The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 143
Appendix A-1 Technical Approach to Payment Simulations: IOM Committee Recommendations for Hospital Wage Index and Physician Geographic Adjustment Factors Prepared by: Kathleen Dalton, Ph.D. RTI International Contents 1 Committee Changes Included in the Payment Simulations, 145 2 Technical Approach, 146 2.1 Use of BLS Data, 146 2.1.1 New BLS-Constructed Indexes, 146 2.1.2 Mapping BLS Areas to CBSA Markets, 146 2.1.3 Adjusted Average Relative Wages vs. a Fixed-Weight Index, 147 2.2 Addition of Benefits Index, 148 2.3 Redefined GPCI Payment Areas, 150 2.4 Smoothing, 150 2.5 County Indicators for Health Professional Shortage Areas, 152 2.5.1 Background, 152 2.5.2 Computation, 155 2.6 Payment Simulations, 155 2.6.1 Hospital Computations, 155 2.6.2 Physician Computations, 159 2.6.3 Payment Impact Computations, 160 2.7 Budget Neutrality, 161 3 Payment Impact Exhibits, 163 3.1 IPPS Hospital Estimates, 163 3.2 Part B Physician and Other Practitioner Estimates, 163 List of Exhibits A-1 Recommended Changes in Index Construction Incorporated into Payment Simulations, 145 A-2 Labor Markets in Source Data and Final Index Construction, 147 143

OCR for page 143
144 GEOGRAPHIC ADJUSTMENT IN MEDICARE PAYMENT A-3 Percent difference in HWI values due to data change alone, plotted against number of hospitals in labor market, 148 A-4 Effects of Adjusting for Independent Area Variation in Benefits, 149 A-5 County Assignments by Region, Type of Payment Locality, and CBSA Market, 151 A-6 County Smoothing Adjustments, by Type of Index, 152 A-7 Commuter-based smoothing adjustments by RuralUrban Continuum Code, 153 A-8 IOM Committee's Recommended Smoothing Adjustments Compared to Current Outmigration Adjustments Under "Section 505" and Related Reclassifications, 154 A-9 Correlation of Adjusted ZIP Codeto-County Address Counts to Population and Beneficiary Statistics, 156 A-10 Distribution of estimated proportion of county population in primary care shortage areas, 156 A-11 Distribution of Counties, Part B Enrollees, and RVUs Billed by Primary Care Practitioners, by Revised HPSA County Status, 157 A-12 Budget Neutrality Factors Imposed on IOM Committee Indexes, 162 A-13 Distribution of payment impact across all IPPS hospitals, 163 A-14 Estimated Change in IPPS Payments, Isolated by Type of IOM Committee Recommendation, 164 A-15 Impact of IOM Committee Recommendations on IPPS Payment, by USDA Rural Urban Continuum Code, 165 A-16 Impact of IOM Committee Recommendations on IPPS Payment, by Census Division and Metropolitan Status, 165 A-17 Impact of IOM Committee Recommendations on IPPS Payment, by Hospital Reclassification Status, 166 A-18 Impact of IOM Committee Recommendations on IPPS Payment, by Special Rural Status, 166 A-19 Impact of IOM Committee Recommendations on IPPS Payment, by Teaching and DSH Status, 166 A-20 Impact of IOM Committee Recommendations on IPPS Payment, by Bed Size, 167 A-21 Distribution of physician payment impact across all counties, 167 A-22 Change in Aggregate Geographic Adjustment Factor, by Type of IOM Committee Recommendation, 168 A-23 County Analysis of the Isolated Payment Effects from Redefining the GPCI Payment Areas, 169 A-24 Physician Payment Impact of IOM Committee Recommendations, by USDA Rural Urban Continuum Code, 170 A-25 Physician Payment Impact of IOM Committee Recommendations, by Revised Health Professional Shortage Area Indicator, 170 A-26 Physician Payment Impact of IOM Committee Recommendations, by County Ranking in Median Family Income, 171 A-27 Physician Payment Impact of IOM Committee Recommendations, by County Ranking in Percent Non-White Population, 171 A-28 Physician Payment Impact of IOM Committee Recommendations, by Rural County Population Density, 171

OCR for page 143
APPENDIX A 145 A-29Physician Payment Impact of IOM Committee Recommendations, by Percent of Total RVUs Accounted for by Primary Care Practitioners, 172 A-30Effect of Level of Physician Work GPCI on Estimated IOM Committee Payment Differences, for Counties Grouped by Revised Health Professional Shortage Area Indicators, 172 1 COMMITTEE CHANGES INCLUDED IN THE PAYMENT SIMULATIONS Recommendations from the committee's Phase I report that have been incorporated into payment simulations are presented in Exhibit A-1 below, grouped as changes in source data, changes in labor markets used for payment areas, and changes in exceptions or adjustments. Not all recommendations could be simulated accurately. For example, the recommendation for new data on geographic variation in commercial rents (a component of the practice expense [PE] geographic practice cost index [GPCI]) could not be included in the simulations because we do not have the recommended data and could not identify a reasonable proxy. The recom- EXHIBIT A-1 Recommended Changes in Index Construction Incorporated into Payment Simulations Regarding the Geographic Practice Regarding the Hospital Wage Type of Recommendation Cost Indexes Index Changes in Data U se health care worker wages instead R eplace hospital reported [Year 1 Recommendations 2-2 of all-employer wages average wages with BLS- and 3-3] Use BLS-constructed indexes from based index health care (Note: recommendations 5-4 and public and nonpublic data for all worker wages for hospital 5-7 were incorporated by CMS occupations reported for physician occupations into 2012 rates and therefore did offices surveys Incorporate separate benefits not need to be simulated) Incorporate separate benefits index index (from cost reports) (from cost reports) Changes in Payment Areas R eplace the 89 payment localities A pply county-based (Market Definitions) (88 excluding territories not included smoothing based on [Year 1 Recommendations 2-1 in this analysis) with CBSA-based commuting patterns across and 4-1] markets for metropolitan counties markets and single rest-of-state areas for nonmetropolitan counties Apply county-based smoothing based on commuting patterns across markets (practice expense and physician work GPCIs only) Changes in Exceptions and A pply county-based smoothing E liminate frontier floors Adjustments based on commuting patterns across (selective replacement with [Year 1 Recommendation 4-2] markets (practice expense and other types of payment physician work GPCIs only) adjustments where needed) Eliminate frontier floors and work Eliminate rural floors for GPCI floors from the index metropolitan areas Eliminate geographic reclassifications (all) NOTES: BLS = Bureau of Labor Statistics; CMS = Centers for Medicare & Medicaid Services; GPCI = geographic practice cost index.

