CHAPTER Hospital inpatient and outpatient services R 3-1 E C O M M E N D A T I O N S The Congress should increase payment rates for the inpatient and outpatient prospective payment systems in 2013 by 1.0 percent. For inpatient services, the Congress should also require the Secretary of Health and Human Services beginning in 2013 to use the difference between the increase under current law and the Commission's recommended update to gradually recover past overpayments due to documentation and coding changes. COMMISSIONER VOTES: YES 16 o NO 1 o NOT VOTING 0 o ABSENT 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-2 The Congress should direct the Secretary of Health and Human Services to reduce payment rates for evaluation and management office visits provided in hospital outpatient departments so that total payment rates for these visits are the same whether the service is provided in an outpatient department or a physician office. These changes should be phased in over three years. During the phase-in, payment reductions to hospitals with a disproportionate share patient percentage at or above the median should be limited to 2 percent of overall Medicare payments. COMMISSIONER VOTES: YES 14 o NO 2 o NOT VOTING 1 o ABSENT 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-3 The Secretary of Health and Human Services should conduct a study by January 2015 to examine whether access to ambulatory physician and other health professionals' services for low-income patients would be impaired by setting outpatient evaluation and management payment rates equal to those paid in physician offices. If access will be impaired, the Secretary should recommend actions to protect access. COMMISSIONER VOTES: YES 17 o NO 0 o NOT VOTING 0 o ABSENT 0 C H A P T E R Hospital inpatient and outpatient services Chapter summary 3 In this chapter From 2009 to 2010, Medicare payments per fee-for-service (FFS) beneficiary for inpatient and outpatient services in acute care hospitals grew by more than 3 percent. As a result, the 4,800 hospitals paid under the Medicare prospective payment system and critical access hospital payment system received $153 billion for roughly 10 million Medicare inpatient admissions and 166 million outpatient services. To evaluate whether aggregate payments were adequate, we consider beneficiaries' access to care, the volume of services provided, hospitals' access to capital, quality of care, and the relationship of Medicare's o Are Medicare payments adequate in 2012? o Rural hospital payments and costs o How should Medicare payments change in 2013? payments to the average cost of caring for Medicare patients. In addition to examining the costs of the average provider, we compare Medicare payments with the costs of relatively efficient hospitals. We also discuss the equity in Medicare payments across regions and across sectors. We examine the equity of rural hospital payments compared with urban hospital payments. We also examine the payment rates for evaluation and management (E&M) clinic visits in hospital outpatient departments (OPDs) compared with rates paid for E&M visits at freestanding physician offices. Assessment of payment adequacy and update recommendation The Commission balanced three factors in reaching its inpatient update recommendation. First, most payment adequacy indicators (including access Report to the Congress: Medicare Payment Policy | March 2012 45 to care, quality of care, and access to capital) are positive. Second, hospitals changed their documentation and coding starting in 2008 in response to the introduction of Medicare severity-diagnosis related groups (MS-DRGs), leading to overpayments in 2008 through 2012. While 2008 and 2009 overpayments are currently being recovered, the 2013 updates must be lowered to recover the overpayments from 2010, 2011, and 2012. Third, while relatively efficient hospitals generated positive overall Medicare margins in 2010, most hospitals have negative overall Medicare margins (-4.5 percent in 2010, projected to reach -7 percent in 2012). Balancing these factors, the Commission recommends reducing the 2013 increase in inpatient payments from the level in current law (expected to be 2.9 percent) to 1 percent. The difference between the update under current law and 1 percent should be used to gradually recover overpayments that occurred between 2010 and 2012 due to documentation and coding changes. This update recommendation will allow Medicare to recover past overpayments and keep 2013 inpatient payment rates adequate. For outpatient services, the Commission also recommends a 1 percent increase in payment rates. On the one hand, growth in the volume of outpatient services has been strong, suggesting the outpatient update in current law (1.9 percent) may be too high. On the other hand, overall hospital margins are negative, suggesting a positive update is appropriate. A 1 percent update would balance these two considerations and help limit growth in the disparity in payment rates between services provided in OPDs and payment rates in other sectors. The Commission maintains that Medicare should seek to pay similar amounts for similar services, taking into account differences in the quality of care and in the relative risks of patient populations. Beneficiaries' access to care--Access measures include the capacity of providers and changes in the volume of services over time. These measures were positive for the period reviewed. o Capacity and supply of providers--The number of hospitals and the range of services offered continue to grow. o Volume of services--Outpatient volume has continued to grow at a robust pace, while per beneficiary inpatient admissions continued to decline. Inpatient admissions per FFS beneficiary declined 1 percent per year from 2004 to 2010 and 1.3 percent from 2009 to 2010. Inpatient use also has declined among non-Medicare patients, and as a result inpatient occupancy has declined as well. The volume of hospital outpatient services per Medicare FFS beneficiary grew on average by 4.2 percent per year from 2004 to 2010 and by 4.0 percent 46 Hospital inpatient and outpatient services: Assessing payment adequacy and updating payments from 2009 to 2010. Part of the growth was due to a shift of services from the inpatient to the outpatient setting. Twenty percent of all outpatient volume growth, however, was due to a shift in physician office visits from freestanding physician offices to hospital-owned physician offices that are deemed parts of OPDs. Hospital-based outpatient physician office visits grew by 6.7 percent from 2009 to 2010. Quality of care--Quality continues to improve on most measures. Hospitals reduced in-hospital and 30-day mortality rates across five prevalent clinical conditions. Patient safety indicators have generally improved, but readmission rates have not improved significantly. Providers' access to capital--Access to capital has been volatile in recent years but appears adequate at this time. As inpatient use and hospital occupancy declined, hospitals slowed the pace of new construction and shifted spending toward outpatient facilities and remodeling existing inpatient facilities. Medicare payments and providers' costs--Overall aggregate Medicare margins improved from -7.1 percent in 2008 to -4.5 percent in 2010. The margins improved for two reasons. First, hospitals faced a decline in their profitability and investment portfolios in the fall of 2008. After the decline in the economy, they constrained cost growth in 2009 and 2010. Second, they made changes in clinical documentation and coding of patients' diagnoses on hospital claims in response to the adoption of MS-DRGs, which increased hospital payments from 2008 through 2010. Starting in 2011, CMS made two payment adjustments related to overpayments from documentation and coding changes. First, CMS adjusted payments in 2011 and 2012 to recover overpayments made in 2008 and 2009. Second, CMS reduced the 2012 update to begin to reduce further overpayments. While the documentation and coding changes contributed to margins improving from 2008 to 2010, changes put in place to recover these overpayments will cause margins to decline from -4.5 percent in 2010 to a projected margin of roughly -7 percent in 2012. o Efficient providers--While Medicare payments are currently less than costs for the average hospital, a key question is whether current Medicare payments are adequate to cover the costs of efficient providers. To explore this question, we have examined financial outcomes for a set of hospitals that consistently perform relatively well on cost, mortality, and readmission measures. We find that Medicare payments more than covered the costs of the median efficient hospital, with the median efficient hospital generating a 4 percent Medicare margin in 2010. Report to the Congress: Medicare Payment Policy | March 2012 47 o Rural hospital payments and costs--In the 1990s, rural hospitals generally had lower Medicare margins due to Medicare payment rules that tended to favor urban providers. After a series of changes in the law (some of which were recommended by the Commission), rural payments increased and rural Medicare margins have exceeded urban margins for the past seven years. Looking forward, we expect the differential between rural and urban margins to grow due to the introduction of a new temporary low-volume adjustment in 2011 and 2012. Addressing differences in payment rates across sites of care In an effort to move toward paying the same rates for the same service across different sites of care, we are recommending equalizing the rate paid for E&M visits in OPDs and freestanding physician offices. Under current policy, Medicare pays 80 percent more for a 15-minute office visit in an OPD than in a freestanding physician office. This payment difference creates a financial incentive for hospitals to purchase freestanding physicians' offices and convert them to OPDs without changing their location or patient mix. Indeed, E&M clinic visits provided in OPDs increased 6.7 percent in 2010, potentially increasing Medicare program and beneficiary expenditures without any change in patient care. To remove this distortion in the payment system, the Commission recommends making payments for E&M visits equal in the physician office and OPD settings. To smooth the transition to lower rates for E&M visits, rates should be equalized over a threeyear transition period. During the transition, we recommend limiting the policy's impact on providers serving a disproportionate share of poor patients and requiring the Secretary of Health and Human Services to study the policy's impact on lowincome patients' access to ambulatory physician and other health professional services. In the future, we plan to examine payment differentials between OPDs and physician offices for other services. ? 48 Hospital inpatient and outpatient services: Assessing payment adequacy and updating payments TABLE 3-1 Hospital services Growth in Medicare inpatient and outpatient spending 2005 2009 2010 Average annual change 2005-2009 Change 2009-2010 Inpatient services Total FFS payments (in billions) Payments per FFS enrollee $106 $113 $116 1.5% 2.7% 2,972 3,290 3,360 2.6 2.1 Outpatient services Total FFS payments (in billions) Payments per FFS enrollee 27 34 37 5.9 8.8 811 1,097 1,181 7.8 7.7 Inpatient and outpatient services Total FFS payments (in billions) Payments per FFS enrollee 133 147 153 2.5 4.1 3,783 4,387 4,541 3.8 3.5 Note: FFS (fee-for-service). Reported hospital spending includes all hospitals covered by Medicare's inpatient prospective payment system along with critical access hospitals. Maryland hospitals are excluded. Fiscal year 2010 payments include partial imputation to account for hospitals that typically do not submit their cost reports to CMS before CMS makes the most recent year available to the public. Although the number of Medicare beneficiaries grew significantly from 2005 to 2009, the number of FFS beneficiaries declined over that time due to the shift of beneficiaries to the Medicare Advantage program. The number of FFS beneficiaries increased slightly from 2009 to 2010. For the purposes of calculating payments per beneficiary we identified populations of beneficiaries eligible for inpatient (Part A) and outpatient (Part B) coverage and excluded enrollees in Maryland. Source: MedPAC analysis of CMS hospital cost reports and MedPAR files. Background Acute care hospitals provide Medicare beneficiaries with inpatient care for the diagnosis and treatment of acute conditions and manifestations of chronic conditions. They also provide ambulatory care through outpatient departments (OPDs) and emergency rooms. In addition, many hospitals provide home health, skilled nursing facility, psychiatric, and rehabilitation services. To be eligible for Medicare payment, short-term general and specialty hospitals must meet the program's conditions of participation and agree to accept Medicare rates as payment in full. Medicare spending on hospitals In 2010, Medicare paid acute care hospitals approximately $116 billion for fee-for-service (FFS) inpatient care and $37 billion for FFS outpatient care (Table 3-1). Acute inpatient and outpatient services represented more than 92 percent of Medicare FFS spending on acute care hospitals. From 2009 to 2010, Medicare inpatient spending per FFS beneficiary--including spending at critical access hospitals (CAHs)--grew, on average, by 2.1 percent, and outpatient spending per FFS beneficiary grew by 7.7 percent. Growth in the overall payment per FFS beneficiary was 3.5 percent; this amount was slightly below the average rate of growth of 3.8 percent from 2005 to 2009. The higher growth in outpatient spending reflects the ongoing shift of services from the inpatient to the outpatient setting, changes in available technology, and the growth in hospital-owned physician practices, which bill for physician office visits as outpatient services. Medicare's payment systems for inpatient and outpatient services Medicare's inpatient and outpatient prospective payment systems (PPSs) have a similar basic structure. Each has a base rate modified for differences in type of case or service as well as geographic differences in wages. However, in addition to different units of service, each PPS has a different set of payment adjustments. Acute inpatient payment system Medicare's acute inpatient PPS (IPPS) pays hospitals a predetermined amount for most discharges. The payment rate is the product of a base payment rate and a relative weight that reflects the expected costliness of cases in a particular clinical category compared with the average of all cases. The labor-related portion of the base payment rate is adjusted by a hospital geographic wage index to Report to the Congress: Medicare Payment Policy | March 2012 49 FIGURE More hospitals opened... 