How to Use RVUs in Physician Pay

Physician Compensation Report, March 29, 2000

The use of relative value units (RVUs) in physician pay systems is rapidly growing, says Theresa Raczak of Med Comp in Chicago, because it leaves out the "financial noise" of payer discounts and bad debts, and focuses on the work physicians do.

Raczak and Seth Garber, M.D., of William M. Mercer Inc. in Seattle, gave a one-day MGMA-sponsored course on designing physician compensation plans in Denver March 10, and will present the course in several cities later this year. Among the many topics they covered were measuring RVUs and using them to figure productivity pay They were joined by Laurel Weinstein, MGMA project manager for its Physician Compensation and Production Survey, who discussed salary trends the last few years and MGMA tools to determine pay ranges for particular physician posts from its survey data.

The 1999 Survey says that 11.2% of the practices that responded (across all specialties) used RVUs in their compensation methodology. These were relatively large practices because they employed 23.9% of the providers. These figures understate the penetration of RVUs among groups using some form of productivity pay, because MGMA data indicate that only about 55% of providers covered by the Survey are paid 50% or more on productivity

How to Count RVUs

RVUs are part of HCFA's Resource Based Relative Value System (RBRVS), which recognizes three components of value in medical procedures: the professional or physician work component, the practice expense component, and the malpractice insurance cost component. A Level 3 office visit (CPT code 99213) is worth .67 work RVUs and 1.41 RBRVS units; the difference is practice and malpractice expenses.

HCFA uses RBRVS to help set Medicare reimbursements for most Part B services. HCFA further modifies pay rates by Geographic Practice Cost Indexes, which can vary by more than 100% between localities.

In a May 1998 MGMA "information exchange" about RVUs, 15 groups said they used RBRVS units to determine production, while 12 groups said they used work RVUs. But it is widely believed that the strong trend is toward work RVUs. The latter are unaffected by regional, payer and cost factors, and depend on physician (and midlevel provider) work. Garber adds that if a group uses encounters or visits to measure productivity RVUs solve the problem of comparing them.

McGraw-Hill developed an RVU system that was widely used several years ago, but is less so today perhaps because HCFA's system must be used for Medicare. There is another RVU system used in California. Raczak says that in adding work RVUs for each physician, a group should:

* Include RVUs from all payers.

* Use HCFA coding modifiers.

* Calculate RVUs for procedures HCFA does not recognize, using gross charges. For instance, if the group's charge for an "executive physical" is $120, and for a Level 4 visit is $80 (a ratio of 1.5:1), then multiply the RVUs for a Level 4 visit (1.1) by 1.5, to get 1.65 RVUs for an executive physical.

* Exclude the technical (non-physician) component of laboratory, radiology, diagnostic and surgical procedures.

* Exclude RVUs attributed to physician extenders such as physician assistants and nurses.

Tips for Osing Benchmarks

Part of validating any physician pay system is to compare it to survey benchmarks. Raczak has this advice on benchmarking RVUs:

(1) Use two sources of benchmarks and average them, not only for RVUs but for all benchmarking.

(2) The RVU per encounter ratio should be calculated. Typical ratios for primary care physicians are in the .7-1.3 range, and much higher for many specialists.

(3) A low RVU/encounter ratio could mean the physician is undercoding; has very healthy patients; is not doing all indicated procedures; or is simply efficient; or it could reflect some office-wide problem, like failure to code in some instances. A high ratio could have the opposite interpretations. Either way, the real reasons should be determined.

(4) When implementing an RVU-based pay system, there is almost always "RVU creep" upward in the RVU/encounter ratio. This occurs even if the previous system was collections, Raczak says. A member of the Denver audience said that when his Florida practice went to RVUs, the RVUs per visit rose from .7 to 1.22, and that payer reimbursements did not rise accordingly As a result, the pay per RVU had to be lower than was modeled before the RVU system was implemented.

(5) Remember that benchmarks are comparators, not best practice standards.

Building a Pay System

Most groups set pay rates at pre-determined levels per RVU. Raczak and Garber emphasize that groups must be able to afford the rates they choose. Raczak gave an actual example of a group that was collecting $63 per RVU and paying physicians an average of $50 per RVU. The $13 difference was woefully inadequate to meet expenses. The solution was to cut the pay rate to $40 per RVU, resulting in pay cuts for four out of five physicians, and a pay hike for the other physician.

Part of setting up any productivity pay system, whether or not measured by RVUs, is to model against current performance. Wide variances between physicians are often found in compensation/RVU and compensation/ collections ratios. When a new plan is implemented, these variances are generally ironed out.