OCR for page 143
146 GEOGRAPHIC ADJUSTMENT IN MEDICARE PAYMENT mendation to incorporate geographic variation in health care worker benefits into the Bureau of Labor Statistics (BLS)-based index also cannot be implemented as envisioned without more detailed data collection; in this case, however, market-level variation in hospital worker benefits is available through Medicare cost reports, and is used as reasonable proxy. 2 TECHNICAL APPROACH 2.1 Use of BLS Data 2.1.1 New BLS-Constructed Indexes BLS base wage indexes for the hospital wage index (HWI) and the nonphysician wage component of the PE-GPCI were computed by BLS staff at RTI's request, in order to make use of data in small markets that were suppressed from the public use files due to privacy rules. Note that this is different from current policy, in which the Centers for Medicare & Medicaid Services (CMS) computes the GPCIs directly from wages in published data. B oth GPCI and HWI used wages reported across all health care employers, defined as North American Industry Classification System (NAICS) code 62. Indexes were con- structed from the mean wage statistic. Note that this is different from current policy, in which CMS uses the published median wage statistic.1 BLS computed index values using fixed employment weights for physician offices (NAICS code 621100) in the nonphysician wage component of the PE-GPCI, and for general hospitals (NAICS code 622100) in the hospital index. For both indexes, weights for all occupations that were reported in their respective NAICS group were used in the computations. Missing values for any given occupation within any given BLS area were handled by renormalizing the weights for nonmissing occupations within the affected market such that the nonmissing weights for that market would sum to 1.00. Note that this is different from current policy, where CMS replaces missing data with the national median wage. 2.1.2 Mapping BLS Areas to CBSA Markets In most parts of the country, the BLS survey data are analyzed by geographic areas that correspond to metropolitan core-based statistical areas (CBSAs) plus multiple nonmetropolitan areas within each state composed of nonmetropolitan counties grouped at the recommenda- tions of that state. The exception is in New England, where BLS data are analyzed by New England City and Town Areas (NECTAs). Unlike CBSAs, which are composed of whole counties, NECTAs can cross multiple counties.2 This causes problems in mapping BLS data to individual counties, both for CMS when it computes current GPCIs and for the IOM's recommended indexes. For consistency with current CMS practice, RTI used a mapping provided to us by the CMS contractor for physician payments (Acumen LLC) to assign BLS wage values to individual counties. 1 Mean wages were used after it was noted that the median wages by market were more frequently suppressed in the publicly available data series; if the indexes are computed by BLS from nonpublic data, however, then median wages would be preferable in order to avoid distortion from occasional extreme values that might reflect data reporting errors. 2 See http://www.census.gov/geo/www/2010census/gtc/gtc_cbsa.html for further definitions and discussion.

OCR for page 143
APPENDIX A 147 Twenty-nine counties are affected by this problem; 15 map to two NECTAs, seven map to anywhere from three to nine NECTAs, and seven map to 10 or more NECTAs. Where this occurs, an employment-weighted average of the relative wages for all NECTAs associated with a given county was computed and assigned to that county. Counties were then remapped to their payment locality, and the relative wage for the locality was computed as an RVU-weighted average of county relative wages. We note, however, that RTI's averaging may not be identical to the averages computed for CMS, and that some of the payment differences we have identified in New England areas may be due to this. It is worth emphasizing this is an area where further review might be helpful; the averaging method is a convenient, but not necessarily optimal, way to handle the problem. Exhibit A-2 provides the number of geographic areas used for index construction in the original BLS data and the number for the final recomputed indexes for this report. GPCIs and simulations of physician payments included data from Puerto Rico but did not include data from other territories. The recomputed HWI and simulations of hospital payments did not include data from Puerto Rico, because the Inpatient Prospective Payment System (IPPS) base rates and wage index are handled somewhat differently in this territory as compared to the 50 states. 2.1.3 Adjusted Average Relative Wages vs. a Fixed-Weight Index The most significant data change recommended by the Institute of Medicine (IOM) is the move in the HWI from a relative average hourly wage to a fixed-weight index. Addition of the benefits index to the BLS-based wage index (see Section 2.2, below) makes the BLS data more comparable to the IPPS average hourly wage data, but major differences remain due to (a) sub- stituting BLS's average wages from all health care employers for wages paid by IPPS hospitals, and (b) substituting a fixed-weight construction with a national average hospital occupation mix as weights, for average hourly wages reflecting each individual hospital's occupation mix with only a partial adjustment to standardize for national average mix of nursing personnel. RTI found that the percent differences between the CMS occupation-mix adjusted hospital index and benefits-adjusted BLS-based index are relatively small in large markets where several hos- pitals contribute to the CMS hospital index, but the differences become quite large (whether EXHIBIT A-2 Labor Markets in Source Data and Final Index Construction BLS-Based Wage IOM Proposed GPCI IOM and CMS HWI Areas Labor Markets Labor Markets Nonmetropolitan 60 48 48 Metropolitan, CBSA-based 368 384 384 Metropolitan, NECTA-based 29 0 0 Subtotal Excluding Territories 457 432 432 Puerto Rico, Metropolitan* 8 8 0 Puerto Rico, Nonmetropolitan* 1 1 0 Total 466 441 432 NOTES: BLS = Bureau of Labor Statistics; CBSA = core-based statistical area; CMS = Centers for Medicare & Medicaid Services; GPCI = geographic practice cost index; HWI = hospital wage index; IOM = Institute of Medicine; NECTA = New England City and Town Area. *Payment areas in Puerto Rico are included in the HWI but are adjusted separately due to special exceptions in the com- putation of the standardized rates for this area. All HWI analyses for the IOM Committee exclude these areas. Payment areas in other territories are excluded from both HWI and GPCI analyses. SOURCES: RTI Analysis of CMS Wage Index Files; Communication from Acumen LLC, received November 6, 2011.

OCR for page 143
148 GEOGRAPHIC ADJUSTMENT IN MEDICARE PAYMENT Percent Difference in Indexes: IOM (Pre-Smoothed) vs. CMS Base Occ-Mix Adjusted 20 10 % Difference 0 10 20 0 20 40 60 80 Number IPPS Hospitals in Market EXHIBIT A-3 Percent difference in HWIExhibit values due to data change alone, plotted against A-3.eps number of hospitals in labor market. NOTE: Both indexes adjusted for budget neutrality. SOURCE: RTI simulations. positive or negative) as the number of hospitals contributing to the CMS hospital index declines. This is easily illustrated in a scatter plot of the percent difference against the number of IPPS hospitals per market (Exhibit A-3). The shape of this plot suggests strongly that hospital wages could be a reasonable proxy for health care wages but only in markets where there is an adequate sample of hospitals; where there are too few hospitals, the IPPS average hourly wage, even after the partial occupation mix adjustment, is too heavily influenced by the hiring patterns of the specific hospitals in that market. The smaller the market, the less accurate the IPPS hospital index is as a measure of local variation of the exogenous price of health care labor. This is a particularly important finding in light of CMS's use of the IPPS hospital index as geographic price adjusters for other institutional settings, and one that lends strong support to the BLS data recommendation. 2.2 Addition of Benefits Index For lack of better data at this time, the source for the independent benefits index applied to both the HWI and the work and practice expense GPCIs is the IPPS hospital cost report wage survey.3 Data from 2009, 2010, and 2011 were combined to provide additional stability to the index. Compensation-related benefits (including payroll taxes, insurance, and pension costs) 3 Worksheet S-3 Parts 2 and 3, as edited and adjusted for inflation by CMS, and published in the wage index public use files. File can be found at http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/ FY-2012-IPPS-Final-Rule-Home-Page.html.