3-1 FIGURE 3-1 More hospitals opened than closed each year from 2002 to 2010 100 Opened Closed Number of hospitals 80 60 40 20 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Fiscal year Note: Hospitals refers to general short-term acute care hospitals. The Commission's reported number of open and closed hospitals can change from year to year because some hospitals may enter Medicare as acute care facilities but later convert to more specialized types of facilities, such as long-term care hospitals. Source: MedPAC analysis of CMS's Provider of Service file, Inpatient Prospective Payment System Final Rule Impact file, and hospital cost reports. Note: Note and Source are in InDesign. account for differences in area wages. Payment rates are Source: updated annually. weight based on its median cost of service compared with the median cost of a midlevel clinic visit. A conversion factor translates relative weights into dollar payment amounts. A more detailed description of the OPPS can be found at: http://www.medpac.gov/documents/MedPAC_ Payment_Basics_11_OPD.pdf. In 2008, CMS implemented a new clinical categorization system called Medicare severity-diagnosis related groups (MS-DRGs). The MS-DRG system classifies Notes about this graph: patient cases in 1 of 749 groups, which reflect similar o Data is in the datasheet. Make updates in the datasheet. principal diagnoses, procedures, and severity levels. o WATCH FOR determined according to The new severity levels areGLITCHY RESETS WHEN YOU UPDATE DATA!!!! Are Medicare payments adequate in whethero The column complication added manually. patients have a totals were or comorbidity (CC) associated with the base DRG (no CC, a nonmajor CC, 2012? o had to manually draw tick marks acute or a major ICC). A more detailed description of theand axis lines because they kept resetting when I changed any data. IPPS, including payment adjustments, can be foundhave to crop judge whether payments for 2012 are adequate to o I can't delete the legend, so I'll just at: To it out in InDesign. http://www.medpac.gov/documents/MedPAC_Payment_ modification. Otherwise hospitals incur, we examine cover the costs efficient if you use the black selection tool, they o Use direct selection tool to select items for Basics_11_hospital.pdf. several indicators of payment adequacy. We consider default when you change the data. beneficiaries' access to care, hospitals' access to capital, o Use paragraph styles (and Hospital outpatient payment systemobject styles) to format. changes in the quality of care, and the relationship of The outpatient PPS (OPPS) pays hospitals a predetermined o Data was from: R:\Groups\MGA\data bookMedicare's payments to hospitals' costs for both average 2007\data book 2007 chp1 and relatively efficient hospitals. Most of our payment amount per service. CMS assigns each outpatient adequacy indicators for hospitals are positive, but margins service to 1 of approximately 850 ambulatory payment on Medicare patients remain negative for most hospitals. classification (APC) groups. Each APC has a relative 50 Hospital inpatient and outpatient services: Assessing payment adequacy and updating payments We assess beneficiaries' access to care by tracking the number of hospitals participating in the Medicare program, the volume of services received, and the proportions of hospitals offering certain specialty services. In general, we find that access to hospital services is good and has expanded from the previous year. More hospitals opened than closed The number of acute care hospitals entering the Medicare program exceeded the number of hospitals exiting the program in 2010, and inpatient bed capacity remained relatively flat. In 2010, 30 acute care hospitals opened and 7 closed (Figure 3-1). It was the ninth consecutive year hospital openings exceeded closings. Approximately 4,800 short-term acute care hospitals participated in the Medicare program in 2010, of which about 1,300 were CAHs (Flex Monitoring Team 2011). Volume of services: Outpatient grew, inpatient declined From 2004 to 2010, the volume of Medicare outpatient services per FFS beneficiary increased at roughly a 4.2 percent average annual rate for a cumulative increase of 28 percent over the seven-year period (Figure 3-2). During the same period, Medicare inpatient discharges per FFS beneficiary declined at roughly a 1.0 percent average annual rate, a cumulative reduction of about 6 percent. To examine changes in volume of services, we used the number of discharges per FFS beneficiary as an indicator of inpatient volume and measured outpatient volume by the number of services per FFS beneficiary. The measurement units differ because the IPPS generally pays for a bundle of services, while the OPPS generally pays for individual services.1 The rapid growth in outpatient services coupled with the decline in inpatient services is consistent with a shift in site of service from inpatient care units to OPDs. Many surgical procedures, such as pacemaker implantation, that were once performed solely as an inpatient service are now often done in an outpatient setting as well. However, growth in the number of outpatient services is not purely a shift in settings from inpatient to outpatient care. About 20 percent of the increase in volume in OPDs is due to increased evaluation and management ("office") visits in OPDs. This increase could be a result of hospitals' FIGURE FIGURE 3-2 3-2 From 2004 to 2010, Medicare Medicare outpatient services grew outpatient discharges.... while hospital inpatient services grew while hospital inpatient discharges per FFS beneficiary declined 30 Outpatient services per FFS Part B beneficiary 25 Cumulative percent change Beneficiaries' access to care: Access remained positive, as hospital capacity generally grew over the period reviewed Inpatient discharges per FFS Part A beneficiary 20 15 10 5 0 -5 -10 2004 2005 2006 2007 2008 2009 2010 Calendar year Note: FFS (fee-for-service). Data include general and surgical hospitals, critical access hospitals, and children's hospitals. Source: MedPAC analysis of MedPAR and hospital outpatient claims data from CMS. Note: Note and Source in InDesign. acquisition of physician practices, which are then deemed part of the OPD. Such acquisitions can result in increased Medicare payments for office visits, even if the care provided does not change. In a freestanding practice, Medicare pays a physician based on the physician fee schedule, which includes a professional component (for Notes about this graph: the value of the physician's work), a practice expense component, o Data is in the datasheet. Make updates in the d and a professional liability insurance component. o I had to force in a hospital's OPD, the x-axis. They For an office visit return the items on Medicare pays a facility fee to the hospital and a reduced o I had to manually draw tick marks and axis line fee for the physician's services. The combined fees paid o Use direct practices tool 80 percent for visits to hospital-based selectioncan beto select items for modi greater than default when you change the data. rates paid to freestanding practices. In 2010, the volume of visitsparagraph stylesoutpatient-based o Use to the higher paid (and object styles) to form practices owned by hospitals grew by 6.7 percent, while visits to the lower paid freestanding practices grew by less than 1 percent.2 This finding suggests that the differential in payment rates is contributing to a shift in the site of service and underscores the need to equalize payment rates across sectors for office visits. The shift away from the inpatient setting is reflected in declining inpatient occupancy rates and a decline in the Report to the Congress: Medicare Payment Policy | March 2012 51 TABLE 3-2 Shares of urban and rural hospitals offering specific services, 2005-2010 Urban Type of service Percentage of hospitals in 2010 Rural Percentage point change 2005-2010 Percentage of hospitals in 2010 Percentage point change 2005-2010 High-tech services Robotic surgery 36% 22 PET or PET/CT scanner 60 10 16 2% 1 4 MRI 93 3 85 9 Palliative care 54 9 22 2 Indigent care clinic 37 9 11 4 Orthopedics 87 5 60 8 Open heart surgery 48 5 4 1 Cardiac catheterization 63 4 7 0 Oncology 76 1 39 2 Geriatrics 53 1 32 -1 Trauma center 46 1 37 4 Skilled nursing 35 -6 43 -3 Home health 61 -3 56 -5 Core services Post-acute services Note: CT (computed tomography). The American Hospital Association's annual survey generally has overall response rates of more than 80 percent, but response rates vary by line of service. Source: American Hospital Association annual survey of hospitals. share of beneficiaries using inpatient services. From 2004 to 2010, the overall hospital bed occupancy rate declined 2 percentage points, from approximately 68 percent to 66 percent.3 In addition, the share of Medicare FFS beneficiaries using inpatient hospital services declined 2 percentage points, from 22 percent to 20 percent from 2006 to 2010. Similarly, from 2006 to 2010, the number of Medicare inpatient bed days per beneficiary declined from 1.9 in 2006 to 1.7 bed days per beneficiary in 2010. For this utilization indicator, we observed wide variation across states. Oregon, Idaho, and Utah had consistently low rates of inpatient utilization (approximately 1 inpatient day per beneficiary) while Mississippi, Kentucky, and New York had consistently high inpatient utilization rates (approximately 2 inpatient days per beneficiary). Hospitals have continued to expand the scope of services they offer. Our analysis of 50 specialized hospital services from 2005 to 2010 found that the share of hospitals and their affiliates providing each of these services increased 52 for most services.4 New technologies, such as robotic surgery and PET services, were among those that grew most rapidly. Core hospital services, such as trauma care, cardiac services, and oncology, generally were offered by more hospitals in 2010 than in 2005. Post-acute care was the only area in which the share of hospitals offering a type of service declined by more than 1 percent. Rural hospitals tended to offer fewer high-tech services but have been expanding their imaging and orthopedic surgery offerings (Table 3-2). The change from 2009 to 2010 was similar to the average change for the six-year period. Access to capital: Access remains positive, as the industry focuses on shifting capacity to the outpatient setting In general, access to capital appears adequate. Access to capital allows hospitals to maintain and modernize their facilities. If hospitals were unable to access capital, it might in part reflect problems with the adequacy of Hospital inpatient and outpatient services: Assessing payment adequacy and updating payments FIGURE Construction.... 3-3 FIGURE 3-3 Spending on hospital construction slowed after 2008 but remains high 40 35 Dollars (in billions) 30 25 20 15 10 5 0 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Fiscal year Note: Spending is for nonfederal hospital construction and deflated to September 2011 dollars using McGraw-Hill's construction cost index. Data for 2011 are an annualized estimate based on data for the first five months of 2011. 16.9 17.8 19.3 23.0 24.4 24.9 26.5 30.6 Source: Census Bureau. http://www.census.gov/const/www/c30index.html. Note: Note and Source are in InDesign. Source: Medicare payments, as Medicare provides about 30 percent of hospital revenues. moderated to just over $27 billion by 2010 (Figure 3-3). Projects for 2010 and 2011 focused on outpatient services, such as emergency departments, imaging After the financial difficulties of 2008, hospitals began centers, and cancer centers, or involved the installation or controlling costs in 2009 in part by reducing their capital Notes about this graph: modernization of health information technology systems expenditure plans (Fitch Ratings 2011, Moody's Investors o Data is in the datasheet. Make updates in the (Carpenter 2011, Robeznieks 2010, Robeznieks 2011). datasheet. Service 2011, Standard & Poor's 2011). For a sample This allocation of capital spending is consistent with the o WATCH FOR GLITCHY RESETS WHEN UPDATE of nonprofit hospitals, Fitch Ratings found that capital YOU declines in DATA!!!!occupancy discussed earlier. inpatient expenditures ascolumn of total were added manually. o The a share totals revenue declined from 7.9 percent of revenues in 2008 to 6.6 percent in 2009 to 5.8 Quality of they kept resetting when I changed o I had to manually draw tick marks and axis lines because care: Overall, quality indicatorsany data. percent in 2010. Using a different methodology, Moody's show improvement o can't delete the legend, so I'll just have concluded Ithat in 2010 hospitals spent just slightly more to crop it out in InDesign. Our analysis of several inpatient quality-of-care indicators than would be necessary to maintain orto selecttheir for modification. Otherwise if you use the black selection tool, they o Use direct selection tool replace items shows generally positive trends. We use five of the inpatient existingdefault when you change the data. capital level of capacity: specifically, that median quality indicators (IQIs), developed and maintained by the spending declined from 1.6 times depreciation expenses Agency for Healthcare Research and Quality (AHRQ), to in 2008 o Usetimes depreciation in (and object styles) to format. to 1.2 paragraph styles 2009 to 1.1 times measure in-hospital and 30-day postdischarge mortality depreciation in 2010. If a hospital were to merely maintain o Data was from: R:\Groups\MGA\data bookrates (Agency for Healthcare Research and Quality 2007a). 2007\data book 2007 chp1 its existing capacity in a given year, the ratio would be We also analyze six of the AHRQ patient safety indicators approximately 1.0. Similarly, after reaching a peak of (PSIs), which measure the frequency of potentially $34 billion in 2008, spending on hospital construction Report to the Congress: Medicare Payment Policy | March 2012 53 preventable adverse events that can occur during an inpatient stay, such as the development of postoperative blood clots or deaths from treatable surgical complications (Agency for Healthcare Research and Quality 2007a, Agency for Healthcare Research and Quality 2007b). To assess sectorwide quality trends, we calculate risk-adjusted rates for these measures across all IPPS hospitals for a rolling fouryear period and determine whether there was a statistically significant change in each rate from the first year to the fourth year of the period. We use the IQIs and PSIs that AHRQ has concluded have the strongest base of clinical and statistical evidence (Agency for Healthcare Research and Quality 2009a). We calculate the IQIs and PSIs using MedPAR inpatient hospital data files for 2007 through 2010 and version 4.1b of the AHRQ mortality and PSI software (Agency for Healthcare Research and Quality 2009b). Most in-hospital and 30-day mortality rates declined In-hospital and 30-day postdischarge mortality rates, as measured by the AHRQ IQIs, declined by a statistically significant amount for four of the five conditions we monitor. From 2007 through 2010, risk-adjusted inhospital and 30-day mortality rates declined by a statistically significant amount for acute myocardial infarction, congestive heart failure, stroke, and pneumonia as measured by the AHRQ methods. The in-hospital and 30-day mortality rate for patients admitted with hip fracture also declined but not by a statistically significant amount. Patient safety indicators improved Rates improved from 2007 to 2010 for five of the six PSIs we analyzed, including iatrogenic pneumothorax, postoperative respiratory failure, postoperative pulmonary embolism or deep-vein thrombosis, postoperative wound dehiscence, and accidental puncture or laceration. The PSI that did not improve from 2007 to 2010 was the rate of deaths among surgical inpatients with treatable serious complications. Caution should be used in interpreting all the reported PSI rates. The PSIs measure rates of very rare events, and it is difficult to detect statistically significant changes in these indicators. In addition, AHRQ and