Appendix E

Data and Methods to Analyze Medicare GME Payments

The committee’s analyses, presented in Appendix F, are based on Medicare cost reports for the latest cost reporting periods beginning on or after May 1, 2010, as of the December 31, 2012, update of the Healthcare Cost Report Information System (HCRIS). Only teaching hospitals that reported having current-year residents in approved training programs were included. Hospitals with no current-year residents that received GME funding through the rolling average were excluded. The final analytic file included 207 cost reports beginning in fiscal year (FY) 2010 (mainly beginning on July 1, 2010) and 885 cost reports beginning in FY 2011 (beginning on or after October 1, 2011). The data were not adjusted to account for differences in the beginning dates of the cost reporting periods.

Most information used in the impact analysis was derived from Worksheet E-4, Form CMS-2552-10 (WS E4). The distribution of resident counts by type of hospital is shown in Table E-1. The type of hospital was assigned according to the Medicare provider number. The unweighted direct graduate medical education (DGME) resident count is the sum of the reported unweighted number of allopathic and osteopathic residents for the current year (WC E4, line 6) and the weighted dental and podiatric resident FTE count for the current year (WS E4, line 10). Unweighted counts for the dental and podiatric residents are not available.

TABLE E-1 Number of Hospitals and Total Direct Graduate Medical Education (DGME) Unweighted Resident Count by Type of Hospital

^a Freestanding hospitals only; residents in units are included in the general acute care count.

ESTIMATE OF THE NATIONAL PRA (Appendix F, TABLE F-2)

1. Determine the national average DGME PRA based on an estimate of total Medicare DGME payments and total DGME-weighted FTE resident count used in the payment determination net of children’s hospitals.

a.  Total Medicare DGME payments = sum of Part A allocation (WS E, line 49) and 80 percent of Part B allocation (0.8 * WS E, Line 50)

b.  Total DGME weighted/capped resident count = sum of adjusted rolling average FTE count (WS 4, line 17, col. 1 + 2) and a derived weighted allowable additional direct GME FTE count (WS 4, line 24 ÷ line 23)

2.  Determine a budget-neutral per-resident amount that, when adjusted by the GAF, would result in estimated payments equivalent to the total DGME payments determined in Step 1.The national average per resident amount (used to determine payment for additional slots beyond the 1996 cap) is adjusted by the geographic adjustment factor (GAF) used in the physician fee schedule.

a.  Use the county/CBSA codes from the cost report to assign the appropriate 2013 GAF to each hospital

b.  Determine the aggregate GAF-adjusted DGME payments using the DGME PRA from Step 1 = Sum of (Step 1a * GAF)_hosp

c.  Determine a budget neutrality factor = Step 1a/Step2b

d.  Determine the budget-neutral DGME PRA = Step 2b * Step 2c/Step 1b

3.  For acute care hospitals only, determine the national average IME PRA based on an estimate of total IME payments for operating plus IME for capital-related costs.

a.  Current allowable IME for operating costs = sum of WS EA, line 28

b.  Current allowable IME for capital-related costs = sum of WS L, Part I, line 6

c.  Total IME capped resident count = Current allowable FTE count (WS EA, line 18)

4.  Determine a budget-neutral per-resident amount that, when adjusted by the GAF, would result in estimated payments equivalent to total IME payments at analytically justified level

a.  Analytically justified IME payments = Step 3a * 0.5 + Step 3b

b.  Determine the aggregate GAF-adjusted IME payments using the GAF determined in Step 2a = Sum (Step 4a * GAF)_hosp

c.  Determine a budget neutrality factor = Step 4a/Step 4b

d.  Determine the budget-neutral IME PRA= Step 4b * Step 4c/Step 3c

ESTIMATED IMPACTS FOR ACUTE CARE PPS HOSPITALS (Appendix F, TABLE F-3)

Hospital Characteristics

1. Number of residents = unweighted DGME current allopathic and osteopathic count (WS E4, line 6) plus weighted dental and podiatric resident FTE count (WS E4, line 10)
2. Medicare share = ratio of Medicare days to total inpatient days for Part A (WS E4, Line 28 column 1) and managed care (WS EA, Line 28, column 2)
3. Medicare discharges = WS S3, column 13, line 14

4. Low-income patient percentage

a.  If the SSI percentage is greater than 0, (SSI percentage (WS L, Part I, line 7) * Medicare days (WS S3, column 6, line 14) + Medicaid days (WS S3, column 7, line 14))/total inpatient days (WS S3, column 8, line 14)

b.  If the SSI percentage is missing, (Medicare days * Medicaid days/total inpatient days + Medicaid days)/total inpatient days

Impacts

The impacts were determined at the hospital level and summarized by aggregating the results by hospital characteristic.

1. Consolidated PRA Payments = From Table F-2, GAF-adjusted DGME PRA * DGME weighted/capped resident count + budget neutral GAF-adjusted IME PRA * IME capped counts
2. Total current GME payments = current DGME payments + current IME payments
3. Current average payment per resident = ∑ current GME payments/∑ total weighted DGME count
4. Change in average payment per resident= ∑(Consolidated payments – current GME payments)/∑ weighted DGME count)
5. Percent difference attributable to IME reduction = ∑ (.5 × current IME payments – current IME payments)/∑ total current GME payments
6. Percent differences attributable to other changes = ∑(Consolidated PRA payments – (current GME payments – 0.5 current IME payments)/∑ total current GME payments

Derived variables pertaining to hospital categories were determined as follows:

• Program size was based on the number of reported residents in the facility (from Worksheet S-3).
• The percentage of primary care residents was determined as the percentage of weighted residents in primary care programs (defined consistent with the Medicare PRA differential as residents in family medicine, general internal medicine, general pediatrics, preventive medicine, geriatric medicine, osteopathic general practice, and obstetrics/gynecology) to the total weighted residents in primary care and other specialty allopathic/osteopathic programs (i.e., exclusive of residents in podiatric and dental programs). Because

• residents in non-primary-care specialty programs are more likely to be weighted at 0.5 FTE, the percentage primary care is overstated.

• Status under cap is a comparison of the hospital’s unweighted GME allopathic and osteopathic resident count cap with the total number of residents reported based on the 1996 cap adjusted for new programs and the reallocation of residency slots. In the 2008 cost reports, there were 44 hospitals with only dental/podiatric residency programs and 26 hospitals with GME costs that did not report a current-year resident count on Worksheet E-3, Part IV.
• Medicare utilization was defined consistent with Medicare’s share for purposes of determining direct GME payments ((Medicare fee-for-service + managed care days)/total inpatient days).

The comparison of 2008 GME costs and payments included the 1,103 hospitals that reported both GME costs and a 2008 resident count for purposes of direct GME payments. Except where noted, the resident counts are taken from Worksheet E-3, Part IV CMS-2552-1996.

This page intentionally left blank.