Time is Not Always on My Side

My Marine Corps father used to tell me that early was on time, on time was late and late was unacceptable. 

Time is truly a cruel taskmaster. And as it was true for me, it is true for many medical practices, as well. It seems almost counterintuitive that anyone would care about how many hours a physician works. One study reported that, on average, physicians work some 50 hours a week but nearly a quarter of them work between 60 and 80 hours a week. They are raised that way. Many studies conducted on the work habits of medical residents show that it is common to work 80 plus hours per week with some studies showing residents logging over 120 hours per week. And while we may not be surprised, this begs the question; “Why does that matter?”  And the answer is, it doesn’t . . . kind of. 

Estimating the amount of time that a physician works is nothing more than another vector to identify potential compliance risks: that is, physicians who are targets for potential improper billing of services and procedures. 

If physicians are punching the clock, then how is it that we can even think about estimating how much they work? And the answer is, nearly every procedure that involves some form of physician work effort (meaning that it has an associated work .) is assigned some number of minutes. For example, 64580 (Neurostimulator Procedures on the Peripheral Nerves) is associated to a total of 79 minutes (or one hour and 19 minutes). This includes 11 minutes of “pre-service” time, 33 minutes of “intra-service” (or skin-to-skin) time, 11 minutes of post-service time and 24 minutes of follow up time.  And all of this is based on relatively complex surveys that are conducted by the RBRVS Update Committee (RUC), which is run by the American Medical Association (AMA).

In general, the association provides a series of questionnaires, if you will, to a bunch of physicians in a bunch of different specialties and ask them to estimate how long it takes to do a bunch of procedures. And from that, they create this huge database of minutes that are sent along to the Centers for Medicare & Medicaid Services (CMS), which subsequently accepts 95 percent of these recommendations to create the work RVUs. Hence the high correlation between the two. 

The accuracy of the studies has come into question in the past because of the very process described above. Physicians are told to estimate how long it takes for a procedure with the understanding that (wink, wink), the longer it takes, the more they get paid. Some studies have suggested that the total time assigned to these procedure codes may be overestimate by some 50 percent, and that’s a big deal. Why? Because, if these time metrics are being used to target physicians (and they are), then more than a significant of docs are being targeted inappropriately. And I know this as a fact because I have worked as a testifying statistical expert defending these physicians against blind indictments where the only evidence of wrongdoing was a calculation of the number of hours they were working in day (or some other period of time).

There is always that famous case of the psychiatrist that was billing for more than 24 hours of services in a day. And one may ask, “How is that possible?” The answer is because he was billing individuals in a group therapy setting (90853; general group psychotherapy) as individual counseling sessions (code 90832, 45 minutes total). For example, let’s say that s/he is counseling 10 patients in a group setting.  The number of minutes for code 90853 is 24. Let’s say that s/he has four sessions per day.  Using the correct code (90853), s/he would be assigned a total of 96 minutes, or 1.6 hours. Instead, s/he bills each of the 10 patients as an individual session for 45 minutes each. Ten patients in four sessions are 40 patients per day. Forty-five minutes for each patient results in a workday assessed at 30 hours. And for some reason, the government just doesn’t believe that is possible.

Time is the great equalizer and that is why it is such an important compliance risk vector in. 

At the outset, time cannot be used to accurately estimate how many hours a provider works because of the inherent problems within the analytical process itself. For example, RUC does not publish variance so we have no idea, for a given procedure, what the high and low values might be. For example, a physician who only does a few procedures but has been doing them over and over for 30 years may be able to do so at the very lowest end of the time range while a new or inexperienced provider may be at the top of the time range. 

I have defended physicians in criminal fraud cases where the government relied solely upon time assessments to levy an accusation of fraud, only to have the physician acquitted in the end. Why? Because of the inherent inaccuracies within the process and the database. But this should be a wake-up call because even though it may not be accurate enough for determining fraud, it is definitely a risk vector because payers and government contractors treat is as such. 

So, knowing what we know now, how do we incorporate time into our compliance awareness strategies? Having access to the time file, it shouldn’t be that difficult. 

First, assign a time value to each procedure the provider reports (or bills for). Next, multiply those minutes times the frequency for which the procedure is billed. Then, take the sum of those products and divide by 60 and you will have the number of assessed hours for that provider for that period. The question I am always asked is this: How much time is too much time? Since there does not appear to be anything in writing about this, I can only go on my personal experience and on what I have been told by former Office of the Inspector General (OIG) investigators and CMS contractors. Any provider that reports more than an annualized amount of 5,000 hours is at a much higher risk of being audited.  That comes out to two times fair market value of 2,000 hours per year. 

When preparing the time report, you should include the total work RVUs, total time and calculated FTEs for each provider. FTEs are calculated as a multiple of 2,000 hours. So, a provider reporting 5,000 hours would come in at 2.5 FTEs while a provider reporting 1,000 hours per year would come in at 0.5 FTE. If you divide the number of assessed hours by the number of days in the period studied, you can calculate the number of assessed hours per day.

Using all the days in the period would be equal to a seven-day work week but this is rarely the case. If you change that variable to say, five, then the number of hours per day are going to increase, as expected, but in any case, the idea is not to manipulate the results, so they look as good as possible but rather to understand how an outside third party interprets the data. It is also important to understand the difference between total time and intra-service time. 

For example, total time includes longer-term follow up care and most often, these services are provided by an APP and not the physician provider. This may also be true for other portions of a service or procedure. When that is the case, then look at the intra-service time. In evaluation and management (E&M) visits, this represents face-to-face time and in surgical procedures, it represents skin-to-skin time. And it may just be that intra-service time is the best and most accurately represents how hard your providers are really working. Again, these time values are for assessment and are likely overstated so before you start pointing the fraud finger (like the government tends to do), remember that the purpose of the time assessment is to give you a prior look at what an outside auditor might be interested in investigating.  Even when a provider is reporting over 24 hours a day, it is rare that this is due to fraud (i.e., billing for services that were not provided) but rather a chink in the time model.  What it should do is put you on notice that there is a higher risk for an audit. 

The time reports are so valuable because they equalize the risk no matter the specialty, no matter the provider’s experience, no matter how special their services are. A surgical specialist may be performing some really arcane procedures, but the reality is, you can still only do so many in a day. And for some specialties, we regularly see overstated time assessments, such as for  Mols or orthopedic surgeons. This is just the nature of the beast and while we may know that there isn’t any funny business going on, an outsider won’t until the audit is complete. 

In the end, you should neither ignore the time analysis nor take is as literal gospel.  It is what it is and what it is, is another vector to assist you with identifying potential coding and billing risk.

After all, that is what risk-based auditing is all about.