other researchers have found that changes over time in providers' coding practices and variations among providers in how patient safety events are captured and reported can affect the accuracy and reliability of some of the PSIs (Agency for Healthcare Research and Quality 2007a, Agency for Healthcare Research and Quality 2007b, Agency for Healthcare Research and Quality 54 2009a, Rosen et al. 2012). Nonetheless, we monitor sector-level trends in selected PSIs as indicators, though not definitive evidence, of increases and decreases in rates of harm to patients resulting from their medical care that can be avoided if providers adhere to known clinical safety practices. In this light, the recent decreases in several of these indicators are encouraging, particularly given recent evidence that, overall, hospitals treating Medicare beneficiaries have significant room for improvement in patient safety (Landrigan et al. 2010). Readmission rates In 2010, CMS reported on the Hospital Compare website that the medians for hospitals' 30-day readmission rates were 18 percent for pneumonia, 20 percent for acute myocardial infarction, and 25 percent for heart failure (Department of Health and Human Services 2010). Those rates have not changed significantly over the past five years (Centers for Medicare & Medicaid Services 2010). However, the literature suggests that financial incentives can induce changes in quality and that progress can be made with readmissions (Jha et al. 2010). To stimulate greater improvement in readmission rates, the Commission recommended that a financial penalty be placed on hospitals with high readmission rates, and the Congress enacted a financial penalty for hospitals with aboveaverage risk-adjusted rates of readmissions. CMS will begin to apply the penalty in fiscal year 2013 (see text box, p. 57, for details). Value-based incentive payments As mandated by the Patient Protection and Affordable Care Act of 2010 (PPACA), CMS released final regulations in 2011 for the hospital value-based purchasing (VBP) program, which will start in fiscal year 2013. For the first year of the VBP program, CMS will reduce all DRG payments to about 3,100 participating hospitals by 1.0 percent to create the pool of funds from which value-based (i.e., performancebased) incentive payments will be made. CMS estimates that this payment adjustment will total $850 million in fiscal year 2013. As required by law, the VBP program must be budget neutral, meaning that the total amount of withheld payments must be redistributed to hospitals participating in the VBP program. In 2013, each hospital's performance score will be based on 12 process measures and 1 patient experience measure (Hospital Consumer Assessment of Healthcare Providers and Systems); in fiscal year 2014, CMS will add three outcome measures (condition-specific mortality Hospital inpatient and outpatient services: Assessing payment adequacy and updating payments rates) to the VBP program. The agency had proposed to also add AHRQ PSI and IQI composite measures, several hospital-acquired condition measures, and a per beneficiary spending measure but decided to drop those measures at least for fiscal year 2014 (Centers for Medicare & Medicaid Services 2011). Our hospital update recommendations below apply to hospital inpatient and outpatient payments. Payments for the other distinct units of the hospital, such as SNFs, are addressed by our update recommendations for those payment systems, which apply to both hospital-based and freestanding providers. In 2008, the Commission suggested measures that should be included in the hospital VBP program, including a robust set of patient safety measures and risk-adjusted outcome measures, such as mortality rates and efficiency measures (Medicare Payment Advisory Commission 2008a). The measures used in the VBP program, and the weighting that different measure domains contribute to a hospital's performance score, should evolve to reflect the program's quality improvement priorities. This progression would involve giving more weight to patient safety and outcome measures. We also have some technical concerns about the measures proposed (see text box, p. 57). Rise in payments per discharge from 2008 to 2010 was partly due to documentation and coding changes Medicare payments and providers' costs In assessing payment adequacy, the Commission also considers the estimated relationship between Medicare payments for and hospitals' costs of furnishing care to Medicare patients. We assess the adequacy of Medicare payments for the hospital as a whole, and thus our primary indicator of the relationship between payments and costs is the overall Medicare margin. This margin includes all payments and Medicare-allowable costs attributable to Medicare patients for the services hospitals provide plus graduate medical education payments and costs. We report the overall Medicare margin across service lines because no hospital service is a purely independent business. For example, we find that operating a skilled nursing facility (SNF) improves the profitability of acute inpatient care services when an in-hospital SNF allows hospitals to safely discharge patients sooner from their acute care beds, thus reducing the cost of the inpatient stay. In addition, the precise allocation of costs presents challenges. For example, under current cost accounting rules, hospitals may allocate too much of their administrative costs to a home health subsidiary, which can distort the apparent profit margins of both the home health agency and the hospital's other service lines. By combining data for all major services, we can estimate Medicare margins without the influence of how overhead costs are allocated. Growth in Medicare hospital payments per discharge under the IPPS depends primarily on three factors: (1) annual payment updates, (2) changes in reported case mix, and (3) policy changes that are not implemented in a budget-neutral manner. In 2010, IPPS hospitals received a 2.1 percent payment update to operating rates and a 1.4 percent update to capital rates. Inpatient payments per case, however, increased 2.5 percent, about 0.5 percentage point more than the update. Per case payments increased faster than the update in 2010 primarily due to increases in reported case mix. Growth in reported case mix was an even bigger factor in the high per case payment increases in 2008 and 2009, when the reported case-mix index (CMI) increased 2.0 percent and 2.6 percent, respectively (Figure 3-4, p. 56). Much of the increase in reported case mix that occurred from 2008 through 2010 was due to the diagnosis documentation and coding changes hospitals made after adoption of the new MS-DRGs in 2008. Implementation of MS-DRGs in 2008 gave hospitals an incentive to change diagnosis documentation and coding to more fully account for each patient's severity of illness. While documentation and coding changes help hospitals measure patient severity more accurately, they also increase the CMI and payments without real increases in patient severity or the resources hospitals must use to furnish inpatient care. The large increase in the CMI (2.0 percent, 2.6 percent, and 0.5 percent, respectively) that occurred in the 3 years after implementation of MS-DRGs in 2008 followed a decade in which the CMI declined in 5 of the 10 years and never grew by more than 1 percent in any one year (Figure 3-4). Analyses by both CMS and the Commission have concluded that the increases in case mix reported from 2008 through 2010 resulted from hospitals' documentation and coding rather than from an actual shift toward patients whose care required greater resources (Medicare Payment Advisory Commission 2010a). This finding explains how Report to the Congress: Medicare Payment Policy | March 2012 55 FIGURE 3-4 FIGURE 3-4 Comparison of growth.... Changes in Medicare payments, costs, and case mix, 1998-2010 10 Cost per case Case-mix index Payments per case 9 Annual percent change 8 7 6 Implementation of MS-DRGs 5 4 3 2 1 0 -1 -2 1998 2000 2002 2004 2006 2008 2010 Year Note: MS-DRG (Medicare severity-diagnosis related group). Changes in case mix are based on national aggregate case-mix indexes calculated for the cohorts of hospitals included in the inpatient prospective payment system (IPPS) in each pair of years. Case-mix index is computed for each year's inpatient claims using the Medicare DRG grouper and weights in place for that year. Source: MedPAC analysis of Medicare cost reports and annual MedPAR claims for IPPS hospitals for fiscal years 1997-2010 from CMS. Note: Note and Source in InDesign. 1998 and less than half any rate since 2001. Growth in outpatient costs also slowed, increasing only 0.1 percent per service unit in 2010 (Table 3-3). The lower cost growth in 2009 and 2010 was partly due Payments per case to lower input price inflation facing hospitals; the 2.1 percentCase-mix index was the lowest rate of increase in increase in 2010 input prices in more than a decade. The slower growth in hospital input prices reflects lower general economy-wide Cost per case inflation for goods and services and slower wage growth in the economy and the hospital industry. Compensation costs for hospital workers, for example, increased on average 2.0 percent in 2010, the smallest increase in more than a decade (Bureau of Labor Statistics, http://www. bls.gov/web/eci/echistrynaics.pdf). Hospitals may also have worked to control cost growth in response to the recession and difficult year they had financially in 2008, Inpatient when the industry experienced historically low total all- 2003 payer margins (1.8 percent) and had steep declines in their Cumulative % change 0 balance sheets. 2004 0.612 Outpatient (from sheet 2009 outpatient) Lower cost growth also Volume per beneficiary Cumulative % change could be the result of a less complex mix of patients, as the overall mix of services for both Medicare inpatient and outpatient services declined. Although the reported inpatient CMI increased, after accounting for documentation changes, inpatient case mix declined slightly as some high-cost surgical services shifted from the inpatient setting to outpatient settings. Outpatient service mix also declined as physician office visits, a relatively inexpensive service, became a larger share of overall outpatient services, resulting in cost growth per service of 0.1 percent. hospitals could record high case-mix growth from 2008 to 2010 without a corresponding increase in cost growth (Figure 3-4). In fact, the rate of cost growth declined in 2009 and 2010 for the reasons discussed. We estimate that Notes about this graph: documentation and coding improvements led to more than $6 billion of o Data is payments in 2008 and 2009;updates in the datasheet. additional in the datasheet. Make CMS has been recovering these overpayments initems and the x-axis. T A B Lwill reflow if I update the data. o I had to force return the 2011 on 2012. They E Cost growth slowed in 2010 However, at least another $11 billion in overpayments 3 -3 o I had to manually draw tick marks have been accumulating in 2010, 2011, and 2012 that and axis lines because they kept resetting when I changed any data Annual cost growth CMS cannoto Use direct selection tool to select items for modification. Otherwise if you use the black selection tool, recover because of a lack of authority under current law. (For a more detailed description of Cost measure 2007 2008 2009 2010 default when you change the data. this issue, see the Commission's comment letter on the o Use paragraph styles (and object styles) toInpatient costs per discharge 4.3% 5.5% 2.9% 2.0% format. 2012 proposed rule, June 17, 2011, at http://medpac.gov/ Outpatient costs per service 5.6 5.1 4.8 0.1* documents/06172011_FY12IPPS_MedPAC_COMMENT. Weighted average 4.5 5.4 3.3 1.6 pdf.) Input price inflation Hospital cost increases fell to their lowest level in a decade in 2010 A combination of low input price inflation and financial pressure on hospitals resulted in a continued slowing of hospital cost growth in 2010. Medicare inpatient costs per case rose only 2.0 percent in 2010, down from 2.9 percent in 2009. This rate is the slowest rate of increase since 56 3.4 4.3 2.6 2.1 Note: Cost growth numbers are not adjusted for reported changes in case mix. Analysis excludes critical access hospitals and Maryland hospitals. The weighted average is based on services provided to Medicare patients in hospitals, including costs for inpatient, outpatient, skilled nursing facility, inpatient rehabilitation, and home health services. *Cost growth was 1.7 percent if adjusted for complexity of services provided. Source: MedPAC analysis of Medicare Cost Report and claims files from CMS. Hospital inpatient and outpatient services: Assessing payment adequacy and updating payments Mortality and readmission measures: Considerations and challenges M ortality and readmissions are outcomes of particular importance to Medicare beneficiaries, providers, and policymakers. Commission staff recently convened two expert panels on these outcome measures: the first, to understand how providers, commercial health plans, and other payers use mortality and readmission measures; the second, to understand the technical properties of specific measures. The predominant view from the first panel was that providers and payers place great value on these risk-adjusted outcome measures and use them to motivate change within their organizations to improve quality. For example, several panelists reported using mortality measures to focus on specific clinical conditions or hospital units with high mortality, guide implementation of corrective actions, and improve performance over time. They also saw value in examining trends in outcomes without risk adjustment to confirm that risk-adjusted outcome trends are not being driven by coding. However, mortality measurement is complicated by the need to identify patients entering hospitals for palliative care or in anticipation of death. Panel members noted that donot-resuscitate orders are not a sufficient indicator of patients' objectives for entering hospitals given that these orders are often issued well into a hospital stay. When examining readmission metrics, the main challenge for the hospital systems was a lack of data on patients who were readmitted to hospitals outside their own system. The second panel discussed the statistical question of how to make reliable estimates for hospitals with a small number of cases. CMS's approach uses a "random effects" method in which the estimated mortality rates and readmission rates are blended toward the national mean before being reported on the Medicare Hospital Compare website. Ideally, the Trend in the overall Medicare margin We define Medicare margins as Medicare payments minus the allowable costs of treating Medicare patients, divided by Medicare payments. In analyzing hospital margins, we exclude CAHs, which are 1,300 rural hospitals paid based on their incurred costs. We also exclude hospitals goal is for observed differences in rates to represent real differences in outcomes and not be subject to random statistical variation from a small number of observations. To minimize the chance of categorizing a hospital as a poor or good performer due to random variation, CMS presents data for each hospital that blends the experience of the subject hospital and the average experience for all hospitals in the country. The smaller the hospital, the less its actual performance information is used