Introduction

This concept describes a method for calculating primary care provider panel size and identifies and provides access to the discussion and results from MCHP research on this subject. The information in this concept comes directly from the deliverable, A Comparison of Models of Primary Care Delivery in Winnipeg, by Katz et al. (2016).

Panel Size - Methods

Panel size refers to the number of patients that receive the majority of their primary care from one provider. However, comparing the panel sizes among the primary care providers is complicated by the fact that providers do not all see patients for the same number of hours per day.

In order to meaningfully compare panel sizes, billing claims for both fee-for-service (FFS) and alternative-funded (see alternate payment plan (APP)) primary care provider services need to be captured equally in the Medical Services data. Using the methods described in the Shadow Billing Data Validation concept, Katz et al. (2016) found that the Medical Services data captures billing claims for FFS and alternative-funded primary care providers equally.

This allowed them to use the Full-Time Equivalent (FTE) National Algorithm developed by Health Canada (see Watson et al. (2004) for more information), to assign an FTE to each provider, regardless of the provider funding method. In this algorithm, every primary care provider's FTE is calculated according to one of the following scenarios:


• if the primary care provider's annual total billings fall between the 40th and 60th percentile (inclusive) of all providers' annual billings, the FTE is assigned a value of 1; this defines the typical full-time primary care provider; or
  
• if the primary care provider's annual total billings fall below the 40th percentile, the FTE is the proportion of total billings divided by the 40th percentile; the FTE ranges from 0 to 1 (exclusive); or
  
• if the primary care provider's annual total billings are above the 60th percentile, the FTE is equal to 1 plus the natural logarithm of the proportion of total billings divided by the 60th percentile; the logarithmic relation prevents high income providers from having a very large FTE: for instance, a provider earning four times the 60th percentile will have an FTE of 2.4 instead of an FTE of 4 (Watson et al., 2004).

For a more detailed description of the FTE calculation, please see the Full Time Equivalent (FTE) Physicians - Calculations concept.

The FTE for each primary care provider was then used in the calculation of each provider's panel size. In Katz et al. (2016), primary care provider panel size was calculated by dividing the total number of patients allocated to a primary care provider by that provider’s FTE. Because the panel size calculation took each provider's FTE into account, the panel sizes of all providers are comparable, regardless of the number of hours per day spent seeing patients. Note: Only primary care providers who reported at least one claim in each of the four quarters of the first year they appear in the research study period were included in the FTE calculation.

Statistical modeling was then performed to determine whether there is a relationship between panel size and the characteristics of primary care providers and patients. For more information on the statistical modeling techniques used in this research, please read the Factor Analysis concept.

For more detailed information on the method used to calculate primary care provider panel size, please see the section titled Primary Care Provider Panel Size in the deliverable.

Panel Size - Research Findings and Results

In Katz et al. (2016), they investigated primary care provider panel size from two perspectives. The first looked at the relationship between patient and provider characteristics and panel size. There are many characteristics that can influence the optimal panel size for any one primary care provider, such as provider sex, provider age, years of practice, percent of patients aged 75 or older, present of patients who are children of teen moms. These characteristics and others, including those that are part of social complexities concept were investigated.

The second perspective looked at the relationship between the different models of primary care delivery and panel size, to determine whether there are differences between fee-for-service (FFS) and alternative-funded providers.

In summary, some of the findings from this research include:


• patients with five or more social complexities have 20% more visits to their primary care provider;
  
• one primary care provider characteristic - male sex - increased panel size by 210 patients;
  
• the next characteristic with the largest impact on panel size was the percent of visits from non-allocated patients, which decreased panel size by 10 patients; and
  
• FFS providers have panel sizes nearly double those of alternative-funded providers.

For more information on the discussion and results related to primary care provider panel size available from this research, please see:


• Table 8.1 – Association Between Patient and Provider Characteristics and Panel Size Adjusted Values - the bold variables in the table with a positive estimate value are associated with an increase in panel size and those with a negative value are associated with a decrease in panel size.
  
• Table 8.2 – Association Between Patient and Provider Characteristics and Panel Size Regression Model Results – this includes components of the Social Complexity Index . Again, only those bold variables in the table have a statistically significant influence on panel size.
  
• Table 8.3 – Clinical Model Panel Size – shows the actual and expected panel sizes based on the statistical models that were developed.