OCR for page 143
APPENDIX A 149 are identified separately on these surveys, and can be used to compute an aggregate market- level average benefit cost per paid hour. This series was then converted to a national index by dividing the market-level hourly benefits figures by the national average hourly benefits figure. Budget neutrality between the base wage and benefits indexes was implemented by normal- izing each to a payment-weighted average of 1.00. For each market, the two index values were then combined using the national weights for the ratio of benefits (exclusive of paid time off) to base wages, as published in the IPPS market basket and the MCI, respectively. Exhibit A-4 shows the effects of adjusting base BLS wages for independent variation in benefits. It summarizes the distribution and regional mean values for the base wage index, the benefits index, and the resulting total compensation index, as computed for the revised HWI and for the new CBSA-based practice expense GPCI. Accounting for variation in benefits tends to raise index values in high-wage markets and lower them in low-wage markets, widening the gap slightly between the lowest and highest index values. Puerto Rico is the only area where addition of the benefits resulted in a substantial reduction in the PE-GPCI, although a review of the data indicated that this could be due to EXHIBIT A-4 Effects of Adjusting for Independent Area Variation in Benefits Hospital Wage Index Practice Expense GPCI Budget- Neutral BLS-Based BLS- Resulting Wage Resulting Based Budget- Budget- Component Budget- Budget- Hospital Neutral Neutral Index Neutral Neutral Wage Benefits Compensation (IOM Benefits Compensation Index Index Index Version) Index Index N (Markets) 431 431 431 441 441 441 Mean (Unweighted) 0.955 0.963 0.959 0.944 0.954 0.946 Standard Deviation .11025 .2222 .1231 .1180 .2411 .1340 Index Values Minimum 0.746 0.550 0.728 0.462 0.227 0.430 5th percentile 0.821 0.696 0.809 0.815 0.676 0.793 25th percentile 0.876 0.814 0.875 0.877 0.811 0.872 50th percentile 0.931 0.933 0.931 0.940 0.933 0.932 75th percentile 1.007 1.049 1.021 1.007 1.050 1.018 95th percentile 1.160 1.392 1.177 1.152 1.382 1.156 Maximum 1.487 1.973 1.591 1.373 1.984 1.501 Average Values, by Region and Rural/Urban Status Northeastmetro. 1.107 1.149 1.121 1.109 1.188 1.127 Nonmetro. 0.921 0.955 0.932 0.92 0.959 0.929 Midwestmetro. 0.975 0.997 0.985 0.982 0.996 0.986 Nonmetro. 0.87 0.911 0.883 0.867 0.916 0.879 Southmetro. 0.962 0.846 0.942 0.956 0.853 0.936 Nonmetro. 0.865 0.759 0.847 0.848 0.767 0.832 Westmetro. 1.134 1.176 1.148 1.088 1.181 1.109 Nonmetro. 0.98 1.026 0.994 0.947 1.019 0.963 Puerto Ricometro. -- -- -- 0.567 0.339 0.521 Nonmetro. -- -- -- 0.873 0.858 0.871 NOTES: BLS = Bureau of Labor Statistics; GPCI = geographic practice cost index; IOM = Institute of Medicine. SOURCES: RTI Simulations; Wage Index Public Use Files published for FY 2010 through FY 2012.

OCR for page 143
150 GEOGRAPHIC ADJUSTMENT IN MEDICARE PAYMENT underreporting of benefits on the Medicare Hospital Cost Report. Nevertheless there are still significant regional differences in relative benefit levels, indicating that it is important to incor- porate benefits into wage index. 2.3 Redefined GPCI Payment Areas Payment areas for the GPCIs were reconfigured as CBSA markets by reaggregating county- level data to CBSA and statewide nonmetropolitan areas, using total county relative value unit (RVU) to weight each county index within the revised market (see Section 2.5.2). Redefining the payment areas into separate metropolitan and nonmetropolitan markets has a systemati- cally negative effect in the index values for nearly all rural counties, but it also has a surprisingly large impact in many metropolitan areas. This is due to the fact that for the 34 current pay- ment localities that are not statewide, the division into urban and rest-of-state areas does not always conform to metropolitan and nonmetropolitan CBSA designations, and consequently the effect of regrouping counties based on CBSA metropolitan areas is less predictable. Among metropolitan counties, converting to CBSA markets reduces the GPCIs for roughly half and increases them for roughly half. In contrast, converting to CBSA markets reduces the GPCIs for 99 percent of nonmetropolitan counties. Exhibit A-5 provides additional detail on the county- level impact of redefining the GPCI payment areas, broken down by region and by the type of current payment locality. 2.4Smoothing The approach recommended by the committee for commuter-based smoothing adjust- ments is described in detail in the Phase I report, where it was illustrated using the 2000 "long form" census data that is used by CMS to implement the "Section 505" outmigration adjust- ments. For these simulations, the IOM obtained special tabulations of data from the most recent 5-year summary "Journey to Work" section of the American Community Survey (ACS).4 Commuter-pattern based smoothing is predicated on the assumption that economic inte- gration across CBSAs or other county-based markets can represent inaccuracies in the labor markets as defined. This is seen most clearly when the wage indexes of adjoining markets are substantially different and employers compete for workers across the county-drawn boundaries. To reduce the number of arbitrary "cliffs" in the wage index--where index values differ sharply at market boundaries but economic integration (as demonstrated by the commuting) is evident at the geographic edges of these markets, the committee recommended computing county- level adjustments based on commuter-weighted averages of the index values in neighboring markets. Where workers commute in or out of counties that are part of the same labor market, no change in the index occurs; where workers commute in or out of markets with relatively little difference in their wage indexes, only small changes result. Where workers commute in or out of markets with large differences (the "cliffs"), large changes occur. Chapter 3 of the IOM's Phase I report provides detailed examples of how these adjustments are calculated for the HWI. 4 Discussion of this survey can be found in Chapter 5 of the first report, Geographic Adjustment in Medicare Payment: Phase I: Improving Accuracy (Washington, DC: The National Academies Press). Unlike the 2000 census "long form" data, the ACS data are from community samples. Complete national county-level data are available only from the 5-year summary files due to sample size issues. Special tabulations were provided to the IOM that were run for the county commuting patterns of all health care workers.

