Calculating Physician Demand Rates

Step 1: Define your Stark Service Area

When I help physicians considering a transition to concierge or direct practice business models, or to add new services (such as HRT, Stem Cells, or to add a new PA, Nurse Practitioner or additional physician or some other marketing endeavor for their practice, my first order of business is to determine their addressable market or "service area". I use standard industry definitions as set forth in compliance regulations. For example, to identify your Stark service area. Before demographics or market share can be run, an appropriate service area must be applied. There is no one-size-fits-all way to define a service area. In fact, there are several ways to do it. Which one is best? The answer is, it depends. That's where my experience and expertise from 35+ years of consulting projects in this domain are drawn. Defining a service area is both art and science.

Most physicians and hospitals define their service areas incorrectly. In many cases, they cast a wider net than they should out of ignorance. There are several ways to accomplish this calculation. As part of "knowing your customer, you must know your service area. Not who they want to serve, but the geography they actually, currently serve. This means the organization has to consider market share from the geography, it cannot be based solely on patient origin. I often do this with a combination of several reports created from the practice management system. If you are my client, you get the assignment as to which reports I will need as an initial assignment immediately after signing an consulting engagement agreement with me. I use these data to create heat maps. A heat map is a two-dimensional representation of data in which values are represented by colors. A simple heat map provides an immediate visual summary of information. More elaborate heat maps allow me and my clients to understand complex data sets.

My first service area identification approach requires the evaluation and interpretation of patient origin by ZIP codes until a specified percentage (75%-85% typically) of total patient origin is reached: 50-60% for primary service area (PSA) and 25% or so for the secondary service area (SSA). When using this method, the service area may end up looking swiss cheese on a map because of the non-contiguous ZIP codes and with holes in the area.   Heat maps that use the zip code data and can then be coupled with age, gender and socioeconomic status. When these are overlaid on top of those zip code areas, I can get at better business intelligence and arrive at an informative level of data upon which better business decisions can be informed.

By law, a hospital is required to follow the Physician Self-Referral Act (commonly referred to as the “Stark Law”) for physician recruitment. Failure to follow the Stark law could result in penalties, costing the hospital or provider millions of dollars. So I use this information and compliance formula when reviewing physician employment contracts offered by hospitals.

Step 2: Competitor analysis

This step is the most time-consuming part of the process; however, the benefits of having an accurate list of key competitors are worth the extra effort. This “detective work,” requires investigation and research to find out where competing physicians practice, what they offer, what they do differently, if they’ve left their market, are winding down, or the level at which they are active in the market. Some resources available to retrieve this data include the National Plan and Provider Enumeration System (NPPES) and CMS’s Base Provider Enrollment File. So many database companies compile this free data and then try to sell it to me and I delete and block their addresses from future contacts. The time it takes to conduct online research and make telephone calls to provider practices as a "mystery shopper" is charged at our research rates. Once my research is complete, we have a better grasp to analyze the market.

Step 3: Using Different Demand Models for different projects

When I need to make an accurate demand rate projection, the first step is to calculate a weighted average of popular demand models. There are several physician demand models available to use for my physician-per-population rate calculation. I use several models based on several factors including recency and market payer mix. For example, if a Stark service area has a high percentage of United plans, the demand model will be weighted higher to more accurately reflect the physician needs in that market. Some of the ones I use include, but are not limited to:

1. American Medical Association
2. Physician Supply Model (PSM) from HRSA (2008, but the premises are still useful)
3. Physician Requirements Model (PRM) from HRSA
4. Hicks & Glenn
5. Mature HMO (1997, and oldie but still informative)
6. Kaiser Foundation
7. GMENAC (1996) (was popular in the 1980s but still in use for certain projects)
8. Number of PAs and APRNs and their scope of practice constraints, if any, as well as retail clinics, employer worksite clinics, and other care outlets available.

See The Complexities of Physician Supply and Demand: Projections from 2015 to 2030 for some interesting data that is relevant to this part of the analysis.

Step 4: Demographically Adjust Your Demand

Once the average demand has been calculated and weighted appropriately, the next step is to demographically adjust the demands by primary care and specialties, as may be relevant to the analysis. Most specialties are adjusted based on the 65+ population in your Stark service area since the older population are higher utilizers of healthcare services. Demographically adjusting the demand is important to account for the specific needs of your population. Below are a few other specialties that we often adjust for:

1. Allergy/Immunology and Pediatrics: 0-18 years
2. Obstetrics/Gynecology: Females 15-44 years
3. Endocrinology: Females 65+ years
4. Plastic and Reconstructive Surgery: Females
5. Joint Replacement Surgeries
6. Cardiac Patients
7. Gerontology: Alzheimer's Patients

Step 5: Calculate a Potential Retirement Analysis

Another factor we take into consideration analyzes both the current demand and a five year projected demand. It is likely that physicians in your region will retire in a few years. They may also be replaced by Millennial physician graduates who work about 13% fewer hours. The physician’s birth date (provided by the hospital) or medical school graduation date (Physician Compare file) is used in the retirement analysis. Based on age or years practicing, selected by the hospital, I may retire physicians out of the future year demand models if that's relevant. Often, hospitals elect to retire physicians at 67+ years old or after 40+ years in practice.

Step 6: Of what value is this data? How should you use it?

Conducting a Situation Assessment of this kind is only the beginning. We use this for strategy planning and feasibility. My intention for all my clients is to keep them safe, compliant, and free of unnecessary distractions and avoid costly penalties, to establish an accurate count of physicians in your market and a better representation of the current and future projected market situation relevant to your project, whether it is a transition to concierge or direct pay practice, to offer new services, add additional practitioners or more staff or larger or smaller space leasing. I also use it to project opportunities arising from the number of retiring physicians in the community.