and the more the national average is used. "In essence, the predicted mortality rate for a hospital with a small number of cases is moved toward the overall U.S. national mortality rate for all hospitals" (Centers for Medicare & Medicaid Services 2011). For all six measures reported on Hospital Compare, more than 90 percent of hospitals are reported to be indistinguishable from the national average. For example, readmission rates for more than 97 percent of hospitals are reported as "no different than the U.S. national rate" for acute myocardial infarction readmissions. As a result, beneficiaries have little useful information on hospital performance and hospitals have little information on where they stand relative to other hospitals and where they could improve. Most panel members agreed that CMS's measures underestimate differences among groups of providers when true differences exist. For that reason, they concluded that the Hospital Compare data should not be used as an input into research studies that compare groups of hospitals. When groups of data are being evaluated, the number of observations is large enough to let the data stand on their own rather than blending the data with national average data. We concur and use metrics developed by the Agency for Healthcare Research and Quality and 3M Health Information Systems for measuring mortality and readmission rates when comparing groups of providers. Questions remain about what methods would be best for reporting an individual hospital's performance. ? in Maryland, which are excluded from the IPPS and paid under a state-wide all-payer prospective payment system. The overall Medicare margin trended downward from 1997 through 2008 and has been negative since 2003 (Figure 3-5, p. 58).5 From 2008 to 2010, however, the overall Medicare margin went up from -7.1 percent in Report to the Congress: Medicare Payment Policy | March 2012 57 FIGURE FIGURE 3-5 3-5 Overall Medicare and MedicareMedicare margins: Hospital inpatient margins Inpatient, outpatient, and overall 15 Inpatient Overall Medicare 10 Margin (in percent) Outpatient 5 0 -5 -10 -15 -20 2000 2002 2004 2006 2008 2010 the -4.8 percent margin for urban hospitals (Table 3-4). Overall Medicare margins at for-profit hospitals remained above those at nonprofit hospitals. In 2010, for-profit hospitals' overall Medicare margins were 0.1 percent compared with -5.7 percent at nonprofit hospitals. ForOutpatient margin profit hospitals also had positive inpatient margins (1.3 percent) and positive outpatient margins (0.1 percent) in Overall Medicare margin 2010 (not shown). Inpatient margin In 2010, the overall Medicare margin was -0.2 percent for major teaching hospitals, increasing from a low point of -1.9 percent in 2008. Major teaching hospitals have higher overall Medicare margins than the average IPPS hospital in large part due to the extra inpatient payments they receive through the indirect medical education and disproportionate share adjustments in the IPPS. A Commission analysis shows that both of these adjustments provide payments that are substantially larger than the estimated effects that teaching intensity and service to Year Note: A margin is calculated as payments minus costs, divided by payments; margins are based on Medicare-allowable costs. Analysis excludes critical access and Maryland hospitals. Medicare inpatient margins include services covered by the acute inpatient prospective payment system. Overall Medicare margin includes acute inpatient, outpatient, hospitalbased home health and skilled nursing facility (including swing bed), and inpatient psychiatric and rehabilitation services, plus graduate medical education. TABLE Overall Medicare margins by hospital group 3 -4 Hospital group 2006 2007 2008 2009 2010 All hospitals -4.6% -6.0% -7.1% -5.1% -4.5% Urban Note: Note and Source in InDesign. Source: MedPAC analysis of Medicare Cost Report file from CMS. -4.7 -6.1 -7.3 -5.2 -4.8 Rural 2008 to -4.5 percent in 2010. The overall Medicare margin Excluding CAHs -4.4 -5.1 -6.0 -4.4 -2.6 Including CAHs -3.3 -3.9 -4.4 -3.3 -1.7 is dominated by inpatient and outpatient services, which Notes about this graph: represent 92 percent of hospitals' Medicare revenues. Nonprofit -5.4 -6.7 -8.2 -6.3 -5.7 o and outpatient datasheet. Make updates in For profit the datasheet. -2.4 Both inpatient Data is in themargins improved in 2010, -3.5 -2.6 -0.1 0.1 although both remained negative. Betweenitems and the x-axis. They will reflow if I update the data. N/A Government* N/A N/A N/A N/A o I had to force return the 2008 on 2010, the margin for Medicare inpatient services rose from -4.8 o I had to manually draw tick marks and axis lines because they kept resetting when I changed any data Major teaching 2.2 0.1 -1.9 -0.5 -0.2 percent to -1.7 percent, and the margin for Medicare Other teaching -5.1 -6.3 -7.4 -5.1 -4.5 o Use direct from -12.7 percent to -9.6 outpatient services went upselection tool to select items for modification. Otherwise if you use the black selection tool, t Nonteaching -8.2 -9.2 -10.0 -7.8 -7.0 percent (Figure 3-5).when you change the and overall default The increase in inpatient data. margins in 2009 was due primarily to increases in reported toNote: CAH (critical access hospital), N/A (not applicable). Data are for all o Use paragraph styles (and object styles) format. case mix; the increase in 2010 was due primarily to lower hospitals covered by the Medicare acute inpatient prospective payment system in 2010 and CAHs where indicated. A margin is calculated cost growth and continued increases in reported case mix as payments minus costs, divided by payments; margins are based on for inpatients and increases in the volume of outpatient Medicare-allowable costs. Overall Medicare margin covers acute inpatient, outpatient, hospital-based skilled nursing facility (including swing bed), services.6 Outpatient margins improved as a result of cost home health, and inpatient psychiatric and rehabilitation services, plus growth being lower than the hospital update in 2010. graduate medical education. The rural margins are shown with and without 2010 Medicare margins by hospital type We further examined the overall aggregate Medicare margin by hospital type. In 2010, the -2.6 percent overall Medicare margin for rural PPS hospitals was higher than 58 CAHs. The margins without CAHs illustrate the profitability of rural inpatient prospective payment system hospitals; the rural margins with CAHs give a fuller picture of rural hospital profitability. *Government-owned providers operate in a different context from other providers, so their margins are not necessarily comparable. Source: MedPAC analysis of Medicare Cost Report file, MedPAR, and impact file from CMS. Hospital inpatient and outpatient services: Assessing payment adequacy and updating payments Historically, other hospital-based units--SNFs, home health agencies, inpatient rehabilitation facility units, and inpatient psychiatric units--have had lower Medicare margins than their freestanding counterparts. However, hospitals with these units have higher overall Medicare margins than hospitals without them. In aggregate, hospitals with some type of post-acute care unit in 2010 had higher overall Medicare margins than hospitals that had no post-acute units, -4.0 percent compared with -7.4 percent. The higher margins for hospitals with post-acute providers could in part reflect the ability of hospitals with an in-hospital SNF or inpatient rehabilitation facility to discharge their patients quicker and improve their inpatient margins. For example, in 2010, the overall Medicare margin for hospitals with a SNF unit was -3.9 percent compared with -4.6 percent for hospitals without a SNF unit--despite the average -67.0 percent margin for hospital-based SNFs. A Commission analysis has shown that hospitals are able to cover their total direct costs for patients who use both inpatient and SNF services. The effect that one service line can have on another service line is the reason we examine hospitals' overall Medicare margins rather than focusing on the profitability of each service line. Cycles of industry-wide financial pressure and cost growth The level of hospitals' cost growth has cycled up and down through four time periods (Figure 3-6). During the first time period (1988-1992), most insurers paid hospitals on the basis of their charges, with little price negotiation or selective contracting. With limited pressure from private payers, hospital margins on private-payer business increased rapidly. In the second cycle (1993-1999), HMOs and other private insurers began to negotiate more assertively with hospitals, and most insurers switched to paying for inpatient services on the basis of DRGs or flat per diem amounts for broad types of services. Because managed care restrained private-payer payment rates, hospitals were under pressure to constrain their costs and FIGURE FIGURE 3-6 3 -6 Costs have risen faster than the market basket in recent 2010 Cost growth falls in 2009 and years as financial pressure increases 12 M Annual change in costs per discharge Market basket index 10 8 Percent change low-income patients have on hospitals' average costs per discharge. In June 2010, the Commission made recommendations to use teaching hospital payments as incentives to train physicians for the skill sets needed by future Medicare beneficiaries (Medicare Payment Advisory Commission 2010b). Nonteaching hospitals, most of which are in urban areas, had the lowest Medicare margins of any hospital group, -7.0 percent in 2010. 6 M 4 2 0 -2 -4 1988 1992 1996 2000 2004 2008 1990 1994 1998 2002 2006 2010 Fiscal year 1988 1989 2003 in costs 0.6 and CMS 1.1 2.6 Source: Medicare analysis of Medicare Cost Report files from CMS Market basket final rules for the inpatient prospective payment system in years 1988 index through 2010. 2.5 3.3 4.3 Note: The market basket index measures annual changes in the prices of the 2001 2002 goods and services hospitals use to deliver care. Cost growth refers to Medicare inpatient allowable costs per discharge. Annual change Note: 1 2 pe 2 4 3 Note and Source in InDesign. the rate of cost growth was below input price inflation from 1994 through 2000 (Figure 3-6). Notes about this graph: By 2000, hospitals had regained the upper hand in price negotiationso Data is in the datasheet. Make updates in the d because of hospital consolidations and consumer backlash against managed care. In the third o I had to force return the rates on cycle (2000-2007), private-payer paymentitems rose the x-axis. They rapidly. Becausehad to manually draw tickmargins and axis line o I of these high rates, all-payer marks for hospitals reached 6.0 percent in 2007 (Figure 3-7, p. o Use direct selection tool to select items for modi 60). Cost growth was high in 2008 (5.5 percent), as many default when you change the data. hospitals started the year with little pressure to constrain o Use picture changed rapidly in September costs. However, the paragraph styles (and object styles) to form 2008 with the collapse of the bond and stock markets. Total all-payer margins in 2008 fell to 1.8 percent, the lowest level in more than two decades. Operating margins fell, investment income declined dramatically, some defined benefit pension plans needed larger contributions from their hospital sponsors, and the economic outlook was uncertain. This situation created financial pressure to constrain costs. In response, hospitals pulled back from the high levels of capital expenditures and employment Report to the Congress: Medicare Payment Policy | March 2012 59 FIGURE FIGURE 3-5 3- 7 Overall Medicare and Hospitals' financial performance Medicare inpatient margins has been improving after poor performance in 2008 12 Margin (in percent) 10 8 6 4 Total all-payer margin 2 Operating margin EBITDA margin 0 2005 2006 2007 2008 2009 2010 EBITDA (earnings before interest, taxes, depreciation, and amortization). A margin is calculated as revenue minus applicable costs, divided by payments. Analysis excludes critical access hospitals and Maryland hospitals. Source: MedPAC analysis of Medicare cost report data from CMS. Note: Note and Source in InDesign. Total inpatient payment rates grew by 1 percent from 2010 to 2012. This relatively low rate of payment growth is due to two adjustments that were made in 2011 and 2012 to correct EBITDA margin for overpayments associated with documentation and codingOperating First, CMS reduced payment rates by 2.9 changes. margin percent in 2011 and left this reduction in place in 2012 to recover overpayments that occurred in 2008 and 2009. Total all payer margin Second, CMS reduced payment rates by 2 percent in 2011 to limit future overpayments. These corrections for past overpayments almost fully offset the market basket-based update in hospital inpatient payment rates. Outpatient payments were not affected by the documentation and coding issues, and those payment rates increased by 4.2 percent from 2010 to 2012. Growth in inpatient payment rate slows We expect the rate of annual cost growth per discharge to increase to roughly 3 percent in 2011 and 2012. Two factors are expected to increase cost growth. First, hospitals' financial performance has rebounded as we discussed earlier. This factor could lead to weaker cost control. We see some evidence of higher cost growth from Census Bureau data through June 2010, from Bureau of Labor Statistics data on employment growth in 2011, and from data from publicly traded hospital systems through the third quarter of 2011. Second, the projected rate of input inflation is expected to rise from close to 2 percent to closer to 3 percent (IHS Global Insight 2011) . Hospital cost growth may increase Year Note: Margins projected to decline in 2012 growth seen in 2007 and 2008 to more moderate levels in 2009 and 2010. As capital and labor growth slowed, Because costs are growing faster than payment rates, cost growth Notes about the lowest level recorded slowed in 2010 to this graph: we project the overall Medicare margin to decline from in more than 10 years, reflecting both slowing input price -4.5 percent in 2010 to roughly -7 percent in 2012. o Data efforts to constrain cost growth. For the decline should growth and hospitals'is in the datasheet. Make updates in This datasheet. not be unexpected. The increase in the first time in 10 years,forcegrowth slowed to the rate of o I had to cost return the items on the x-axis. They will reflowto 2010 was largely due to hospitals' margins from 2008 if I update the data. input price inflation (Figure 3-6, p. 59). documentation and coding changes, and the expected o I had to manually draw tick marks and axis lines because they kept resetting when I changed any data decline in margins in 2011 and 2012 will reflect the Cost growtho Use direct selectionresponse select items for modification. Otherwise if you use the black selection tool, t may start to increase in tool to to the reduction in payment updates required to correct for these rebound in hospitals'whenall-payer margin in 2010, default total you change the data. documentation and coding changes. which reached 6.4 percent, the highest level in more than o Use paragraph styles (and object styles) to format. 