As chair of the American Academy of Dermatology’s Patient Access and Payer Relations Committee, I traveled around the United States for three years with other well-versed dermatologists and explained the value of dermatology to insurance company medical directors (thanks to dermatologists James Zalla, Scott Collins, Howard Rogers, Alexa Kimball, Clifford Lober, Sabra Sullivan, Mark Lebwohl, Beth Lertzman, Bruce Brod, Carrie Kovarik, Brent Moody, George Hruza, and Carl Johnson).
We showed them the statistics, clinical guidelines, and clinical photos and explained how cost effective dermatologists are in treating skin disease. We argued against using blunt tools, like average provider expense, as a proxy for quality. I thought it was a pretty compelling story, but the medical directors always asked for a reproducible quality metric. Almost no one in specialty medicine has reproducible quality metrics, and these are very difficult to develop.
Finally, on April 28, in JAMA Dermatology, the American College of Mohs Surgery ( ACMS ), with its Improving Wisely Study Group partner Johns Hopkins University, published a quality metric using the ratio of Mohs layers, which can be used by physicians, payers, and patients alike (2017 Apr 28. doi: 10.1001/jamadermatol.2017.1450 ). The effects of this metric on practice behavior can be monitored in real time.
The average number of layers taken for Mohs surgery of head, neck, hands, feet, and genitalia was calculated for all physicians reporting the codes to Medicare from 2012 to 2014. The Accreditation Council for Graduate Medical Education training programs were separately analyzed, since theoretically, they should get referrals of the more complex and difficult cases.
The average proved to be 1.74 stages per case, with a median of 1.69. Of 2,305 physicians billing for Mohs surgery, there were 137 extremely high outliers in at least 1 of 3 years, and 49 persistent high outliers (greater than two standard deviations in all three years), who averaged more than 2.41 layers per case. There were also 92 extremely low outliers (1.28 stages per case, in at least 1 of the 3 years), 20 of whom were persistent in all 3 years. High outliers were more likely to work in a solo practice setting.
The Improving Wisely program is based on the concept that reducing unnecessary variations in care can improve patient safety and quality of care while also reducing costs. It works on the premise that many outliers are unaware they are an outlier, no one wants to be an outlier, and confidential, collegial education and peer mentoring within a medical specialty society can reduce unnecessary variations in care.
Last month, all ACMS members received confidential data reports with their own personal ratio in relation to the entire cohort. Educational and mentoring resources are available for members, and outliers are encouraged to engage with the ACMS to identify opportunities for modifying and improving their practice patterns.
Now, these numbers must not be taken as an indictment of anyone. They are for educational purposes, and the goal is to help identify physicians who are unaware of their deviation and bring these outliers back into the norm. Supporting this premise, solo practitioners were at greatest risk of being outliers, which may be explained by lack of collegial interaction, peer review, and feedback.
In addition, there may be good reasons for being an outlier depending on one’s patient population, and the ACMS is interested in examples. The Mohs College is devoting considerable resources to help outliers. Much of this variation may also result from incorrect coding or processing of specimens. Nonetheless, no patient, payer, or physician wants unnecessary surgery or avoidable charges.
Low outliers are particularly puzzling, since someone would need godlike abilities to almost never have a positive margin in Mohs. I have heard of some practices whose patients all present in the morning, one layer is performed, and the rest of the day is spent processing and interpreting their slides. (Mohs is time consuming.)
I am also aware of some rural providers who travel to distant sites, take a layer, return to the city to process the tissue, and return a few days later to complete the case. This may not indicate bad care, just an unusual practice pattern or adaptation to difficult circumstances. However, it must also be noted that not completing cases on the same day could result in increased payments because of the coding system that reimburses more for first stages than for additional ones.
You must be aware that all these numbers, with a two-year lag, are available to physicians, payers, and patients. If you don’t know your ratio of first to additional Mohs layers, I encourage you to look your numbers up and calculate your ratio (CPT code 17311 plus 17312/code 17311). The easiest website to use is provided by the Wall Street Journal. If you are an outlier, you should ask yourself why, and consider some peer review and other appropriate changes. If your patterns change, they will be noticed quickly, since Johns Hopkins and the Improving Wisely program has leveraged their relationship with the Robert Woods Johnson Foundation to gain more immediate access to current Medicare data.
Everyone is hoping we see normalization of the patterns in layer usage, since this will give great credibility to Mohs surgeons and be better for our patients and the health care system in general. Kudos to the ACMS and the Jama Dermatology paper’s senior author, John Albertini, MD, of Winston-Salem, NC, in particular, for making this benchmark become a reality.
Dr. Coldiron is in private practice but maintains a clinical assistant professorship at the University of Cincinnati. He cares for patients, teaches medical students and residents, and has several active clinical research projects. Dr. Coldiron is the author of more than 80 scientific letters, papers, and several book chapters, and he speaks frequently on a variety of topics. He is a past president of the American Academy of Dermatology. Write to him at dermnews@frontlinemedcom.com.
