Using Quantitative Measures to Guide Workforce Planning (Part 3 of 3-Part Series)

Posted by: Joel Sauer on Tuesday, May 3, 2016
Chalk Talk

As was described in Part 1 of this blog series, workforce planning requires a review of both qualitative measures (culture, expectations, future plans, strategy, etc.) and quantitative metrics. This latter group includes specific data we can generate around demand, capacity, role needs and future projections that will impact our workforce. 

Measuring Current Work Output

One of the first steps in the quantitative analyses for workforce planning is measuring the current work output of the providers. Figure 1 provides a sample of output measurement using work RVUs as the metric. The data provide us with a three year lookback of wRVU production by sub-specialty, along with a comparison to our peers. At the bottom of the figure, you can see where this group ranks in its wRVU output compared to its like peer group. Several important data points emerge from this comparison. 

First, we can see that the group is experiencing a downward trend in overall wRVUs. There are multiple factors that can lead to this, which deserve further exploration and consideration, but some typical culprits are loss of market share, CMS changes to the wRVU schedule, capacity constraints, and so on. 

Second, we see that from the physicians’ side, the three-year average production puts them at the 59th percentile of their peer group; however, for the latest year that ranking dips to the 53rd percentile, or just above the median.  This is important when we consider our qualitative analyses and expectations for work. If that bar is set at the 75th percentile, the data show our group has significant capacity—not equally distributed—but in total. If our expectation is at the median level, then our group is at capacity.

Our future physician workforce will not look or act the same as it has historically. This isn’t bad or good. It just is.”

Third, we can see the output of the advanced practice providers (APPs) compared to peers.  Here, too, the group ranks above the median when compared to its peers, with a slight upward trend over the three years. At the same time, the production per APP is relatively low when considering that an APP with a relatively full schedule billing under his/her own provider number can generate around 3,000 to 3,500 wRVUs.  If the group has set a strategic objective of creating expanded specialty clinics, then this capacity needs to be considered. 

Figure 2 shows the sub-specialty mix of our sample group, as compared to the MedAxiom database as a whole. In this example, the group syncs up almost identically to the mix seen across the nation. If this were not the case, the data could point us to opportunities or the need for further exploration.

Figure 2

Specialty Mix

Group

MedAxiom

Electrophysiology

15%

14%

General Non-Invasive

31%

29%

Interventional

38%

39%

Invasive

15%

17%

 

100%

99%

Taking another view on work using a different group example, Figure 3 shows output measures from the office practice, specifically new patients and return visits. Here, we see that from a new patient perspective, the group is performing above the MedAxiom median. When looking at return office visits, the group is well below the median. This could indicate that the group operates mainly in a consultative fashion and is in a market where follow-up cardiac care is provided by primary care. It could also point to an opportunity for chronic disease management through standardized specialty clinics, such as atrial fibrillation, chest pain and heart failure.  This opportunity is supported by the fact that the group has no APPs in its workforce.

This comparison also shows that the group has just one electrophysiologist out of nearly 14 physicians. When looking back at the data in Figure 2, this is about half of the ratio of the MedAxiom database overall, indicating that there is probably an opportunity for expanded EP coverage, an important consideration as part of workforce planning.

Similar data points can be run for key volumes, such as cath, PCI, echo, nuclear and EP procedures. These, too, provide insight germane to future workforce needs.

Market & Demographic Data

One of the main drivers of demand is, obviously, the population of our primary and secondary service areas (PSA & SSA, respectively). You can find lots of free data on populations and demographics through the US Census Bureau (www.census.gov). Utilizing these data provides both a historical snapshot and projections for changes in population, looking ahead five years.  Figure 4 summarizes the population data for a sample CV program, stratified by age, since demand for cardiovascular services increases in an older population.

Rarely does a program control 100 percent of its market, so population statistics alone tell just part of the story. Also necessary to consider is our program’s market penetration, as measured by market share. Most CV centers are able to get relatively accurate and current inpatient market share data. The outpatient data are more difficult to come by. If outpatient market data aren’t available, we’re forced to use the inpatient data as a surrogate. Sometimes “good enough” wins out over perfect.