20 years, as shown in Figure 3-7. The figure also shows a Hospital-level financial pressure and hospital costs corresponding increase in operating profits and operating The effect of financial pressure on hospitals' costs is not cash flow as measured by earnings before interest, taxes, only evident over time; it is also evident when comparing 7 depreciation, and amortization (EBITDA). This increase hospitals facing different levels of financial pressure to suggests that hospitals were under less financial pressure constrain costs. Some hospitals have strong profits on in 2011, and some indicators suggest that hospital hiring non-Medicare services and investments and are under and capital costs are increasing. Following a slowdown in little pressure to constrain their costs. Other hospitals, with hospital employment growth in 2009 and 2010, hospital thin profits on non-Medicare services, face overall losses employment in 2011 grew at 1.8 percent (Bureau of Labor (and possibly closure) if they do not constrain costs and Statistics 2011). 60 Hospital inpatient and outpatient services: Assessing payment adequacy and updating payments generate profits on Medicare patients. To determine the effect of financial pressure on costs, we grouped hospitals into three levels of financial pressure from private payers: high, medium, and low based on their non-Medicare profits (margins) and other factors from 2005 to 2009. For these years, the hospitals under high pressure had non-Medicare profits of less than 1 percent, while the low-pressure hospitals had non-Medicare margins of more than 5 percent. We found that hospitals under high pressure from 2005 to 2009 ended up with lower costs per discharge in 2010 than hospitals under low levels of financial pressure during the same five-year period. For more details on our analytic methods, see our prior year's analysis of payment adequacy (Medicare Payment Advisory Commission 2011c). Key findings from our analysis of financial pressure on hospitals are: o o o High pressure = low cost: The 25 percent of hospitals under the most financial pressure had median standardized costs per case that were roughly 10 percent lower than the national median for all 2,893 IPPS hospitals with available data. Because of their lower costs, hospitals under pressure generated a median overall Medicare profit margin of 5 percent, which is 10 percentage points above the national median. Low pressure = high cost: The 60 percent of hospitals that were under a low level of financial pressure had median standardized costs per case that were 4 percent above the national median. Because of higher costs, they generated a median Medicare profit margin of -9 percent, which is 4 percentage points below the national median. For profits have different incentives: For-profit hospitals tended to keep their median standardized costs per case at the national median even when they were under little financial pressure. This finding suggests that if both types of hospitals receive high payment rates from private payers, the higher revenues tend to result in higher costs in nonprofit hospitals, but in for-profit hospitals a larger share of the revenue is retained as operating profit for shareholders. The overarching conclusion is that costs are at least partially under hospitals' control, and those hospitals with the strongest cost control can generate profits treating Medicare patients. This conclusion has been supported in recent literature that finds hospitals in more competitive markets tend to control their costs more than hospitals in less competitive markets (Robinson 2011). The next question is whether some set of hospitals can have both low costs and high-quality outcomes. Relatively efficient hospitals The goal of this analysis is to examine payment adequacy for the group of hospitals that perform relatively well on both cost and quality metrics while serving a broad spectrum of patients. The variables we use to identify relatively efficient hospitals are hospital-level mortality rates (AHRQ IQIs), readmission rates (3M Health Information Systems potentially preventable readmissions), standardized inpatient costs per case, providers' payer mix, and the annual level of total FFS Medicare service use per capita in the county where the hospital is located. As data and risk-adjustment methodologies improve, our measures of efficiency will continue to evolve. Our assessment of efficiency is not in absolute terms but rather relative to other IPPS hospitals. Ideally, we would limit our set of efficient hospitals to those that not only had high in-hospital quality and low unit costs but also low overall costs to the Medicare system during the year. To avoid having hospitals from high-use areas in our analysis, we removed hospitals from the population studied if they were in counties in the top 10 percent of annual Medicare FFS service use per FFS beneficiary.8 This method reduces the chance that a hospital will appear to have low unit costs of service simply because it is in an area with a high volume of admissions of low-cost patients who could be treated on an outpatient basis. We further restricted the population of hospitals that we evaluated for efficiency by removing the 10 percent of hospitals with the smallest shares of Medicaid patients. This process reduces the likelihood of including hospitals in our efficient group simply because they had a favorable selection of patients. Our goal in this screening process is to improve our ability to identify hospitals that can provide good outcomes at a reasonable cost while serving a broad spectrum of patients (including Medicaid) without driving up the overall volume of hospital and nonhospital services provided. We assigned hospitals to the relatively efficient group or the control group according to each hospital's performance on a set of risk-adjusted cost and quality metrics during the period 2007-2009. We then examined the performance of the two hospital groups in fiscal year 2010. Categorizing hospitals as relatively efficient Report to the Congress: Medicare Payment Policy | March 2012 61 TABLE 3-5 Performance of relatively efficient hospitals Type of hospital Relatively efficient during 2007-2009 Number of hospitals Share of hospitals 188 Other hospitals 1,943 9% 91% Composite 30-day mortality (AHRQ) 82% 102% Readmission rates (3M Health Information Systems) 96 100 Standardized cost per discharge 91 102 Composite 30-day mortality (AHRQ) 83% 101% Composite 30-day readmission (3M Health Information Systems) 95 101 Standardized cost per discharge 89 102 103 98 Historical performance, 2007-2009 Relative risk-adjusted: Performance metrics, 2010 Relative risk-adjusted: Relative percent of patients highly satisfied, 2010 (H-CAHPS(R)) Median: Overall Medicare margin, 2010 4% Non-Medicare margin, 2010 6 8 Total (all payer) margin, 2010 5 4 Note: -5% AHRQ (Agency for Healthcare Research and Quality), H-CAHPS(R) (Hospital Consumer Assessment of Healthcare Providers and Systems). Relative percents are the median for the group as a percentage of the median of all hospitals. Per case costs are standardized for area wage rates, case-mix severity, prevalence of outlier and transfer cases, interest expense, low-income shares, and teaching intensity. Composite mortality was computed using the AHRQ methodology to compute riskadjusted mortality for six conditions (acute myocardial infarction, congestive heart failure, pneumonia, gastrointestinal hemorrhage, stroke, and hip fracture). We then weighted the scores for each type of discharge by the share of discharges in that particular hospital. We removed hospitals with low Medicaid patient loads (the bottom 10 percent of hospitals) and hospitals in markets with high service use (top 10 percent of hospitals) due to concerns that socioeconomic conditions and aggressive treatment patterns can influence unit costs and outcomes. Source: MedPAC analysis of impact file, MedPAR, and Medicare cost report data from CMS, and CMS Hospital Compare data. Hospitals were identified as relatively efficient if they met four criteria every year of the 2007 to 2009 period: o Risk-adjusted mortality levels were in the best twothirds. o Risk-adjusted readmission rates were in the best two-thirds. o o 62 Standardized costs per discharge were in the best two-thirds. Risk-adjusted mortality or standardized costs per discharge were in the best one-third. The objective is to identify hospitals that consistently performed at an above-average level on at least one measure (cost or quality) and that always performed reasonably well on all measures. The rationale for this methodology is discussed in detail in our March 2010 report (Medicare Payment Advisory Commission 2010b). Examining performance of relatively efficient and other Of the 2,131 hospitals that met our screening criteria, 188, or about 9 percent, were found to be relatively efficient during the 2007-2009 period. The set of relatively efficient providers was a diverse array of hospitals, including large teaching hospitals and smaller rural hospitals. CAHs were excluded hospitals from 2007 to 2009 Hospital inpatient and outpatient services: Assessing payment adequacy and updating payments from the analysis because they are not paid under the IPPS and have different cost accounting rules. We examined the performance of relatively efficient hospitals for 2007-2009 on three measures by reporting the group's median performance divided by the median for the set of 2,131 hospitals in our analysis (Table 3-5). The median efficient hospital's relative risk-adjusted 30-day mortality rate from 2007 through 2009 was 82 percent of the national median, meaning that the 30-day mortality rate for the efficient group was 18 percent below the national median. The median readmission rate for the efficient group was 4 percent below the national median. Standardized cost per discharge for the efficient group was 9 percent below the national median. The group of efficient hospitals tends to be larger than average but otherwise had diverse characteristics. For a more complete description, see our March 2011 report (Medicare Payment Advisory Commission 2011b). Historically strong performers had lower mortality and The composite mortality level for the efficient group was 17 percent below the national median in 2010. In addition, the risk-adjusted 30-day readmission rate was 5 percent lower in the efficient group. The efficient group also performed slightly better than other hospitals on patient satisfaction. The share of patients who gave the median hospital a top rating in 2010 was 3 percent higher than the national median (69 percent) for the efficient group and 2 percent lower than the median (66 percent) for the comparison group. readmissions in 2010 Historically strong performers continue to have lower Hospitals that were low-cost and lowmortality providers from 2007 through 2009 continued to have lower costs in 2010. The median standardized Medicare cost per discharge in the efficient group was 11 percent lower than the national median, compared with 2 percent higher for the other group. The lower costs allowed the relatively efficient hospitals to generate higher overall Medicare margins. The median hospital in the efficient group had an overall Medicare margin of 4 percent, while the median hospital in the comparison group had an overall Medicare margin of -5 percent. Among the relatively efficient hospitals, 65 percent had positive Medicare margins compared with 35 percent for other hospitals. The distribution for the efficient hospitals ranged from -2 percent to 9 percent at the 25th and 75th percentiles, respectively. For the comparison group, the distribution of Medicare margins ranged from costs in 2010 -15 percent to 4 percent at the 25th and 75th percentiles, respectively. Among the relatively efficient hospitals 50 percent were under high or medium financial pressure to constrain their costs compared with 40 percent for the other hospitals. This result suggests that some of the efficient hospitals may have been pressured to constrain their inpatient costs, while others made the choice to constrain their costs to generate financial reserves for the future. Rural hospital payments and costs PPACA requires that the Commission analyze the adequacy of Medicare payments to rural providers as part of a larger report on rural health care. To prepare for that larger rural report, we present additional data on rural hospital payments and costs. The key question is whether Medicare payment rates are inappropriately low (or inappropriately high) in rural areas relative to urban areas. If rural payment rates are too low or too high relative to urban areas, the current set of special rural payments may need to be adjusted. The Commission conducted a similar review of rural payment adequacy in 2001 as part of a larger report on rural health care. In 2001 the Commission noted that rural PPS hospitals' inpatient margins were lower than urban PPS inpatient margins and the gap had increased from less than 1 percent in 1992 to more than 10 percent in 1999 (Medicare Payment Advisory Commission 2001).9 As a result of the 2001 report, the Commission made a series of payment recommendations including raising the base payment rate for rural providers up to the urban level, increasing disproportionate share (DSH) payments by moving closer to the formula used for urban hospitals, and introducing a low-volume adjustment for hospitals with few total discharges. The Congress enacted payment changes that were similar to these policy recommendations. The Congress also enacted several additional increases in rural payments including adjustments to the wage index, enhancing the sole community hospital (SCH) program, enhancing the Medicare-dependent hospital program (MDH), and adding a more generous low-volume adjustment, which we will discuss later. The SCH and MDH programs pay hospitals based on their historical costs of providing inpatient care updated for inflation, if those payments are higher than standard IPPS rates.10 Report to the Congress: Medicare Payment Policy | March 2012 63 FIGURE 3-8 FIGURE 3- 8 Rural-urban margin gap... Rural-urban margin gap was closed by 2004 20 Rural inpatient margin Margin (in percent) 15 Urban inpatient margin 10 5 0 -5 2001 MedPAC rural report Low-volume adjustments became much more generous in 2011 In our 2001 rural report, the Commission recommended that the Congress direct the Secretary to create a lowUrban outpatient volume adjustment formargin that are more than a hospitals specified distance from other facilities. The Congress Rural inpatient margin enacted a low-volume adjustment in 2003 and, as the Commission recommended, left implementation up to the Secretary. The Secretary then determined that only hospitals with fewer than 200 total discharges and that are more than 25 miles from another hospital warrant a low-volume adjustment. Because many of the smallest hospitals have elected CAH status, the low-volume adjustment applied to fewer than 10 IPPS hospitals in 2010. In 2010, the Congress enacted a new, more generous, lowvolume adjustment for IPPS hospitals. Rather than leave 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 the eligibility criteria up to the Secretary, the Congress Fiscal year mandated that inpatient payments increase for any hospital with fewer than 1,600 Medicare discharges and that is 15 Source: MedPAC analysis of Medicare cost reports. or more miles from another IPPS hospital. In practice, the program is not focused on isolated hospitals because hospitals eligible for the low-volume adjustment can be any distance from CAHs. The adjustment increases payments to IPPS hospitals with 200 or fewer Medicare The payment changes have closed the gap between rural Note: Note and Source in InDesign. discharges by 25 percent; the adjustment decreases and urban Medicare hospital margins, and rural hospitals linearly until it phases out for hospitals with 1,600 or now have Medicare inpatient margins that are higher than more Medicare discharges. For example, a hospital with urban margins by 2.6 percentage points (Figure 3-8). 