OCR for page 143
APPENDIX A 151 EXHIBIT A-5 County Assignments by Region, Type of Payment Locality, and CBSA Market Statewide Payment All Payment Localities Non-Statewide Payment Localities Localities Urban Rest-of-State Region and Type Number Percent Number Percent Number Percent Number Percent of CBSA-Based of Total of Total of Total of Total Market Counties RVUs Counties RVUs Counties RVUs Counties RVUs Northeast Metropolitan 17 5.6% 34 31.4% 72 19.5% 123 18.6 markets State nonmetro. 20 0.9% 4 0.2% 70 3.0% 94 1.4% counties Midwest Metropolitan 185 25.6% 24 18.4% 76 8.2% 285 17.6% markets State nonmetro. 477 5.5% 3 0.04% 197 3.4% 677 3.0% counties South Metropolitan 240 33.5% 48 32.3% 270 43.5% 558 36.1% markets State nonmetro. 480 11.7% 1 0.1% 390 6.0% 871 5.9% counties West Metropolitan 75 15.1% 15 17.7% 50 15.0% 140 15.8% markets State nonmetro. 331 3.0% 0 68 1.3% 399 1.4% counties Puerto Rico Metropolitan 68 1.0% 0 0 0 0 68 0.3% markets State nonmetro. 10 <0.05% 0 0 0 0 10 <0.05% counties National Metropolitan 585 79.3% 121 99.7% 468 86.2% 1,174 88.3% markets State nonmetro. 1,318 20.7% 8 0.3% 725 13.8% 2.051 11.7% counties All Counties 1,804 100.0% 129 100.0% 1,193 100.0% 3,225 100.0% NOTES: CBSA = core-based statistical area; RVU = relative value unit. SOURCE: RTI Analysis of CMS GPCI County Data File for 2012. For the simulations in the year 2 report, smoothing adjustments have been computed for the revised HWI, for the CBSA wage component of the practice expense GPCI and work GPCI, and for the payment locality-based GPCIs. Exhibit A-6 shows the distribution of smoothing factors applied to the new HWI and the CBSA GPCIs. For most counties, smoothing adjustments are very small. RTI found find that smoothing behaves exactly as was expected: commuters tend to move from lower-wage areas to higher-wage areas (making positive adjustments more common

OCR for page 143
152 GEOGRAPHIC ADJUSTMENT IN MEDICARE PAYMENT EXHIBIT A-6 County Smoothing Adjustments, by Type of Index For GPCIs For Hospital PE, Wage PE, Service Physician Wage Index Component Component 25% Work PE Number of Counties Affected 1,573 3,220 3,220 3,220 Value of Smoothing Factors Mean 1.0050 1.0055 1.0028 1.0009 Standard Deviation 0.0133 0.01710 0.0099 0.0037 Minimum 0.9720 0.8983 0.9280 0.9682 1st Percentile 0.9860 0.9770 0.9801 0.9908 10th Percentile 0.9960 0.9960 0.9977 0.9990 50th Percentile 1 1 1 1 90th Percentile 1.0190 1.0233 1.0129 1.0045 99th Percentile 1.0670 1.0747 1.0382 1.0128 Maximum 1.1240 1.1984 1.1393 1.0563 NOTES: Smoothing adjustments computed for CBSA markets. GPCI = geographic practice cost index; PE = practice expense. SOURCES: RTI simulations; American Community Survey 5-Year Journey-to-Work data. than negative ones), commuting across market boundaries is more common in counties that are adjacent to other markets, and the largest smoothing adjustments are computed for rural counties adjacent to metropolitan areas. Exhibit A-7 confirms this in a graph of the average smoothing factors by USDA RuralUrban Continuum Code. The committee considers the smoothing adjustment to be a type of refinement to the labor markets, one that reduces inaccuracies caused by the inherent limitation of representing economic markets by fixed political boundaries. The IOM committee's version of smoothing adjustment is similar in many ways to the outmigration adjustment that CMS now computes for hospitals that are not reclassified, but the CMS adjustments are only positive; commuting patterns from a higher to a lower-index market are not included in the computations. Exhibit A-8 compares the size of commuter-based smoothing adjustments to the size of CMS' outmigration adjustments as well as reclassifications. 2.5 County Indicators for Health Professional Shortage Areas 2.5.1Background The Health Resource Services Administration (HRSA) identifies Health Professional Shortage Areas geographically (by census tract) and by specific institution (Federally Qualified Health Centers or other safety net providers).5 HRSA also maps these designated areas or populations to counties, and provides a three-level county shortage area indicator that is published annu- ally in the Area Resource File (ARF). In the ARF variable, counties are identified only as "not a shortage county," a "full shortage county," and a "partial shortage county." Many counties are identified as "partial," particularly in metropolitan areas, and "partial" status gives no indication 5 See http://bhpr.hrsa.gov/shortage.

OCR for page 143
APPENDIX A 153 For Hospital Wage Index (Unit of Observation Is Hospital) metro pop>1M, central cty metro pop>1M, other cnty metro pop2501M metro pop<250K nonmetro adj u_pop>20K nonmetro not adj u_pop>20K nonmetro adj u_pop 2.520K nonmetro not adj u_pop 2.520K nonmetro adj u_pop2.5K nonmetro not adj u_pop2.5K .98 1 1.02 1.04 1.06 Smoothing Factor Plots exclude extreme values. Shaded areas incorporate 25th to 75th percentile. For PE-GPCI (Wage Component) (Unit of Observation Is County) metro pop>1M, central cty metro pop>1M, other cnty metro pop2501M metro pop<250K nonmetro adj u_pop>20K nonmetro not adj u_pop>20K nonmetro adj u_pop 2.520K nonmetro not adj u_pop 2.520K nonmetro adj u_pop2.5K nonmetro not adj u_pop2.5K .98 1 1.02 1.04 1.06 Smoothing Factor Plots exclude extreme values. Shaded areas incorporate 25th to 75th percentile. RUCC categories not available for Puerto Rico. Exhibit EXHIBIT A-7 Commuter-based smoothing A-7.epsby RuralUrban Continuum Code. adjustments NOTES: Smoothing adjustments computed for CBSA markets. Adjustments have been made budget-neutral to offset the effect of a larger number of positive than negative adjustments. SOURCE: RTI simulations; American Community Survey 5-Year Journey-to-Work data.

