Guest Post: Get CMI data ready for when physicians ‘squirrel’ to evade efforts

By Howard Rodenberg, MD, MPH, CCDS

Editor’s note: Today is National Squirrel Day, so the ACDIS editorial team thought we’d give you some surprise Sunday reading to go along with the celebration. Rodenberg is the adult physician advisor for CDI at Baptist Health in Jacksonville, Florida. Contact him at howard.rodenberg@bmcjax.com. Advice given is general. Readers should consult professional counsel for specific legal, ethical, clinical, or coding questions. Opinions expressed are that of the author and do not represent HCPro or ACDIS.

Everyone’s probably seen the literature rack at the tourist welcome centers along the highway between one state and the next. These small wooden fixtures contain stack upon stack of neatly slotted brochures designed to lure travelers into making just one more stop to see the pride of (fill in the name of your small town here). Picking up these brochures has long been an addiction of mine. There’s a whole tote bag in my bedroom closet full of information about sights I have yet to see. I figure I’ve got about 15 years before someone calls Hoarders and puts me on TV.

This is how I knew that I needed to stop one day in Greenville, Illinois, and take a selfie next to the World’s Largest Golf Tee. Why I know that WaKeeney, Kansas, is the Christmas City of the Plains, that Clark’s Fish Camp in Jacksonville, Florida, houses the Nation’s Largest Private Taxidermy Collection, and why I know about the White Squirrels of Olney, Illinois.

I bring up squirrels because I recently had occasion to use the term in the context of CDI. I was involved in a discussion about using case-mix index (CMI) data on individual physicians to “call out” their performance.

As a CDI professional, I know why this could be important. If you have physicians who consistently underperform, it may be helpful to confront them individually, either privately to let them know you’re watching, publicly to leverage peer pressure, or with their supervisor to use third-party influence to achieve CDI goals. All of these are perfectly acceptable ways to address individual physician deficiencies. (They will all also annoy the very physicians you’re trying to influence, but the decision to “call out” individual providers is a philosophical decision each CDI program makes for itself.)

From the physician side, however, our first reaction is to turn into a squirrel. Squirrels scamper. They avoid, dodge, evade, and run for cover. And when cornered, they give you a look so gosh-darned cute you simply can’t hurt them no matter how many times they’ve raided your bird feeder. Similarly, when confronted with adverse information about themselves, doctors squirrel. We do so because we’ve been taught to squirrel in medical training as a way to avoid confrontation or being caught in the wrong. We have the intellectual heft to pull it off and we’ve been doing it our entire professional lives.

Let me give you an example from my own experience. I’m an emergency department doc by trade and training, and I’ve sometimes been on the wrong end of those time studies looking at patient throughput. My times were longer than the mean; in fact, close to the bottom of the group. So, what happened? Did I take it to heart and resolve to change my ways? Did I take a serious look at my practice style and gratefully acknowledge the input of my health care colleagues, especially those who issue the mandates but cannot actually do my job?

Of course not. I squirreled.

“Thanks for the information. I’ve taken a look and I have some thoughts. First, I work all nights. As you know, there are no mid-level providers at night. As a result, every doctor has relatively more patients to see per provider, and it’s well known that the more patients you’re caring for at any one time the more time it takes to care for each.

You’ll notice the data shows that all of us night shift doctor times are longer than the day shift people. You’ll also notice that when I do get to the patient, I see them faster than anyone else on night shift. That’s because instead of signing in for the patient when they get placed in a room, I don’t do it until I actually see them to avoid any errors in orders or documentation in the interest of patient safety. (The phrase “patient safety” always gets you off the hook.)

Also, as I’m sure you know (you have to throw in at least a sentence or two of “collegial language”), patient flow in the ER is non-linear, so the fact that patients come in as a bolus in late afternoon and early evening rather than presenting to an empty ER early in the morning may skew the data as well. Thanks again for the note, and I look forward to continuing our efforts in patient care.”

I never heard anything again and several months later they dropped the measurement altogether. I imagine that I wasn’t the only squirrel in the forest.

So, if you’re going to call out physicians, you have to be prepared for this. They will find reasons to prove that you’re wrong, and you have to realize that sometimes they might actually be right. For example, let’s talk about an orthopedic surgery group. The metrics of some doctors look great, with high CMIs; the CMI of others is much lower. Call out the slackers, right?

Not so fast, though. What kind of procedures do the lower performers do? If the higher CMI guys do hips and the lower folks do shoulders, they’re not really lower at all; they just have different patient populations. The same can be said of a hospitalist who works nights and mostly does admits with only the rare discharges. How about a cardiologist who does few admits but mostly consults? The surgeon who doesn’t do his or her documentation but has the mid-level provider do it for them? What does his or her CMI actually mean?

How can you make sure that your “call-out” is really valid? First, you need to make sure the patient populations are roughly the same. You can probably do some fair comparisons on adult and pediatric hospitalists, as with decent volumes you would think that the “luck of the draw” would give you roughly equivalent groups.

On the surgical side, however, that’s harder to do, especially if you have only one or two physicians who do a certain procedure at your hospital. Having tools that provide some peer cohort facility measurements can be helpful as well, but be aware that the squirrels will still find ways to gnaw away at the acorn of data. 

You’ll also need to make sure that the documentation is really under the control of the physician. I say this fully cognizant that from a coding point of view, the physician is clearly the one on the line. But in reality, much of inpatient documentation may be done by mid-level providers, and the physician simply adds an attestation and signs off on the note. In an ideal world, the doc would review every notation with a nit comb, but it simply doesn’t happen. (Procedures make money, not post-op notes.) So, be prepared for the squirrel that says documentation is the mid-level’s problem, and instead of a correction, offer to extend your educational efforts accordingly.

Another trap to avoid is looking at short-term data. There’s a virtual cornucopia of factors that can affect physician documentation, such as vacation, sick call, seasonal variations in patient populations, and simple physician fatigue. I believe that a full three months of data is the minimum you should use to evaluate physician performance, but the more data you have, biannually or even yearly, the more reliable your trends will be. More data over longer time frames adds validity to your interpretations and tends to defuse some of the firestorms that result.

I’m personally very interested in statistics (nerd alert!), and I’ve even tried to apply the concept of statistical significance to identifying outlying providers based on their CMI compared with peers. I’ve not found any sharp demarcations between our providers, but if a clear outlier is present, the use of statistical techniques can help support your concern.

The bottom line is that if you’re going to call a squirrel a squirrel, you had better do your homework first and be prepared for evasion, rationalization, and the like. Squirreling is part of human nature, and doctors are particularly good at it. Be prepared. Oh, and don’t fall for the cute thing.