200 Medicare discharges receives a 25 percent add-on; Rural hospital overall Medicare margins, which combine with 900 Medicare discharges, a 12.5 percent add-on; and revenues and costs for inpatient, outpatient, and postwith 1,600 Medicare discharges, no add-on. In 2011, 529 Notes about this graph: acute care services, have also improved relative to urban hospitals received a low-volume adjustment (Centers for o hospitals receive special hold-harmless the datasheet. margins. RuralData is in the datasheet. Make updates in Medicare & Medicaid Services 2011). This adjustment payments for outpatient services, which maintainon the x-axis. They will reflow if I update the data. o I had to force return the items a floor raises several issues: on the profitability of outpatient services equivalent to o I had to manually draw tick marks and axis lines because they kept resetting when I the low- any data changed o The empirical support for the magnitude of the margin in 1998, before implementation of the OPPS o Use SCHs selection percent add-on volume adjustment is unclear; use the black larger system. In addition, direct receive a 7 tool to select items for modification. Otherwise if youthe adjustment is selection tool, t default when you change net result than past estimates of the effect of volume on inpatient payment to their outpatient payments. Thethe data.of the costs special inpatient and outpatient payments is that overall o Use paragraph styles (and object styles) to format. per discharge. Medicare margins for rural hospitals are now higher than o The adjustment is added on top of SCH and MDH margins for urban hospitals, and margins are higher as cost-based payments, both of which increase hospitals become more rural. In 2010, urban hospitals had payments based on a hospital's historical costs and an aggregate overall Medicare margin of -4.8 percent, reflect any impact of historically low volume on its compared with margins of -3.4 percent for rural hospitals base-year costs per case. Therefore, a hospital can in micropolitan areas, -0.9 percent for rural hospitals in be paid its historical costs, plus inflation, plus a lowareas adjacent to urban areas, and 0.8 percent for rural volume adjustment of up to 25 percent. hospitals that are in the most rural areas (Table 3-6). In contrast to Medicare margins, total (all-payer) margins o The adjustment is not well targeted. It is based on tend to be higher for urban providers. Medicare discharges rather than total discharges. -10 64 Hospital inpatient and outpatient services: Assessing payment adequacy and updating payments TABLE 3-6 Does payment adequacy in 2010 differ between rural and urban areas? Urban Number of IPPS hospitals Micropolitan 2,264 Rural adjacent to urban Rural nonadjacent 130 587 185 Overall Medicare margin -4.8% -3.4% -0.9% Inpatient Medicare margin -2.0 -0.6 4.4 4.7 6.4 6.3 0.7 3.9 Total (all payer) margin Note: 0.8% IPPS (inpatient prospective payment system). Urban is a county in a metropolitan statistical area (MSA). Micropolitan areas refer to counties that are associated with a city of over 10,000 people but are located outside of an MSA. Rural adjacent areas are in counties without a city of 10,000 people but are adjacent to an MSA. Rural nonadjacent counties are not located next to an MSA and do not have a city of 10,000 or more people. We did not report frontier counties separately because they include only 26 IPPS hospitals. The rural IPPS margins do not include data for critical access hospitals, which receive 1 percent above costs. Source: MedPAC analysis of Medicare cost report files. Economies of scale depend on total discharges (not just Medicare discharges), so the adjustment has a weaker connection to a provider's economies of scale problem than an adjustment based on total discharges. Basing the adjustment on Medicare discharges also discriminates in favor of hospitals with large numbers of private-payer patients and against hospitals with larger shares of Medicare discharges (Table 3-7). Table 3-7 shows Medicare, non-Medicare, and total discharges for two hospitals and simulates how the lowvolume adjustment would have affected those hospitals in 2011. In this simulation, both hospitals have 2,200 total discharges and therefore might be expected to have similar problems related to economies of scale. Hospital A, with a 70 percent Medicare share, receives only a 1 percent low-volume add-on because it has almost 1,600 Medicare discharges (the upper limit for the Medicare low-volume adjustment). But hospital B receives a 17 percent add-on because it has a relatively small share of Medicare patients. The Commission's analysis for the 2001 rural report and current analysis using 2010 payments and costs TABLE 3-7 found higher unit costs per case-mix-adjusted discharge for hospitals with up to 500 total discharges. A policy that applied to hospitals with fewer than 500 total discharges would benefit roughly 200 hospitals, compared with the more than 500 hospitals that benefit from the current lowvolume policy. A key question is the degree to which the current lowvolume adjustment biases payment in favor of lowvolume hospitals by paying more than the estimated effect of low volume on costs per discharge. Table 3-8 (p. 66) shows the 2010 Medicare inpatient margins for rural hospitals, dividing the hospitals into quintiles based on their total patient volume. We report actual 2010 margins and simulated margins as if the 2011 lowvolume adjustment had been in effect. The lowest volume rural hospitals (quintile 1) tend to have higher inpatient margins (without the 2011 low-volume adjustment) than the next two quintiles of rural hospitals. This result is in part due to special payments many of them receive under the SCH and MDH programs. An additional low-volume adjustment will exacerbate those differences in 2011 and Low-volume policy favors hospitals with larger non-Medicare shares, 2011 Type of hospital Hospital A: high Medicare share (70%) Hospital B: low Medicare share (30%) Medicare discharges Non-Medicare discharges Total discharges Low-volume adjustment 1,550 650 2,200 1% increase 650 1,550 2,200 17% increase Source: MedPAC analysis of CMS data. Report to the Congress: Medicare Payment Policy | March 2012 65 TABLE 3-8 Estimated effect of the new low-volume adjustment Medicare inpatient margins Total (all-payer) volume of discharges Rural: Actual 2010 Rural: Simulated with low-volume adjustment* Lowest quintile 0.8% Second quintile 0.1 9.4 Third quintile -2.9 2.4 Fourth quintile 0.1 0.7 Highest quintile 1.6 1.6 All hospitals 0.6 2.8 Note: 14.0% *The margin with 2011 low-volume adjustment is a simulated margin where payments are adjusted to what they would have been if the low-volume adjustment had been in effect in 2010. The cut points for the volume quintiles for rural hospitals are 1,349; 2,145; 3,291; and 5,124 total discharges. Source: MedPAC analysis of Medicare cost report files. The margin with the low-volume adjustment is simulated using 2010 cost report data adjusted for the low-volume effect. 2012, creating greater inequity among classes of hospitals. For example, the inpatient margins of the smallest rural hospitals would have increased to 14.0 percent in 2010 if they had received the low-volume adjustment that was adopted in 2011. To focus on isolated providers and be empirically justified, the low-volume adjustment would have to be restricted to hospitals that were not close to any other acute care hospital (IPPS or CAH) and be based on total discharges rather than Medicare discharges. In addition, the lowvolume adjustment is duplicative for hospitals that receive MDH and SCH adjustments, which are also intended to compensate small hospitals for factors that result in higher costs, such as their low volume of patients. How should Medicare payments change in 2013? Each year, we provide update recommendations for services covered by Medicare's inpatient operating and outpatient prospective payment systems.11 These recommendations apply only to acute care inpatient and outpatient services; updates for services furnished in hospital-owned rehabilitation, home health, skilled nursing, and psychiatric units are based on separate recommendations for those types of Medicare services. 66 Current law: Projected increase in inpatient rates would be 2.9 percent For both the acute IPPS and the OPPS, the update in current law for fiscal year 2013 equals the projected increase in the hospital operating market basket index minus an adjustment equal to the Secretary's forecast of the 10-year average productivity growth in the country and a -0.1 percent budgetary adjustment. The operating market basket index is a projection of input price inflation for the goods and services hospitals use in producing inpatient and outpatient services. CMS's latest forecast of the change in this index for fiscal year 2013 is 2.9 percent, but it will update the forecast twice before using it to revise payments in 2013. The productivity forecast is currently 0.9 percent. The net result is a current law update of 1.9 percent (2.9 - 0.9 - 0.1). In addition, CMS temporarily reduced payment rates by 2.9 percent in fiscal year (FY) 2011 and FY 2012 to recoup overpayments in FY 2008 and FY 2009 due to hospitals' changes in documentation and coding. Therefore, if no further documentation and coding change adjustments were needed, inpatient payments would increase by a projected 1.9 percent plus 2.9 percent or 4.8 percent in total. However, CMS has also stated that an additional -1.9 percent documentation and coding change adjustment is needed to prevent further overpayments, and that: "While we are not at this time stating when we will make the remaining required 1.9 percent prospective adjustment, we consider it feasible Hospital inpatient and outpatient services: Assessing payment adequacy and updating payments Policy changes between 2010 and 2013 increase some payments and decrease others A number of payment policy changes in recent years affect our projection of 2012 hospital margins as well as payments to hospitals in 2013. Inpatient payments CMS and the Congress made a variety of policy changes affecting the acute inpatient prospective payment system (IPPS) for fiscal years (FYs) 2011, 2012, and 2013. Among them are the series of adjustments CMS made in FY 2011 and FY 2012 to account for increases in payments due to hospitals' changes in medical record documentation and coding. In 2009, CMS completed its implementation of Medicare severity-diagnosis related groups (MS- DRGs) and cost-based relative weights. CMS and the Commission concur that hospitals responded to the financial incentives of the MS-DRG system by changing medical record documentation and diagnosis coding, which resulted in assignment of cases to higher weighted MS-DRGs. Because this change in assignments increased payments without an accompanying increase in resources used, it resulted in unintended increases in payments. As a part of the TMA, Abstinence Education, and QI Programs Extension Act of 2007 (TMA), the Congress mandated a payment reduction of 0.6 percent in FY 2008 and an additional 0.9 percent reduction in FY 2009 to offset the effects of changes in documentation and coding projected by the CMS Office of the Actuary. To the extent that the TMA reductions differ from the actual effects of hospitals' coding improvements, the Secretary of the Department of Health and Human Services (HHS) is required by law to adjust hospital payments in FY 2010, FY 2011, and FY 2012 to recover (restore) any overpayments (underpayments) that occurred in FY 2008 and FY 2009. The Secretary is also required to adjust payment rates further to prevent overpayments from continuing. Analyses by both CMS and the Commission found that hospitals' changes in documentation and coding increased payments by 2.5 percent in 2008 and by a cumulative 5.4 percent by 2009. After accounting for the statutory adjustments of -1.5 percent taken in 2008 and 2009, the net overpayments to hospitals were 1.9 percent in 2008 and 3.9 percent in 2009, or 5.8 percent cumulatively. To recover the 5.8 percent in overpayments that occurred in 2008 and 2009, CMS decided to make a temporary adjustment of -2.9 percent in 2011 and to leave that adjustment in place in 2012. (Without action, payments will go back up by 2.9 percent in 2013.) In addition to recovering past overpayments, CMS concluded that to fully prevent future overpayments, it must reduce payments by a total of 5.4 percent. CMS has taken 0.6 percent in 2008, 0.9 percent in 2009, and 2.0 percent in 2012 and indicated it will take the remaining 1.9 percent in future years. The Patient Protection and Affordable Care Act of 2010 (PPACA) mandated several policy changes that affect inpatient hospital payments for FY 2011, FY 2012, and FY 2013. Among them are five permanent and two temporary policy changes. Two of the five permanent policies affect hospital wage indexes. o The first permanent policy is the frontier wage index policy, which states that the wage index for the most rural states (frontier states) cannot be less than the national average. We are not aware of any empirical support for this policy, which implicitly assumes that the frontier states always have wage rates that are equal to or above the national average. Because of this policy, hospitals in frontier states (Montana, Nevada, North Dakota, South Dakota, and Wyoming) that have a wage index less than 1.0 are granted a wage index equal to 1.0. The frontier wage index policy began in FY 2011, and the Commission estimates that in FY 2012 payments for the 48 urban and rural hospitals affected by this policy will increase by $43 million in aggregate. o The second permanent policy is the rural floor policy, which states that urban areas cannot have a lower wage index than rural areas of their state. We are not aware of any empirical support for this policy, which implicitly assumes that rural areas always have wages that are equal to or below urban areas. To pay for the additional payments that some hospitals receive because of the "rural floor," PPACA mandated that the Secretary of HHS enact a national budget-neutrality factor. The adjustment (continued next page) Report to the Congress: Medicare Payment Policy | March 2012 67 Policy changes between 2010 and 2013 increase some payments and decrease others (cont.) can be substantial. For example, when the rural Nantucket Cottage Hospital deactivated its critical access hospital status, thus becoming the only rural IPPS hospital in Massachusetts, it set the rural floor for all Massachusetts hospitals at the wages paid on Nantucket, a very high-cost island community. CMS estimated that this change yielded $274 million in extra payments to 60 urban hospitals in Massachusetts--a nearly 9 percent increase in inpatient payments. These extra payments will be offset by lowering payments to other IPPS hospitals across the country. In aggregate, the rural floor policy can reduce payments to hospitals that do not receive this benefit by up to 0.5 percent. The Commission recommended eliminating these special wage index adjustments and adopting a new