OCR for page 143
162 GEOGRAPHIC ADJUSTMENT IN MEDICARE PAYMENT EXHIBIT A-12 Budget Neutrality Factors Imposed on IOM Committee Indexes Hospital Wage Index IOM committee proposed index made neutral to CMS final post-reclassification 1.0175011 GPCI Components, If Using Current Payment Localities for CMS and IOM Committee Indexes IOM committee PE smoothed wage component, made neutral to CMS wage component 1.004498 IOM committee PE smoothed service component, made neutral to CMS service component 0.999000 IOM committee smoothed work GPCI, made neutral to CMS work GPCI 0.999600 GPCI Components, If Using Current CBSA Markets for IOM Committee Index Only IOM committee PE smoothed wage component, made neutral to CMS wage component 1.004240 IOM committee PE smoothed service component, made neutral to CMS service component 0.998920 IOM committee smoothed work GPCI, made neutral to CMS work GPCI 0.999600 NOTES: CBSA = core-based statistical area; CMS = Centers for Medicare & Medicaid Services; GPCI = geographic prac- tice cost index; IOM = Institute of Medicine; PE = practice expense. SOURCE: RTI Simulations. or it can be implemented at the end of the rate-setting process by altering the underlying national standardized rate or conversion factor. In keeping with CMS's current approach, we have imposed budget neutrality to current CMS payments using across-the-board adjustments to the index values. For the GPCIs, budget neutrality factors are computed by estimating aggregate payments under both the IOM committee recommendations and under CMS policy and dividing the IOM committee estimate by the CMS estimate. All index values constructed from IOM committee recommendations are then divided by this factor to achieve a budget-neutral index, and IOM committee payments are then reestimated using this adjusted index. This is computationally equivalent to saying that final payment-weighted averages of the two indexes being compared will be 1.00. Adjusting the HWI for budget neutrality requires several more steps to account for (a) the labor related share that is applicable to operating costs and (b) a separate wage index adjust- ment that is applicable to capital costs. Specifically, the wage index is applied to either 62 or 68.8 percent of the operating rate, but it is applied to a variable proportion of the capital rate because it is based on an exponential function (as described in Section 2.6.1). For operating costs, the neutrality factor would have to be adjusted as follows: Neutrality Factor = NF = IOMpmts CMSpmts HWI Neutrality Adjuster = [(1 NF) labor share] + 1 The neutrality adjustment for combined operating and capital payments is only approximately equal to this adjusted neutrality factor. While it is possible to compute separate HWI operating and capital neutrality factors, for ease of computation RTI simply iterated (starting at a value between 62 and 68.8 percent of the computed payment ratio) and recomputed payments until arriving at a value that resulted in equivalency between the two sets of payments. Because there are several steps in the payment computations, and because RTI simulated payment impacts for subsets of the recommendations as well as for all of them combined,

OCR for page 143
APPENDIX A 163 multiple neutrality factors had to be computed.14 The final neutrality adjustment factors for the HWI and each of the three GPCIs are presented in Exhibit A-12. The HWI neutrality factor is larger than the GPCI neutrality factors because of the shift from hospital to BLS data. By construction, the employment-weighted average of the BLS-based index is always 1.00--or put another way, the index is normalized to a value 1.00 based on employment. In contrast, the CMS HWI is normalized to a value of 1.00 based on hospital hours paid (although CMS budget neutrality factors ultimately adjust this to a value of 1.00 based on payment dollars). The HWI neutrality adjustment effectively renormalizes the BLS-based index to a value of 1.00 based on payment dollars. 3 PAYMENT IMPACT EXHIBITS Exhibits A-13 through A-30 are offered to provide additional detail on the payment simula- tion results. They are divided into Section 3.1 (related to impact on IPPS hospital payments) and Section 3.2 (related to impact on physician payments). They are presented without com- mentary, but are offered to supplement the analyses provided in Chapter 2 of the main report. 3.1 IPPS Hospital Estimates 3.2 Part B Physician and Other Practitioner Estimates Combined Effects of All Recommendations 400 300 Number of Hospitals 200 100 0 20 15 10 5 0 5 10 15 20 % Change in IPPS Payments In nonmetropolitan areas (64% positive) In metropolitan areas (57% positive) EXHIBIT A-13 Distribution of payment impact across all IPPS hospitals. SOURCE: RTI simulations. Exhibit A-13.eps 14 Forexample, before adding the independent benefits index to the base wage indexes, it was necessary to impose a budget neutrality adjustment on both such that the payment-weighted average of both would equal 1.00.

OCR for page 143
164 GEOGRAPHIC ADJUSTMENT IN MEDICARE PAYMENT EXHIBIT A-14 Estimated Change in IPPS Payments, Isolated by Type of IOM Committee Recommendation Percent Difference in Estimated IPPS Payments (IOM Committee Relative to Current Policy) Policy (isolated effect Data of eliminating (isolated effect Market adjustments, All of move to BLS- (isolated effect reclassifications (combined effects based wages) of smoothing) and floors) of all IOM Compares Compares committee payments under payments under recommendations-- benefits-adjusted benefits-adjusted Compares payments BLS-based index BLS-based index payments under under final with no other after smoothing CMS final post- IOM committee adjustments to payments reclassified index recommended to payments under benefits- to payments HWI compared to under CMS's adjusted BLS- under CMS's payments under occupation-mix based index occupation-mix current CMS adjusted index before smoothing adjusted index policy) Distribution Across Hospitals Minimum 13.5% 1.8% 15.0% 16.7% 5th percentile 3.7% 0.5% 1.8% 6.8% 25th percentile 1.6% 0.1% 1.5% 1.9% 50th percentile 0.5% 0.1% 1.4% 0.7% 75th percentile 2.4% 0.1% 0.4% 3.0% 95th percentile 4.5% 1.4% 8.2% 5.4% Maximum 16.8% 7.7% 21.9% 14.3% Average by Region and Metropolitan Status Northeastmetro. 0.9% 0.05% 1.9% 2.7% Nonmetro. 1.8% 0.4% 2.5% 0.3% Midwestmetro. 0.5% 0.1% 0.9% 1.4% Nonmetro. 1.9% 0.3% 2.6% 0.5% Southmetro. 0.8% 0.1% 1.1% 1.9% Nonmetro. 0.9% 0.4% 2.6% 0.2% Westmetro. 2.5% 0.01% 0.5% 2.1% Nonmetro. 2.0% 0.1% 2.3% 3.1% NOTES: BLS = Bureau of Labor Statistics; CMS = Centers for Medicare & Medicaid Services; HWI = hospital wage index; IOM = Institute of Medicine; IPPS = Inpatient Prospective Payment System. SOURCE: RTI Simulations.

