Capitation Research Paper

Patrick C. Alguire, MD, FACP
Director, Education and Career Development

Capitation payments are used by managed care organizations to control health care costs. Capitation payments control use of health care resources by putting the physician at financial risk for services provided to patients. At the same time, in order to ensure that patients do not receive suboptimal care through under-utilization of health care services, managed care organizations measure rates of resource utilization in physician practices. These reports are made available to the public as a measure of health care quality, and can be linked to financial rewards, such as bonuses.

Capitation is a fixed amount of money per patient per unit of time paid in advance to the physician for the delivery of health care services. The actual amount of money paid is determined by the ranges of services that are provided, the number of patients involved, and the period of time during which the services are provided. Capitation rates are developed using local costs and average utilization of services and therefore can vary from one region of the country to another. In many plans, a risk pool is established as a percentage of the capitation payment. Money in this risk pool is withheld from the physician until the end of the fiscal year. If the health plan does well financially, the money is paid to the physician; if the health plan does poorly, the money is kept to pay the deficit expenses.

When the primary care provider signs a capitation agreement, a list of specific services that must be provided to patients is included in the contract. The amount of the capitation will be determined in part by the number of services provided and will vary from health plan to health plan, but most capitation payment plans for primary care services include the following:

  • Preventive, diagnostic, and treatment services
  • Injections, immunizations, and medications administered in the office
  • Outpatient laboratory tests done either in the office or at a designated laboratory
  • Health education and counseling services performed in the office
  • Routine vision and hearing screening

It is not unusual for large groups or physicians involved in primary care network models to also receive an additional capitation payment for diagnostic test referrals and subspecialty care. The primary care physician will use this additional money to pay for these referrals. Obviously, this puts the primary care provider at greater financial risk if the overall cost of referrals exceeds the capitation payment, but the potential financial rewards are also greater if diagnostic referrals and subspecialty services are controlled. Alternatively, some plans pay for test and subspecialty referrals via fee-for-service arrangements but are more typically paid via contractually agreed-upon fee schedules that are discounted 10% to 30%, compared to the local usual and customary fees.

Below is an example of a capitation rate schedule. It is for illustrative purposes only and does not imply a standard for comparison purposes. The jargon used by managed care organizations for the capitation rate is PMPM (per member, per month).

Member's Age

Capitation per 
Member, per Month


Payment per Member,
per Month













> 20




Other plans may have different schedules based on patient sex, different categories of ages, and different withhold amounts.

Back to Main Menu

Model and Data Sources

This analysis constructs income statements for a practice with 6 clinicians, including 5 full-time equivalent (FTE) physicians and 1 FTE advanced practice provider (APP); FTEs may be allocated between multiple part-time providers. The outcome variable of interest is the breakeven comprehensive capitated rate inclusive of shared savings and quality incentive payments.

The income statement was developed after interviews with 3 public and private payers and 2 practice administrators from midsize practices. Multiple data sources were used to establish model assumptions, including published medical literature, the Bureau of Labor Statistics, the Centers for Disease Control and Prevention, the Medical Group Management Association, surveys from the American Academy of Pediatrics and American Academy of Family Physicians, and proprietary data from 200 independent pediatrics practices across 40 states. As a robustness check, the final model was reviewed by 2 additional practice administrators, 1 commercial payer, and a pediatric practice consultant.

For simplicity, the model imposes a number of constraints. First, it makes a direct conversion from FFS to full capitation, inclusive of quality and cost incentive payments. Second, it shifts all patients in the practice to capitated payments simultaneously. Third, the capitated rate holds the panel size constant throughout the year and includes only responsibility for basic point-of-care testing, such as rapid strep, hemoglobin, and urinalysis. Fourth, any revenues from hospital consultations or circumcisions are excluded. Finally, the model simulates a 50/50 payer mix between Medicaid and commercial payers.11,12

Expense and Revenues Under FFS and Capitation

Published panel sizes vary widely, depending on practice style and the age distribution of the panel.13⇓⇓⇓⇓18 The model assumes a panel size of 1700 and an average of 3.24 visits per patient per year.15,19,20Table 1 summarizes the core model assumptions, along with the range for each variable identified from multiple sources.

Total expenses are often reported as 60% of actual revenue in pediatrics and family medicine.14,21⇓⇓24 To increase transparency and generalizability, the model separates staff salary and fringe benefit expenses from other overhead. Pediatricians’ salaries vary with differences in practice ownership, payer mix, productivity, and geographic location.14,25,26 The model uses the Bureau of Labor Statistics national median salary of $180 000.25 Practice administrator salaries varied for similar reasons, and the model uses a salary of $92 000.27,28 The median salaries for APPs and registered nurses are consistently reported at $95 000 and $65 000, respectively.2931 Median salaries for administrative and clinical support staff vary based on duties but converge ∼$34 000.32⇓⇓35 Fringe costs as a percentage of staff salary are 15% for clinicians and the practice administrator and 30% for support staff.14,36 The model assumes a support staff to physician ratio of 3.2.14,20,37 Overhead excluding staff expenses was set at 30% of revenue, resulting in total overhead of 62% of revenue.