wage index system to avoid geographic inequities that can occur due to current wage index policies (Medicare Payment Advisory Commission 2007). o The third permanent policy is PPACA's mandate to apply budget and productivity adjustments in determining annual hospital payment updates. The adjustments began with a 0.25 percentage point reduction to the hospital payment update for the second half of FY 2010. A 0.25 percentage point reduction was also applied in FY 2011. For FY 2012, the reduction is 0.1 percentage point, but it is paired with a reduction for productivity growth equal to the 10-year moving average of nonfarm multifactor productivity for the period ending in FY 2012 (1.0 percentage point). Therefore, in FY 2012 the payment update based on 3.0 percent projected input price inflation is reduced to 1.9 percent. Adjustments for documentation and coding changes and other factors brought the 2012 net change in payment rates down to 1 percent. For FY 2013, the payment update will be reduced by 0.1 percentage point as well as the projected 10-year moving average of nonfarm multifactor productivity for the period ending in FY 2013 and any adjustments to prevent further accumulation of overpayments due to documentation and coding changes. o The fourth permanent policy mandated by PPACA was the value-based purchasing (VBP) program. Beginning in FY 2013, the VBP program will redistribute a pool of dollars equal to 1 percent of inpatient DRG payments ($850 million in FY 2013) to hospitals based on their overall performance on a set of quality measures. The size of the VBP redistribution pool is mandated to increase 0.25 percentage point each year, reaching a maximum of 2 percent of DRG payments in FY 2017. o The fifth permanent policy mandated by PPACA was the hospital readmissions reduction program. Also beginning in FY 2013, this policy will reduce payments to hospitals that have higher than expected risk-adjusted readmissions. (See our hospital readmissions discussion.) Two temporary hospital payment policies PPACA authorized will expire at the conclusion of FY 2012. o First, PPACA mandated the expansion of the lowvolume adjustment policy for FY 2011 and FY 2012. This policy is intended to provide additional payments to rural hospitals that have a low volume of Medicare (not all payers) inpatient discharges and that are 15 miles or more from the nearest IPPS hospital. We estimate that the expansion of the lowvolume adjustment increased payments to rural hospitals by approximately $380 million in FY 2011 and $365 million in FY 2012. o Second, PPACA also authorized creation of the lowspending county hospital payment policy for FY 2011 and FY 2012. This policy provides additional payments to hospitals in counties with relatively low levels of Medicare spending per beneficiary. In both years, approximately 400 hospitals qualified for the additional payments and, as mandated, shared the fixed pool of dollars available ($150 million for FY 2011 and $250 million for FY 2012).12Absent legislative action, both programs will expire at the end of FY 2012. Two non-PPACA hospital payment policies are due to expire during the policy window stretching from FY 2011 to FY 2013. (continued next page) 68 Hospital inpatient and outpatient services: Assessing payment adequacy and updating payments Policy changes between 2010 and 2013 increase some payments and decrease others (cont.) o First, Section 508 of the Medicare Prescription Drug, Improvement, and Modernization Act of 2003, which gave eligible hospitals an opportunity for a one-time reclassification to a different labor market and allowed them the opportunity to increase their payments, expired at the end of FY 2011. CMS estimated that the expiration of Section 508 would decrease overall inpatient spending by more than $220 million in one year. o Second, the Medicare-dependent hospital (MDH) program will expire at the end of FY 2012. As a part of the MDH program, eligible hospitals can receive an additional payment to augment their standard IPPS payments if they are rural, if they have fewer than 100 beds, and if at least 60 percent of the inpatient days or discharges are covered under Medicare Part A. We estimate that the MDH program will provide $120 million in additional payments in FY 2012. New readmission policy starting in 2013 As required by PPACA, the hospital readmission reduction program will be implemented beginning in FY 2013. Under the readmission reduction program, hospitals that have excess Medicare readmissions for selected conditions will have their IPPS payments reduced. In FY 2013 and FY 2014, the readmission reduction program applies to just three conditions: acute myocardial infarction (AMI), heart failure, and pneumonia. In FY 2015, the program will be expanded to at least four additional conditions, including chronic obstructive pulmonary disease, coronary artery bypass graft surgery, percutaneous transluminal coronary angioplasty, other vascular conditions, and other conditions the Secretary may deem appropriate. The Secretary will use the National Quality Forum- endorsed risk-adjusted 30-day readmission measures for AMI, heart failure, and pneumonia currently reported on Hospital Compare. The Secretary plans to use three years of data to evaluate each hospital's readmission performance; a hospital must have at least 25 initial Medicare admissions for an individual condition to be evaluated. Hospitals whose Medicare risk-adjusted readmission rates for any of the three conditions are greater than the national average rates for the conditions (in other words, those that have "excess readmissions") will have their 2013 IPPS payment rates reduced. The payment penalty will be applied to IPPS payments for all Medicare discharges, not just discharges for the measured conditions. The payment penalty is calculated as the sum of base DRG payments for excess readmissions (based on the initial discharges that resulted in readmissions) divided by the sum of base DRG payments for all Medicare cases.13 The payment penalty is capped at 1 percent of a hospital's base DRG payments in 2013, 2 percent in 2014, and 3 percent in 2015 and thereafter. Two aspects of the readmissions penalty are counterintuitive. The first is that CMS's current estimates of risk-adjusted readmission rates are based on a method that blends the experience of the subject hospital with the average experience in the country. The smaller the hospital, the less of its information is used and the more of the national average is used. If CMS continues to use this method in the readmission reduction program, it will tend to underestimate excess readmissions, especially for small hospitals that have high readmission rates. This underestimate would have the effect of reducing potential penalties. The second counterintuitive aspect of the policy tends to work in the opposite direction and could increase potential penalties. The formula in the law produces a higher count of excess readmissions than if the calculation were based on taking the difference between actual and expected readmissions, thus producing a higher estimate of Medicare spending on excess readmissions. The law, however, is explicit in how it lays out the size of the penalty. The two counterintuitive aspects of the policy tend to somewhat offset each other. Therefore, any reexamination of how the readmission policy functions should consider both aspects. Outpatient payments Outpatient policy changes for rural and cancer hospitals change our projections of margins in FY 2012. First, sole community hospitals and other rural hospitals with 100 (continued next page) Report to the Congress: Medicare Payment Policy | March 2012 69 Policy changes between 2010 and 2013 increase some payments and decrease others (cont.) or fewer beds receive hold-harmless outpatient payments through 2011. Payment rates for these hospitals were based on the higher of the current outpatient prospective payment system (PPS) rates or the hospital's historic payment-to-cost ratio applied to its current reported outpatient costs. For example, if a hospital received payment equal to 95 percent of its costs for care in 1998 before implementation of the outpatient PPS and its outpatient PPS payments in the current year were below this level, the hospital would receive holdharmless payments. In 2011, a hospital's hold-harmless payments were equal to 85 percent of the difference between the hospital's historic cost-based payments and its outpatient PPS payments. As of January 2012, these adjustments expired, which will result in a decline in outpatient payments for some rural hospitals. Second, PPACA directed the Secretary to study whether the outpatient costs incurred by 11 cancer centers exceed those incurred by other hospitals. CMS found that the cancer centers incur higher costs for outpatient services than do other hospitals. In response, CMS has increased the outpatient PPS payment rates for the 11 cancer hospitals. These hospitals already received payments that were roughly 20 percent above base PPS rates due to an outpatient hold-harmless policy, and this new adjustment increased payments by an additional 10 percent. The net result is that outpatient payments to cancer hospitals are roughly 30 percent above base payment rates. Because this change is budget neutral, outpatient payments to all other PPS hospitals are expected to decrease by 0.2 percent ($71 million) in FY 2012. Health information technology The American Recovery and Reinvestment Act of 2009 initiated the electronic health record (EHR) to make all or most of the adjustment in FY 2013, when a +2.9 percent adjustment will be factored into rates to offset the one-time FY 2012 recoupment adjustment" (Centers for Medicare & Medicaid Services 2011). If CMS made a -1.9 percent adjustment for documentation and coding in 2013, the projected change in payment rates under current law would be 2.9 percent. The 2.9 percent increase in payment rates would reflect the sum of a 2.9 percent market basket increase, -0.9 percent 70 incentive payment program to provide payment incentives for hospitals and physicians to adopt EHR technology. Hospitals began earning payments under this program in FY 2011, and payments will continue each year until FY 2017. Under the law, a hospital can earn as many as four years of incentive payments if it is deemed a meaningful user of EHRs--based on meeting specified criteria concerning the capabilities of its EHR system released in CMS's Medicare and Medicaid EHR Incentive Program final rule (Centers for Medicare & Medicaid Services 2010).14 The payment each hospital receives will equal the sum of an initial payment amount per hospital ($2 million base amount) plus a discharge-related amount of $200 per patient discharge for all discharges between the 1,150th and 23,000th discharge, both multiplied by the hospital's share of Medicare days. Therefore, hospitals' EHR incentive payments vary with the shares that their Medicare inpatient days represent of their total days. Under the Medicare portion of this program, payments to hospitals decline in value over the course of four consecutive payment years. According to this mandated formula and assumptions we have made about the share of hospitals that will meet the EHR meaningful use criteria, we estimate that the Medicare EHR program will distribute approximately $3 billion in additional payments in 2012. We also estimate that the average large hospital (more than 400 beds) will receive payments of $2.7 million in its first year of participation and the average smaller hospital will receive payments of about $1.6 million in its first year. The law also stipulates that, in FY 2015, hospitals that fail to meet the meaningful use criteria will be penalized through the IPPS. ? productivity adjustment, -0.1 percent budget adjustment, 2.9 percent expiration of temporary documentation and coding adjustment, and -1.9 percent prospective documentation and coding adjustment. While it is in the process of recovering 2008 and 2009 overpayments due to documentation and coding, CMS needs additional legislative authority to recover overpayments that occurred or are occurring in 2010, 2011, and 2012. Hospital inpatient and outpatient services: Assessing payment adequacy and updating payments spending implication of this recommendation is based on Medicare spending projections that were made prior to a sequester, as the recommendation was developed and voted on before the sequester was triggered and became current law. If a Medicare sequester does occur, it will change the spending implication of the recommendation. RECOMMENDATION 3-1 The Congress should increase payment rates for the inpatient and outpatient prospective payment systems in 2013 by 1.0 percent. For inpatient services, the Congress should also require the Secretary of Health and Human Services beginning in 2013 to use the difference between the increase under current law and the Commission's recommended update to gradually recover past overpayments due to documentation and coding changes. RATIONALE 3-1 The Commission balanced three factors in reaching its inpatient update recommendation. First, most payment adequacy indicators (including access to care, quality of care, and access to capital) are positive. Second, hospitals' documentation and coding changes led to overpayments in 2010, 2011, and 2012. Updates must be lowered to recover these overpayments. Third, while relatively efficient hospitals generated positive overall Medicare margins in 2010, most hospitals have negative overall Medicare margins (-4.5 percent in 2010 and projected to reach -7 percent in 2012). Balancing these factors, the Commission recommends reducing the 2013 increase in inpatient payments from the level in current law (currently expected to be 2.9 percent) to 1 percent.15 The difference between the update under current law and 1 percent should be used to gradually recover overpayments that occurred due to documentation and coding changes, which will allow Medicare to recover past overpayments and keep 2013 inpatient payment rates adequate. For outpatient services, the Commission also recommends a 1 percent increase in payment rates. On the one hand, growth in the volume of outpatient services has been strong, suggesting the outpatient update in current law (1.9 percent) is too high. On the other hand, overall hospital margins are negative, suggesting a positive update is appropriate. A 1 percent update would balance these two considerations and also help limit growth in the disparity in payment rates between services provided in OPDs and payment rates in other sectors. The Commission maintains that Medicare should seek to pay similar amounts for similar services, taking into account differences in the quality of care and in the relative risks of patient populations. IMPLICATIONS 3-1 Spending o This recommendation would decrease Medicare spending by more than $2 billion in 2013 and would save more than $10 billion over five years. The Beneficiary and provider o This recommendation should have no negative impact on beneficiary access to care and is not expected to affect providers' willingness and ability to provide care to Medicare beneficiaries. Equalizing payment rates for outpatient office visits in freestanding physician offices and outpatient departments As we considered an update to outpatient payment rates, we also considered ways to limit the differences in payment rates between hospitals and physician offices for the same (or similar) services. This