A frequently asked question by cardiology programs across the country is how to determine the right mix of electrophysiologists (EPs) to cardiologists. Although the answer to this question is multifactorial and certainly has regional and even local nuance, we can look to the ratio of the overall MedAxiom database for a baseline. When doing so, we find in Figure 1 that there are 395 EP physicians compared to 2589 non-EP cardiologists in the overall database, for a ratio of 1 to 6.55 FTEs.
Certainly this ratio should not be a surrogate for detailed local workforce planning and analysis, but it can be used for reality checking your program’s EP service.
For the first time since 2011, EP wRVU productivity increased in 2014, though only slightly at just 1.1 percent (Figure 2). EP was the only one of the four cardiology subspecialties to see a gain in wRVU production in the survey year (Figure 3). Despite this increase, wRVU production amongst EP physicians is down more than 10 percent from the high water mark of 12,991 per full-time equivalent (FTE) physician in 2011.
Most, if not all, of the EP declines in wRVUs can be attributed to coding changes known as “bundling” by the Centers for Medicare & Medicaid Services (CMS). Bundling refers to formerly separate CPT codes that are combined together into a single code. When this occurs, the aggregate wRVUs — and therefore the reimbursement — almost always decreases. Figure 4 shows recent bundling activities and the impact on wRVUs.
As can be seen in Figure 3, EP holds the top spot in terms of wRVU production of all the cardiology subspecialties. However, in terms of overall compensation, EP ranks second behind interventional cardiology (Figure 5) and is last in terms of compensation per wRVU (Figure 6). This latter statistic may be influenced in part by groups that have significant sharing as part of their internal distribution plans. Even in the integrated and employed models, a large proportion of groups have control of the total compensation pool and can therefore determine how it is split at the individual doctor level (Figure 7). Figure 8 then shows that a significant portion of these groups include some form of sharing as part of their internal distribution plans. With EP as the highest wRVU producer, this sharing will mathematically bring down the resulting calculation of compensation per wRVU.
The gain in overall compensation for EP shown in Figure 5 is due nearly entirely to an increase amongst integrated (employed or in a professional services agreement with a hospital or health system) physicians, as EP doctors in private practice saw gains of less than 1 percent. Like the other subspecialties, EP physicians in the integrated model fare substantially better than their private peers, outearning private EP physicians by nearly $115,000 annually (Figure 9). This despite the fact that private EP physicians outproduce integrated EPs by over 650 wRVUs annually (Figure 10). Predictably then the integrated cohort earns nearly $14 more per wRVU than private EP physicians (refer back to Figure 6).
One of the strongest indicators of a healthy cardiology practice and a strong determinant of nearly every other procedure measure is total new patients to the practice. As shown in Figure 11, after two years of relatively low new patient volumes, 2014 saw an increase back to the 2010 level of 567 total new patients to the practice per FTE cardiologist. This is a good sign for EP physicians, as this new population of patients will undoubtedly drive up EP procedure volumes, as has been the case in the past.
When looking at volume ratios, it is important to consider the most appropriate denominator, particularly when trying to determine accurate trends. For instance, the historical denominator for most cardiology volumes has been the number of cardiologists, as measured by FTEs. However, this denominator is rife with problems when trying to consider an apples-to-apples comparison. For instance, we know from data that not all physicians take the same amount of time off each year and the variance is considerable. Figure 12 shows that at the 25th percentile, cardiologists are taking off around 5 weeks per year, whereas at the 75th percentile, that jumps to 8 weeks or twice as much.
To suggest that physicians in these two cohorts are the same for procedure comparison purposes seems fraught with inequity and may lead to bad conclusions. In addition to time off, there are significant differences in individual productivity, geographic differences, age distribution within the group, and a host of other imbalances. For these reasons, a better denominator would be one that attempts to accurately measure the population of cardiology patients, not the number of physicians serving them.
To capture this, in 2013 MedAxiom began tracking a new cardiology metric called panel size, which attempts to measure unique active patients in a cardiology practice. Because of the significant sharing of patients that occurs within cardiology, MedAxiom measures panel size at the group level, not at the individual cardiologist level. The definition of active patients in the panel is “unique patient Evaluation & Management (E&M) visits within an 18-month period.” Thus, a patient seen four times in the office during the defined 18-month period would count as “1” for purposes of panel size. Likewise, a patient who was never seen (and billed an E&M code), but who had an EKG interpreted by cardiology, would not be considered an active patient in the panel size calculation.
Although far from a perfect definition and certainly likely to evolve over time, at present panel size may represent the most accurate measurement of patients managed within a cardiology practice and therefore may provide the most accurate denominator for volume analyses. Because this measure was introduced with the 2013 MedAxiom Provider Compensation & Production Survey, trending is not yet available, but will be with the 2015 survey. Figure 13 shows several key EP volumes using panel size (per 1,000 patients) as the denominator.
Another measure of the active cardiology patient population is the total number of cognitive (E&M spectrum of the CPT codes) encounters within a group. The logic for using this measure as an estimate of the active cardiology patient population is that patients requiring active management will inevitably end up in the exam room. We exclude nursing visits (99211) in this calculation, as a patient who is only seeing non-provider (physician or APP) staff would not be considered “actively managed.”
Using total cognitive encounters as the denominator and then calculating a ratio of procedures to cognitive encounters results in the trends shown in Figure 14 for several key EP volumes. Of note is the very precipitous drop in ICD implant volumes seen from 2011 to 2012, a direct result of the Department of Justice’s (DOJ) focus on these expensive devices that occurred at that same time. Despite a slight recovery, ICD implant volumes have never returned to their pre-DOJ investigation levels.
On the positive trending side is the ratio of ablations to cognitive encounters, which has now trended up for five straight years (Figure 14). Pacemaker inserts are at best stagnant if not declining slightly.
Electrophysiology physicians saw very slight gains in overall compensation from 2013 to 2014, and continue to be the second highest-paid cardiology subspecialty, trailing only interventional cardiology. Reversing a 5-year trend, EP wRVU production increased in 2014, besting the 2013 median levels by just over 1 percentage point. Although this gain is very modest, it may signal the end of CMS bundling efforts that have hit EP wRVUs hard over the past four years, which would be a welcome reprieve. EP key volumes show a mixed bag with ICD implant volumes continuing to erode, while ablation procedures trend upward.
Joel Sauer is Vice President, Consulting at MedAxiom Consulting. He works with organizations across the country in the area of physician/hospital partnerships. His work includes full-service line development, co-management arrangements, and integration transactions. Mr. Sauer may be reached at firstname.lastname@example.org.