Putting all the Data Together

Now that we’ve assembled our data, it’s time to put it together in a way that helps drive our workforce planning. Figure 5 takes our new patients and total cognitive encounter data and calculates provider demand, if the group expectation is median (50th percentile) performance against the MedAxiom peer group. It also compares this provider demand based both on what the group considers the FTE status and the calculated FTE status.

As mentioned in Part 2 of my blog, this calculated FTE status attempts to normalize all FTEs for comparison purposes, by presuming a median number of days worked (a measure MedAxiom collects and reports) and comparing that total to the group’s actual days worked.  Doing this normalizes post-call days, afternoons off, vacation, etc. On our sample practice, you can see in Figure 5 that our fictitious group reports nearly a full FTE more than the “calculated FTE.” This means that in terms of days worked (or conversely, days off), our sample group is below the median (works less days), which results in a lower effective FTE count.

As can be seen in Figure 5, the data show a mixed bag. Because the group is over-performing the median for new patients, the calculated demand shows a shortage of providers; however, when looking at cognitive encounters in total, the calculation suggests a surplus. The results are nearly identical, whether looking at the practice-designated FTE or the calculated FTE. Again, these data point not necessarily to an opportunity in terms of cardiologists, but more for the advent of APP utilization.

Moving to wRVUs, Figure 6 has a similar comparison to above, calculating demand based on our wRVU production. Keeping in mind that the group’s expectation is median production, the data show appropriate staffing levels. Obviously, if the group moved the target to 75th percentile expectation, this calculation would show a significant surplus.

Figure 7 shows a demand calculation based on group panel size. Panel size is defined as unique patients seen during an 18-month period, excluding test interpretations (EKG, echo, etc.), where the physician never had a face-to-face encounter. This measure, tracked by MedAxiom in its survey, is the best indication of a group’s patient population. As can be seen for our sample group, the calculation shows a workforce in pretty close sync with demand.  When considering time off (calculated FTE), the data show a potential full FTE deficit.

Unfortunately, the most recent calculation of the required need for cardiologists to population ratio is from 2003 by Thomson Healthcare (formerly Solucient). This pegs that ratio as 4.22 cardiologists per 100,000 adult population (18 years and older). In Figure 8, then, we see how the population statistics from Figure 4 provide a demand forecast based on our sample group’s primary and secondary market control. When looking simply at the population, the data would suggest a significant surplus of physicians, even looking out five years with expected population growth.

In addition to all the patient demand indicators is the simple calculation of physician roles that need to be filled. Obviously, these roles need to be carefully scrutinized by leadership and must match up with our strategic and organizational objectives, but presuming their necessity, these roles need to be considered in our workforce planning. Figure 9 shows the coverage needs (roles) for our sample group of 14 cardiologists (13.5 FTEs). Here, you can see that in order to fill all the slots, our group has just enough physician workforce to get by. At high demand times, the group will simply have to leave some roles uncovered.

Summary

With the passage of MACRA and other radical changes to Medicare reimbursement, groups cannot simply focus on volume as part of its workforce planning. Careful attention needs to be given to the new value economy and the impact this transition will have on our care team globally. The entire conversation needs to take place within the context of our organization’s overall strategic plan and the specific objectives we’ve set out to achieve. 

Given this, it is no longer sufficient to simply consider our workforce when a provider leaves or cuts back. We must create a plan that anticipates the future and our desired plan. To do this, we need both qualitative and quantitative data, looking at multiple perspectives and considering the entire care team, not just our physicians. For groups facing imminent physician slow-downs or retirements, it is wise to have a plan in place that spells out the rules for such schedule changes and the economic impact that follows. It is far better to create the plans before the request is made, so the decisions can be made without the bias of the person asking. 

Lastly, our future physician workforce will not look or act the same as it has historically. This isn’t bad or good, it just is. The new generation will have different expectations for work capacity, life balance, family responsibilities and professional job duties. This, too, needs to be part of the future plan algorithm—unless the plan is to simply turn out the lights.

 


Joel Sauer is Vice President of MedAxiom Consulting. His work includes full-service line development, co-management arrangements, workforce planning, compensation planning and integrations.

Leave a Comment

Go Back