Overhead costs excluding staff expenses are calculated as a percentage of total revenue in the capitated model. The model increases nonstaff overhead costs from 30% to 35% to allow for additional expenses such as electronic health record upgrades and reinsurance.3840 Fewer resources may be needed for billing-related administrative functions in capitated environments, but staff must still confirm valid insurance coverage, and detailed quality reporting is required. Because total expenses are calculated as a percentage of payments collected (not receivable) in the FFS model, the total expenses account for a similar proportion of revenues in the capitated model (63%).

Vaccines are a large expense for practices. Given the Centers for Disease Control and Prevention vaccine schedule for children 0 to 18 years of age and current vaccine prices, average vaccination cost is ∼$135 per patient per year.41,42 Practices typically break even or gain small profits from vaccinations billed to commercial payers, whereas the opposite is true for those billed to Medicaid.43 Because the modeled practice is a 50/50 payer mix, vaccines are excluded from both the FFS and capitation models.

Copayments may contribute meaningfully to practice revenues in both FFS and capitation. The model incorporates an average patient copayment of $8. This amount is calculated by multiplying the average commercial copays ($23) by the proportion of commercial patients (50%) and the likelihood that commercial plan requires copayments (66%).44 Many states allow (or will soon allow) nominal copayments for Medicaid beneficiaries, but these have been excluded from the model.45,46

In the FFS model, practice revenues are tied to physician and APP encounters. Practice revenues are calculated by multiplying the average number of visits per day, the number of providers, the number of clinical days per year, and the average payment per encounter.47 The hypothetical practice assumes 25 patients per day per provider, 220 clinical days annually, and $100.00 average payment per visit.14,4850 APPs may independently bill at 85% of physician fees, although APP roles may vary by practice.51 The FFS model accounts for rejected claims, no-shows, and uncompensated visits by writing off 10% of expected revenue. The model includes a 5% practice margin in both the FFS and capitated scenarios so that the practice is able to build and maintain financial reserves for upgrades or unexpected expenses.

In the capitated model, practice revenue is driven by attributed panel size and the average capitated payment. Instead of billing payers for individual patient encounters, capitated practices receive a risk adjusted base PMPM payment for each attributed patient. Particularly in pediatrics, the capitated rate should account for the age of covered patients.52 The model excludes vaccination, but if included in capitated payments, rates must be substantially higher and must allow for vaccine price increases, which occur annually if not more frequently.

Providers typically receive incentives for performance relative to quality and cost benchmarks. Quality and cost bonuses are paid when practices reach predetermined performance thresholds and as a percentage of spending compared with targets, respectively. For transparency, our model converts quality and shared savings payments into PMPMs. Incentive payments and copayments are added to the base PMPM to calculate total revenue.

In capitation: Net Income = Patient Co-payments + Capitation Base Rate + Utilization Incentives + Quality − Operating Expenses.

Sensitivity Analysis

Practices vary widely in organization and style. Our model presents an “average” practice, but several core assumptions may vary significantly between practices, including physician salary, panel size, and overhead. As a robustness check, an economic simulation was constructed where physician salary, panel size, and overhead less staff expenses were allowed to vary across the range of values in Table 1. Each simulation generated 500 different practices. The model was iterated 50 times to generate a total of 25 000 practices.

Practice Transformation

Because revenues are not tied to face-to-face physician or APP encounters, capitation models allow providers greater clinical autonomy than FFS, and all practice staff may contribute to patient care at the level to which they are legally entitled.13,37,53⇓⇓56 We modified the practice’s staffing ratios to reflect 2 published PCMH transformations to assess its effect on the PMPM.37,54 In 1 example, an additional APP mental health provider, 2.5 nurses, 1.5 clinical support staff, 0.5 administrative support staff, and 0.5 of a practice administrator were added. These additions increased the staffing ratio by 37% (3.2 to 4.4). In a second example, 2 nurses and 2 clinical support staff were added. These additions increase the staffing ratio by 25% (3.2 to 4.0).

Capitated Rates for a Real-World Payer

For illustrative purposes, we obtained capitated rates and program information for Capital District Physicians Health Plan (CDPHP), a health plan in upstate New York that serves nearly a half million commercial and Medicaid members.57 CDPHP provided data from 2013, including the base capitated rate, quality incentives, and shared savings incentives payments. We applied Medicaid, commercial, and 50:50 blended rates to our model with and without staffing changes, to assess the financial impact on our hypothetical practice.


List of Core Model Assumptions and Practice Parameters

0 thoughts on “Capitation Research Paper”


Leave a Comment

Your email address will not be published. Required fields are marked *