In this chapter, we summarize the purpose of our project, provide an overview of how physician work is valued under the Medicare physician fee schedule, describe our project objectives and how the results might be used, and explain the organization of the remainder of the report.

The Centers for Medicare & Medicaid Services (CMS) uses the resource-based relative value scale (RBRVS) to pay physicians and nonphysician practitioners for their professional services. Under RBRVS, payment for a specific service is broken into three elements: physician work, practice expense, and malpractice expense. Each component is valued separately in relative value units (RVUs). Total...

In this chapter we provide a descriptive analysis of how time and work are distributed for the core study procedures in current CMS estimates across each component of the BBM: intra-service work, pre- and immediate post-service work, and, for surgical services with global periods, related post-operative E&M visits. This chapter begins with a description of the CMS estimates and the BBM. We then review a number of descriptive analyses. These analyses set the stage for understanding the relative importance of the individual work components in the current system and the analyses that we undertake in the remainder of the report...

We begin this chapter with a description of the core procedures that are included in our models. We then provide an overview of our modeling approach and how we explore different modeling issues. We conclude by summarizing the steps that we use to predict RVUs for total work and the work components. The details on the prediction models for each work component are provided in the relevant chapters of the report.

As an illustration of our models, we focus on the CPT category of surgical procedures. This broad category includes surgery (where there is an incision of some kind) and...

One critical aspect of all our validation models is intra-service time. In this chapter, we focus on the intra-service times. We begin by first discussing complexities in defining intra-service time using data from external databases. Next, we introduce an analytic technique that combines existing CMS estimates and time estimates from external databases. This analytic technique helps us address situations where for a given procedure there are few observations in the external data sources. Using this technique, we use data from two data sources (SPARCS and Medicare anesthesia claims) to estimate intra-service time, what we call RAND time. Finally, we externally...

Under the RBRVS, physicians’ activities immediately before or after intra-service (“skin-to-skin”) effort are bundled with payment for the procedure itself. In the case of surgical procedures, the four pre-service and immediate post-service activities include (1) patient evaluation, (2) positioning, and (3) scrubbing, as well as the (4) immediate post-operative care. This chapter describes these four pre-service and immediate post-service work components and develops models that can be used to identify procedures with more or less pre-service or immediate post-service work than is expected based on observable characteristics of the procedure, including RAND intra-service time. The time values for the four...

This chapter describes the post-operative E&M work component and its valuation using the reverse BBM. We describe corrections that we make in the work values derived from the CMS estimates and how this work component can be predicted for both the typical POS and all POS using characteristics of services from the CMS estimates, the revised time estimates, and Medicare administrative data.

Post-operative E&M visits related to surgical procedures are bundled into total work for CPT procedure codes with a 10- or 90-day global period. These E&M visits are not reimbursed separately under RBRVS when they are performed by the...

This chapter describes the different models of intra-service work and intra-service intensity (IWPUT). First, we give background on the distinction between intra-service work and IWPUT and different ways they can be generated. Then, we provide the rationale for our predictive models. We also discuss why certain procedure characteristics may be correlated with intra-service work and intensity and show the actual correlations. We compare the intra-service work and IWPUT estimates derived from the CMS time file to our estimates.²⁷Based on these findings, we discuss the rationale for three different models to predict intra-service work and compare the RAND model results...

In prior chapters, we explored various options for estimating the individual work components using the BBM. In this chapter, we focus on total work RVUs. For Models 1 and 2, we combine the results from the earlier chapters to estimate total work RVUs using the BBM. In Model 3, we use a single prediction model to predict total work RVUs using the same type of variables as we did for intra-service work. We start with an overview of the models and how they vary. We then provide the results and summary statistics that can be used to compare the models....

In this chapter, we discuss selected topics that are not explicitly addressed in our models but that could potentially affect how the work RVUs are estimated and/or implemented in the Medicare physician fee schedule. The RAND time estimates are based on circumstances when only a single procedure was performed; in this chapter, we first discuss our estimates of the incremental time associated with performing multiple procedures or add-on procedures to a base code. Next in this chapter we explore whether the small percentage of Medicare beneficiaries who have a core procedure as an outpatient and are subsequently admitted as an...

In this chapter, we first review the key findings from the RAND validation models. We then discuss how the RAND models could be used by CMS. We provide examples of how the values could be used in the valuation process and describe the input we received from four specialty panels on procedures. Key issues we need to address are how to account for CPT coding and practice pattern changes, how to maintain relativity with other codes that were not included in our analyses, and whether our approach could be applied to other types of codes. Finally, we end with some...