OCR for page 143
APPENDIX A 165 EXHIBIT A-15 Impact of IOM Committee Recommendations on IPPS Payment, by USDA RuralUrban Continuum Code Difference Under IOM Committee Recommendations Payments In Frontier States In Other States Under Current Number Number Policy of % of % County RuralUrban Continuum Code (Billions) Hospitals Difference Hospitals Difference Metropolitan, population >1 million, central $56.4 14 0.9% 1236 0.2% Metropolitan, population >1 million, other $2.7 0 -- 162 1.4% Metropolitan, population 250K1 million $24.5 4 1.3% 636 0.1% Metropolitan, population <250K $14.5 24 5.8% 400 0.9% Non-metro, urbanized pop >20K, adjacent $4.3 1 6.1% 267 0.7% Non-metro, urbanized pop >20K, not adjacent $2.6 11 9.7% 118 1.4% Non-metro, urbanized pop 2.520K, adjacent $2.2 5 7.3% 293 0.9% Non-metro, urbanized pop 2.520K, not adj. $1.7 12 9.7% 180 1.1% Non-metro, urbanized pop <2.5K, adjacent $0.1 0 -- 24 2.7% Non-metro, urbanized pop <2.5K, not adjacent $0.1 1 11.5% 30 3.9% NOTES: IOM = Institute of Medicine; IPPS = Inpatient Prospective Payment System; USDA = U.S. Department of Agriculture. SOURCE: RTI Simulations. EXHIBIT A-16 Impact of IOM Committee Recommendations on IPPS Payment, by Census Division and Metropolitan Status Difference Under IOM Committee IPPS Recommendations Payments In Metropolitan In Nonmetropolitan Under Areas Areas Current Policy Number of % Number of % Location ($ Billions) Hospitals Difference Hospitals Difference New England $6.6 120 9.4% 23 1.6% Middle Atlantic $16.9 320 0.6% 69 1.0% East North Central $18.1 401 2.6% 120 2.5% West North Central $7.4 169 1.1% 99 5.1% South Atlantic $22.2 421 3.2% 170 0.9% East South Central $7.8 153 5.2% 170 2.6% West South Central $11.6 366 1.8% 183 2.5% Mountain $5.1 159 1.6% 66 4.5% Pacific $13.4 380 3.0% 29 3.8% NOTES: IOM = Institute of Medicine; IPPS = Inpatient Prospective Payment System. SOURCE: RTI Simulations.

OCR for page 143
166 GEOGRAPHIC ADJUSTMENT IN MEDICARE PAYMENT EXHIBIT A-17 Impact of IOM Committee Recommendations on IPPS Payment, by Hospital Reclassification Status IPPS Difference Under IOM Committee Payments Recommendations Under In Frontier States In Other States Current Hospital Reclassification or Policy Number Of % Number Of % Adjustment Status ($ Billions) Hospitals Difference Hospitals Difference Reclassifications (MGCRB) $19.5 2 4.1% 606 1.8% "Lugar" Hospitals $0.5 0 -- 53 1.4% Section 505 Outmigration Adjustments $4.6 2 11.7% 268 0.5% Frontier Floors $0.0 46 7.4% N/A -- Metropolitan Area Rural Floors $9.6 N/A -- 261 3.1% No Exceptions $73.2 22 0.5% 2158 1.0% NOTES: IOM = Institute of Medicine; IPPS = Inpatient Prospective Payment System; MGCRB = Medicare Geographic Classification Review Board. SOURCE: RTI Simulations. EXHIBIT A-18 Impact of IOM Committee Recommendations on IPPS Payment, by Special Rural Status Difference Under IOM Committee Payments Recommendations Under In Frontier States In Other States Current Policy Number of % Number of % Hospital Status ($ Billions) Hospitals Difference Hospitals Difference Sole Community Hospital (All) $5.9 32 6.0% 410 0.3% Medicare-Dependent Hospitals (All) $1.6 0 -- 211 2.0% Rural Referral Centers (Those Not SCH $5.5 2 8.0% 174 1.1% or MDH) All Other (Rural) $1.7 4 10.4% 219 1.3% All Other (Nonrural) $94.5 34 3.3% 2332 0.1% NOTES: IOM = Institute of Medicine; IPPS = Inpatient Prospective Payment System; MDH = Medicare-dependent hospi- tal; SCH = sole community hospital. SOURCE: RTI Simulations. EXHIBIT A-19 Impact of IOM Committee Recommendations on IPPS Payment, by Teaching and DSH Status IPPS Difference Under IOM Committee Payments Recommendations Under In Frontier States In Other States Current Policy Number of % Number of % Payment Status ($ Billions) Hospitals Difference Hospitals Difference Teaching Only $19.3 14 4.1% 402 0.3% Teaching and Disproportionate Share $44.2 5 4.2% 615 0.0% Disproportionate Share Only $35.2 25 4.8% 1,767 0.1% All Other (Nonteaching, Non-DSH) $10.2 28 3.5% 562 0.4% NOTES: DSH = disproportionate share hospital; IOM = Institute of Medicine; IPPS = Inpatient Prospective Payment System. SOURCE: RTI Simulations.

OCR for page 143
APPENDIX A 167 EXHIBIT A-20 Impact of IOM Committee Recommendations on IPPS Payment, by Bed Size Difference Under IOM Committee Recommendations In Frontier States In Other States Payments Under Current Policy Number of % Number of % Hospital Size ($ Billions) Hospitals Difference Hospitals Difference 50 beds $2.3 21 5.8% 654 0.6% 51100 beds $3.3 17 7.2% 614 0.7% 101300 beds $42.4 25 4.4% 1,419 0.1% 301500 beds $30.6 6 4.9% 444 0.8% >500 beds $27.4 3 0.1% 201 0.5% Not availablea $0.03 -- -- 14 0.3% NOTES: IOM = Institute of Medicine; IPPS = Inpatient Prospective Payment System. aNew hospitals; number of beds listed as "1" and data on bed days available are missing in 2012 impact file. SOURCE: RTI Simulations; Final IPPS Payment Impact Files for FY 2012. Combined Effects of All Recommendations 600 Number of Counties 400 200 00 30 20 10 0 10 20 % Change in Physician Payments In metropolitan areas (51% positive) In nonmetropolitan areas (1% positive) EXHIBIT A-21 Distribution of physician payment Exhibit impact across all counties. A-21.eps

