The quest for CPT understanding, continued
A few weeks ago, I wrote about the early days of my quest to understand the new CPT codes for 2013. It was facetious, but I did begin with the idea that I would figure them out so that I could write about them, and perhaps even make a teaching video.
The weeks have gone by, and I have made an effort to learn the new coding and to figure out what changes I need to make in my own office to implement them. In my private practice, I don’t participate in any insurance networks, but I would like to comply with the new changes so that my patients receive fair reimbursement for the treatment they pay for upfront. Because I also work in a clinic, I am required to be in the Medicare network, and I do so as a non-participating psychiatrist, so these patients will be directly impacted by fees that change depending on the content of their sessions.
The question has been raised as to whether out-of-network doctors must comply with the documentation requirements, whether solo psychiatrists who do psychotherapy and have small caseloads will be subject to chart audits, or whether Medicare will bother auditing, nitpicking, denying, or fining small-time docs who miss a few bullet points or simply don’t conform their notes to the Medicare requirements. Questions have also been raised as to what insurance companies will allow – will they reimburse for a 60-minute (53 minutes, actually) psychotherapy session and a 99214 – medical management of a moderately complex problem – in the same visit? What would make more sense, to code a psychiatric evaluation as an E/M code with the many requirements for documentation of bullet points in the history and exam (the numbers of which depend on the code one uses), to code by time and make sure it’s documented that half of the session was spent on counseling and coordination of care, or to code a psychiatric diagnostic exam with medical services, 90792 – similar to the old code of 90801 – without the E/M documentation requirements, but with possibly a lower fee? And oddly enough, the Medicare fees for 2013 pay higher rates if the psychiatric diagnostic exam is done without medical services, so social workers will be reimbursed more for their evaluation than psychiatrists will be, at least at this writing. I have no answers to these questions, nor do I know of anyone who does.
So let me tell you what I did to learn the old coding and what I’ve done to learn the new coding. The old coding was self-explanatory. Mostly, I code 90807, a 50-minute psychotherapy session with medication management. I often run over and go closer to 60 minutes, but that’s fine, and sometimes a patient is ready to go at 45 minutes, and that works, too. For education, I read the sentence. I use a different code if the session is only 30 minutes, or if a condition does not require medications, but reading the description was enough to get it.
For the new codes, I’ve done the following: I’ve gone through the National Council’s 99 slides. I’ve listened to an APA webinar. I’ve watched Dr. Ron Burd’s Vimeo. I’ve attended the Maryland Psychiatric Society’s seminar for 2.5 hours, talked briefly to Dr. Chet Schmidt, taken the APA’s online CPT course, followed another psychiatrist’s CPT blog (and even gone to meet her!). I’ve read more charts and algorithms than I can count, and I’ve read the sections of the E/M manual (several times over) pertaining to psychiatry that was made available on the APA’s free online course. With what I’ve learned, I have written 4 articles and produced a 4-part video tutorial on how to approach this coding for outpatient psychiatrists. In order to do that, I needed to use PowerPoint, iPhoto, Garageband, iMovie, and upload to YouTube and I am quite proud that I figured these out with minimal guidance from one of my techy co-bloggers. Finally, I finished the APA’s online CME course, and after all that, I failed the 9 question post-course quiz.
As the process has gone on, I have become more and more upset. Something as simple coding a psychotherapy session has become unnecessarily complex, and each session now needs 2, or even 3 different codes and the algorithms to figure out what the right codes are and involve pages of charts that layer in on one another. Figuring out an E/M code is like ordering from a Chinese menu, but you’d need an Excel spread sheet to get your lunch.
