the absolute worst line of thinking a physician could internalize is “the EHR will handle it”. EHR’s possibly offer a solution, but only if set up in the clinical setting to be utilized properly. EHR’s can lead to such things as templated and cloned documentation, which in established patterns can bring exponentially increased risk. Remember that the main reasoning behind the government’s push for EHR is simplicity and portability of retrieval, and not as a panacea to documentation of services.

Because I tend to look at RAC audits of physicians from this angle, I’d like to offer a preview of next week’s edition of The RAConteur, where I’ll begin taking a look at the most common CERT E/M errors and offer analysis you can use to make better decisions about code selection, beginning with subsequent hospital service codes 99231 thru 99233. I can promise the reader ahead of time that finding permanent and lasting answers will require discipline, but much less than is required to plunge your fist to the bottom of a bucket of sand.