OCR for page 143
168 GEOGRAPHIC ADJUSTMENT IN MEDICARE PAYMENT EXHIBIT A-22 Change in Aggregate Geographic Adjustment Factor, by Type of IOM Committee Recommendation Percent Difference in Payments Data Market Policy (isolated effect (isolated effect (isolated effect of adjusted of moving to of removing BLS data) CBSA markets) index floors) Compares payments using all IOM committee All (combined Compares recommended effects payments changes Compares of all IOM using adjusted including payments under committee BLS data but CBSA markets, current CMS recommendations-- keeping payment compared policy, including payments localities, to payments frontier and under final to payments using all IOM Alaska floors, IOM committee under CMS committee compared to recommended GPCIs but recommended payments under GPCIs compared excluding changes except CMS GPCIs but to payments under frontier and the CBSA excluding all current CMS Alaska floors markets index floors policy) Distribution Across Counties Minimum 3.5% 10.5% 26.2% 26.1% 5th percentile 0.8% 4.5% 4.1% 6.0% 25th percentile 0.4% 2.8% 0.0% 3.4% 50th percentile 0.0% 2.0% 0.0% 2.2% 75th percentile 0.3% 0.4% 0.0% 0.7% 95th percentile 1.0% 2.5% 0.0% 3.3% Maximum 6.8% 12.0% 0.0% 17.9% Average by Region and Metropolitan Status Northeastmetro. 0.4% 0.5% 0% 0.9% Nonmetro. 0.3% 2.0% 0.0% 1.7% Midwestmetro. 0.3% 0.2% 0.1% 0.3% Nonmetro. 0.6% 2.9% 0.2% 3.1% Southmetro. 0.2% 0.4% 0.0% 0.2% Nonmetro. 0.4% 2.7% 0.0% 3.1% Westmetro. 0.2% 0.2% 0.2% 0.3% Nonmetro. 0.7% 2.3% 0.8% 3.0% Puerto Ricometro. 1.6% 0.2% 0.0% 1.6% Nonmetro. 0.2% 0.6% 0.0% 0.8% NOTES: BLS = Bureau of Labor Statistics; CBSA = core-based statistical area; CMS = Centers for Medicare & Medicaid Services; GPCI = geographic practice cost index; IOM = Institute of Medicine.

OCR for page 143
APPENDIX A 169 EXHIBIT A-23 County Analysis of the Isolated Payment Effects from Redefining the GPCI Payment Areas All IOM All IOM IOM Committee Committee Committee Recommendations Recommendations Recommendations Including Including Excluding Market Market Market Redefinition Redefinition Redefinition --compared to-- --compared to-- --compared to-- IOM CMS CMS Recommendations GPCis GPCIs Excluding Without Without Market Index Index Redefinition Floors Floors All Counties Number of counties w/pmt reduction 2,464 2,467 1,658 Percent counties w/reduction 79% 79% 53% Percent national RVUs 51% 54% 50% Median % reduction 2.3% 2.4% 0.4% Aggregate average % reduction 1.5% 1.6% 0.4% Number of counties w/pmt increase 654 651 1,460 Percent counties w/increase 21% 21% 47% Percent national RVUs 49% 46% 50% Median % increase +1.5% +1.8% +0.3% Aggregate average % increase +1.6% +2.1% +0.6% Metropolitan Counties Only Number of counties w/pmt reduction 515 547 510 Percent metro. counties w/reduction 44% 48% 44% Percent national RVUs 45% 48% 48% Median % reduction 1.2% 1.3% 0.3% Aggregate average % reduction 1.2% 1.3% 0.4% Number of counties w/pmt increase 636 604 641 Percent metro. counties w/increase 66% 52% 56% Percent national RVUs 55% 52% 52% Median % increase +1.6% +2.0% +0.5% Aggregate average % increase +1.6% +2.1% +0.6% Nonmetropolitan Counties Only Number of counties w/pmt reduction 1,949 1,920 1,148 Percent counties w/reduction 99% 98% 58% Percent national RVUs 99.9% 99% 63% Median % reduction 2.5% 2.6% 0.4% Aggregate average % reduction 2.7% 0.2% 0.4% Number of counties w/pmt increase 18 48 819 Percent counties w/increase 1% 2% 42% Percent national RVUs 0.01% 1% 37% Median % increase +0.2% +0.6% +0.2% Aggregate average % increase +0.8% +0.5% +0.3% NOTES: CMS = Centers for Medicare & Medicaid Services; GPCI = geographic practice cost index; IOM = Institute of Medicine; pmt = payment; RVU = relative value unit.

OCR for page 143
170 GEOGRAPHIC ADJUSTMENT IN MEDICARE PAYMENT EXHIBIT A-24 Physician Payment Impact of IOM Committee Recommendations, by USDA RuralUrban Continuum Code In Frontier States In All and Alaska Other States Total Part B Practitioner Proportion % Proportion % Payments Billed Change in Billed Change in RuralUrban Continuum Code ($ Billions) RVUs Payments RVUs Payments metro, pop>1M, CBSA central countiesa $38.9 0.005 2.0% 0.492 1.0% metro, pop>1M, CBSA other countiesa $2.3 -- -- 0.031 3.0% metro, pop 250K1M $16.3 0.001 2.0% 0.221 0.0% metro, pop <250K $9.1 0.006 4.0% 0.120 1.0% adjacent nonmetro, urban pop >20K $3.2 0.0002 5.0% 0.045 3.0% not adjacent nonmetro, urban pop >20K $1.7 0.001 6.0% 0.022 3.0% adjacent nonmetro, urban pop 2.520K $1.9 0.0004 6.0% 0.026 3.0% not adjacent nonmetro, urban pop 2.520K $1.4 0.001 7.0% 0.019 4.0% adjacent nonmetro, urban pop <2.5K $0.1 0.0001 6.0% 0.002 2.0% not adjacent nonmetro, urban pop <2.5K $0.2 0.0002 6.0% 0.002 3.0% Puerto Rico (not coded) $0.2 -- -- 0.003 2.0% All counties $75.4 0.016 2.0% 0.984 0.0% NOTES: CBSA = core-based statistical area; IOM = Institute of Medicine; RVU = relative value unit; USDA = U.S. Depart- ment of Agriculture. aDesignation as "central" and "other" derived from CBSA indicators for metropolitan counties, as published in Area Resource File 2009. SOURCE: RTI simulations. EXHIBIT A-25 Physician Payment Impact of IOM Committee Recommendations, by Revised Health Professional Shortage Area Indicator In Frontier States In All and Alaska Other States Total Part B County Shortage Indicator, Practitioner % % as Defined by Estimated Payments Proportion Change in Proportion Change in Population in Bonus Areas ($ Billions) Billed RVUs Payments Billed RVUs Payments In Metropolitan Counties Non-HPSA (0%) $31.4 0.001 5.0% 0.410 1.0% Partial HPSA (20%) $34.7 0.012 1.0% 0.448 0.0% Partial HPSA (20% to 80%) $0.5 0.00001 6.0% 0.007 1.0% Partial HPSA (80% to <100%) $0.2 0.0003 5.0% 0.002 1.0% Full HPSA (100%) $0.1 0.002 1.0% In Nonmetropolitan Counties Non-HPSA (0%) $3.3 0.001 6.0% 0.045 3.0% Partial HPSA (20%) $4.0 0.002 6.0% 0.055 3.0% Partial HPSA (20% to 80%) $0.3 0.00003 4.0% 0.004 3.0% Partial HPSA (80% to <100%) $0.5 0.0003 6.0% 0.007 3.0% Full HPSA (100%) $0.3 0.0004 6.0% 0.004 4.0% In All Counties $75.3 0.016 2.0% 0.984 0.0% NOTES: HPSA = Health Professional Shortage Area; IOM = Institute of Medicine; RVU = relative value unit. SOURCES: RTI simulations; CMS ZIP primary care bonus area file (2012).

OCR for page 143
APPENDIX A 171 EXHIBIT A-26 Physician Payment Impact of IOM Committee Recommendations, by County Ranking in Median Family Income In Frontier States In All and Alaska Other States Total Part B Practitioner Proportion % Proportion % Counties Arrayed by Payments Billed Change in Billed Change in Median Household Income ($ Billions) RVUs Payments RVUs Payments Lowest quartile $2.9 0.000 7.8% 0.041 2.4% 25th to 50th percentile $8.7 0.000 6.5% 0.121 1.7% 50th to 75th percentile $19.9 0.004 5.4% 0.268 0.5% Top quartile $43.9 0.011 1.0% 0.555 0.7% All counties $75.4 0.016 2.4% 0.984 0.0% NOTES: IOM = Institute of Medicine; RVU = relative value unit. SOURCES: RTI simulations; HRSA Area Resource File (2009). EXHIBIT A-27 Physician Payment Impact of IOM Committee Recommendations, by County Ranking in Percent Non-Whitea Population In Frontier States In All and Alaska Other States Total Part B Practitioner Proportion % Proportion % From Counties Arrayed by Payments Billed Change in Billed Change in Percent Non-White Population ($ Billions) RVUs Payments RVUs Payments Lowest quartile $2.0 0.001 5.4% 0.027 2.1% 25th to 50th percentile $7.7 0.005 5.6% 0.102 0.7% 50th to 75th percentile $29.0 0.005 3.7% 0.381 0.3% Top quartile $36.6 0.005 1.7% 0.473 0.0% Total $75.3 0.016 2.5% 0.984 0.0% aComputed as (1 %white non-Hispanic). NOTES: IOM = Institute of Medicine; RVU = relative value unit. SOURCES: RTI simulations; HRSA Area Resource File (2009). EXHIBIT A-28 Physician Payment Impact of IOM Committee Recommendations, by Rural County Population Density In Frontier States In All and Alaska Other States Total Part B Counties by Rural/Urban Status Practitioner Proportion % Proportion % and Grouped by Persons per Payments Billed Change in Billed Change in Square Mile ($ Billions) RVUs Payments RVUs Payments Frontier Counties (6 persons/sq. mi.) $0.2 0.0011 5.6% 0.002 4.0% Other rural, below 50th percentile $1.4 0.002 5.6% 0.018 3.0% Other rural, above 50th percentile $6.7 0.0006 7.3% 0.094 3.0% For comparison: all urban $67.0 0.013 1.4% 0.870 0.0% All counties $75.3 0.016 2.4% 0.984 0.0% NOTES: IOM = Institute of Medicine; RVU = relative value unit. SOURCES: RTI simulations; HRSA Area Resource File (2009).

OCR for page 143
172 GEOGRAPHIC ADJUSTMENT IN MEDICARE PAYMENT EXHIBIT A-29 Physician Payment Impact of IOM Committee Recommendations, by Percent of Total RVUs Accounted for by Primary Care Practitioners In Frontier States In All and Alaska Other States Total Part B From Counties Arrayed by Practitioner Proportion % Proportion % Primary Care Practitioner Payments Billed Change in Billed Change in RVUs as Percent Total RVUs ($ Billions) RVUs Payments RVUs Payments Lowest quartile $47.5 0.007 3.8% 0.618 0.2% 25th to 50th percentile $23.9 0.008 0.5% 0.311 0.0% 50th to 75th percentile $3.2 0.0008 6.8% 0.044 1.3% Top quartile $0.7 0.0003 6.6% 0.010 2.1% All counties $75.3 0.016 0.0% 0.984 0.0% NOTES: Primary Care practitioners include all those self-identified identified as internists, geriatricians, family p ractitioners and pediatricians, plus RVUs billed nurse practitioners and physician assistants. IOM = Institute of Medicine; RVU = relative value unit. SOURCES: RTI simulations; 2010 RVU file provided to RTI by Acumen, LLC. EXHIBIT A-30 Effect of Level of Physician Work GPCI on Estimated IOM Committee Payment Differences, for Counties Grouped by Revised Health Professional Shortage Area Indicators Percent Difference in Part B Payments by Level of Physician Work Adjustmenta Counties Grouped by Estimated Share Current Range of Optional Policies Proportion of Their Population That Policy: Proportion Primary Is Located in CMS Bonus 25% of 0% of 100% of Part B Care Payment ZIP Codes Work GPCI Work GPCI Work GPCI Enrolleesb RVUsc Metropolitan Counties 0% 0.7% 0.3% 1.8% 0.365 0.384 20% 0.1% 0.0% 0.5% 0.369 0.437 20% to 80% 0.8% 0.3% 2.2% 0.011 0.008 80% to <100% 1.3% 0.4% 4.0% 0.012 0.005 100% 1.4% 0.0% 5.6% 0.004 0.003 Subtotal 0.4% 0.1% 1.1% 0.761 0.836 Nonmetropolitan Counties 0% 2.8% 1.0% 8.1% 0.089 0.063 20% 3.0% 1.1% 8.6% 0.096 0.075 20% to 80% 2.5% 1.2% 6.3% 0.010 0.005 80% to <100% 3.0% 1.2% 8.6% 0.032 0.013 100% 3.7% 2.0% 9.0% 0.012 0.006 Subtotal 2.9% 1.1% 8.4% 0.239 0.164 All Counties 0.0% 0.0% 0.0% 1.000 1.000 NOTES: CMS = Centers for Medicare & Medicaid Services; GPCI = geographic practice cost index; IOM = Institute of Medicine; RVU = relative value unit. aFor each choice of percent physician work adjustment, value is defined as difference between payments estimated with GPCIs computed using all of the IOM committee's recommendations, relative to payments estimated under current CMS policy including all floors. bShare of Medicare beneficiaries enrolled in Part B fee-for-service program, from calendar 2009 (most recent county data available for download as of January 2012). cShare of national total Part B RVUS billed in 2010 by physicians identified as internists, geriatricians, family practitioners and pediatricians, plus RVUs billed nurse practitioners and physician assistants. SOURCES: RTI simulations; 2010 RVU file provided to RTI by Acumen, LLC.