effort is the start of a broader effort by the Commission to move toward having the same payment for the same service provided to similar patients across sites of care. The issue of E&M payment rates is particularly timely because of the increase in physician employment by hospitals in recent years. Many factors have been cited for this trend: o Financially, physicians are faced with rising costs associated with private practice, including new technology such as electronic health records and the administrative costs of dealing with insurers, each of which has its own requirements for submitting claims. Also, they may not have the leverage with insurers to negotiate payment rate increases that keep pace with rising expenses. Further, physicians of all specialties desire to avoid the uncertainty of changes in professional liability insurance premiums (Ginsburg 2011b, O'Malley et al. 2011). o Many physicians--especially younger ones--desire a different work-life balance and more lifestyle flexibility than has been typical in the past (BDC Advisors 2010, Ginsburg 2011a, Healthcare Financial Management Association 2011, Kocher and Sahni 2011, O'Malley et al. 2011). Hospital employment may enable physicians to work fewer and more predictable hours and to focus on the clinical aspects of medicine. They may be willing to give up their autonomy in exchange for these benefits. Report to the Congress: Medicare Payment Policy | March 2012 71 TABLE 3-9 Differences in program payments and beneficiary cost sharing for midlevel outpatient office visit provided in freestanding practices and hospital-based entities, 2011 Service provided in hospital-based entity Service provided in freestanding physician practice* Physician facility rate* Outpatient PPS rate** Total, hospitalbased rate Program payment $55.18 $39.42 $60.10 $99.52 Beneficiary cost sharing +13.79 +9.85 +15.03 +24.88 68.97 49.27 75.13 124.40 Total payment Note: PPS (prospective payment system). The Current Procedural Terminology code for this visit is 99213. * Paid under the Medicare physician fee schedule. **Paid under the outpatient PPS. Source: MedPAC analysis of payment rates from the outpatient PPS and physician fee schedule in 2011. o Hospitals often choose to employ physicians to ensure a stable stream of tests, admissions, and referrals to specialists who perform their services at the hospital. o PPACA creates a Medicare shared savings program for accountable care organizations (ACOs), which are integrated health care systems composed of physicians and health care facilities that take responsibility for controlling spending and increasing quality. ACOs could be established by hospitals or by groups of physicians working together. Hospitals may be acquiring physician practices to position themselves to establish ACOs. o Physicians and hospitals can benefit financially from hospital employment of physicians. Large hospital systems can use their market power to obtain higher rates for physician services from private insurers in some markets (Ginsburg 2010). In addition, for most services that can be provided in a physician office or OPD, total Medicare payments (program payments and cost sharing) are substantially higher if the service is provided in an OPD rather than in a physician office. The combination of higher private insurance payments and higher Medicare payments may allow hospitals to offer physicians comparable incomes as employees, even if the hospital has higher overhead than freestanding practices. As more physicians become employed by hospitals, billing of services is likely to shift from freestanding physician practices to OPDs. Because most services have higher payment rates under the OPPS than under Medicare's 72 physician fee schedule (PFS), the result of such a shift is higher program spending and beneficiary cost sharing. We start our evaluation of this issue by examining differences in payment rates for E&M office visits provided in OPDs and physician offices. For example, in 2011 Medicare paid 80 percent more for a 15-minute visit--Current Procedural Terminology (CPT) code 99213--provided in an OPD than in a freestanding office of a physician or other health care professional paid under the PFS. This payment difference creates a financial incentive for hospitals to purchase freestanding physician offices and convert them to OPDs without changing their location or patient mix. We have seen a 6.7 percent increase in the number of these visits furnished in OPDs from 2009 to 2010. Thus, Medicare expenditures and beneficiary cost sharing could increase without any difference in patient care. In this section, we consider a policy of making Medicare payments for E&M office visits equal whether they are provided in OPDs or in physician offices. In the future, we plan to examine payment differentials between hospitals and physician offices for other services. Comparing Medicare's payments for services in physician offices and outpatient departments Services covered under the PFS have two payment rates: one rate for when the physician provides the service in his or her office (the nonfacility rate) and another rate for when the physician provides the service in a facility such as an OPD or other provider-based entity (the facility rate).16 An outpatient facility or organization that has Hospital inpatient and outpatient services: Assessing payment adequacy and updating payments provider-based status is considered part of a hospital, and provider-based status is generally available for hospital-owned entities that are on the hospital campus or within 35 miles of the hospital campus. In general, the nonfacility rate is higher than the facility rate in the PFS because physician practice costs are higher when physicians provide care in their offices than in facilities, as they have to cover their direct costs (e.g., equipment, supplies, and staff). When a service is provided in a physician office, there is a single payment for the service. However, when a service is provided in a facility, Medicare makes a payment to the facility in addition to a payment to the physician. For example, if a 15-minute E&M office visit for an established patient (CPT code 99213) is provided in a freestanding physician office, the program pays the physician 80 percent of the nonfacility payment rate from the PFS, and the patient is responsible for the remaining 20 percent. In 2011, the nonfacility rate for this service was $68.97; the program pays $55.18 and the patient is responsible for $13.79 (Table 3-9). If the same service is provided in an OPD-based entity, the program pays 80 percent of the PFS facility rate and 80 percent of the outpatient PPS rate, and the patient is responsible for 20 percent of both rates.17 The PFS facility rate in 2012 is $49.27, and the OPPS payment is $75.13, for a total payment of $124.40. The program pays $99.52, and the patient is responsible for $24.88 (Table 3-9). Potential spending effects of services moving from physician offices to hospital-based entities Medicare data on the site of care for E&M office visits suggest that the increase in hospital employment of physicians has been associated with a shift of services from offices to OPDs. In 2004, 8 percent of specialists and 23 percent of primary care physicians were employed by hospitals (Kocher and Sahni 2011). In 2008, the percentages of specialists and primary care physicians employed by hospitals had increased to 15 percent and 31 percent, respectively. The proportion of E&M office visits provided in OPDs reflects this increased hospital employment of physicians. The percentage of E&M office visits provided in OPDs increased from 5.1 percent in 2004 to 7.3 percent in 2010. However, growth in the percentage of E&M office visits that are provided in OPDs has accelerated, increasing at an annual rate of 3.5 percent from 2004 through 2008, by 9.9 percent in 2009, and by 12.9 percent in 2010.18 As more physicians become employed by hospitals, it is likely that more services will migrate from physician offices to OPDs (or other hospitalbased entities), which would increase Medicare spending. The magnitude of the increased Medicare spending is difficult to estimate for some OPD services where the packaging of ancillary services differs between the PFS and the OPPS. The OPPS packages many ancillary services and supplies with their associated procedures for payment purposes, whereas the PFS often pays separately for ancillary items and services (Medicare Payment Advisory Commission 2011b). However, we have greater confidence in estimating the potential effect of a shift of E&M office visits from offices to OPDs because the level of packaging is relatively low for these services, about 2.5 percent of the total cost. The potential effect on Medicare spending of a large shift in these visits from freestanding physician practices to hospital-based clinics that are billing as part of an OPD is significant. If the percentage of E&M office visits that are provided in OPDs grows at 12.9 percent (as it did in 2010) over 10 years, about 24.5 percent of E&M office visits will occur in OPDs in 2020. Such a shift would increase program spending by $2.0 billion per year and beneficiary cost sharing by $500 million per year (assuming 2010 payment rates). Options for equalizing payment rates for E&M office visits across settings Variations in payment rates among different ambulatory care settings raise questions about how Medicare should pay for the same (or similar) services in different settings. Medicare should strive to ensure that patients have access to settings that provide the appropriate level of care. If the same service can be safely provided in different settings, it may be undesirable for a prudent purchaser to pay more for that service in one setting than in another. Payment variations across settings may encourage arrangements among providers that result in more care being provided in higher cost (and higher paid) settings, thereby increasing total Medicare spending. Therefore, to be a prudent purchaser of medical care, the Commission believes that Medicare should base payment rates on the resources needed to treat patients in the most efficient setting, adjusting for differences in patient severity, to the extent that severity differences affect costs. The easiest way to address this issue is to set payment rates in the OPPS and PFS so that payments are equal whether a service is provided in a freestanding practice or in an OPD. However, for many services, we are concerned Report to the Congress: Medicare Payment Policy | March 2012 73 that such a policy would fail to account for some important differences between physician offices and OPDs: o Hospitals incur costs to maintain standby capacity for handling emergencies and to comply with additional regulatory requirements. Hospitals are subject to the Emergency Medical Treatment and Active Labor Act, which requires them to screen and stabilize (or transfer) patients who believe they are experiencing a medical emergency, regardless of their ability to pay.19 This mission may make the cost of certain services performed in OPDs higher than in physician offices, which typically do not provide emergency care. In addition, hospitals are required to meet conditions of participation in the Medicare program that likely increase hospital costs; these conditions do not apply to physician offices. o Patient complexity may differ in these two sectors. For many services, greater patient complexity may result in higher costs of care. o For services covered under both the OPPS and the PFS, the OPPS typically packages the cost of ancillary services and supplies to a greater extent than does the PFS. For many services, these factors can cause higher costs in OPDs than in physician offices.20 Therefore, we chose to narrow our focus for equalizing payment rates across these two sectors to E&M office visits, which are indicated by CPT codes 99201 through 99215. For these services, we believe it is reasonable to set payment rates equal in the PFS and the OPPS because: o o 74 Hospitals should not need to maintain standby capacity for E&M office visits that are not provided in an emergency department, nor should requirements to stabilize patients presenting at the emergency room affect the costs of furnishing E&M office visits. To a large extent, differences in resource needs because of patient complexity for these visits are reflected in their coding structure, which classifies visits based on their length and complexity. For example, CPT code 99213 is for visits that typically include 15 minutes of face-to-face time between the physician and patient, whereas CPT code 99214 is for visits that typically include 25 minutes of face-to-face time between the physician and patient and involve a more detailed history and examination. This coding structure is the same whether the visit is provided in a physician office or in an OPD.21 o On the basis of our analysis of 10,000 OPD claims that included an E&M office visit, the cost of ancillary services that are packaged with these visits when provided in an OPD is about 2.5 percent of the visits' total cost, which means that ancillaries add about $2 to the payment rate of the average E&M office visit provided in OPDs; therefore, the content of the unit of payment is similar across settings. We conclude that the E&M visits are a service in which rates should be equalized between PPS hospital OPDs and other sites of care that use the physician fee schedule. The payment rate for both settings should be based on the cost of the most efficient setting where high-quality care can be provided. In this case, our best proxy for the cost of efficiently delivering E&M services is the E&M rate paid to physician offices. We realize that over time adjustments to E&M rates in the physician fee schedule will also affect the price paid in OPDs. Although fee schedule payment rates for primary care services such as E&M visits have increased over the past several years, the Commission has recommended further improvements to the accuracy of fee schedule payments (see text box, p. 76). To ensure that payments for E&M services are equal across PPS settings, Medicare should set the OPPS rate equal to the difference between the nonfacility practice expense and the facility practice expense in the physician fee schedule. Under this formula, total Medicare payment rates would be the same whether the E&M visit occurs in an OPD or in a nonfacility ambulatory site such as a physician office (Table 3-10). The payment to physicians for their work would not change and payments to costbased providers such as CAHs would not change under the proposal. RECOMMENDATION 3-2 The Congress should direct the Secretary of Health and Human Services to reduce payment rates for evaluation and management office visits provided in hospital outpatient departments so that total payment rates for these visits are the same whether the service is provided in an outpatient department or a physician office. These changes should be phased in over three years. During the phase-in, payment reductions to hospitals with a disproportionate share patient percentage at or above the median should be limited to 2 percent of overall Medicare payments. Hospital inpatient and outpatient services: Assessing payment adequacy and updating payments TABLE 3-10 Payment rates to physicians and OPDs for a midlevel E&M office visit under current payment rates and policy that aligns payment rates across settings, 2011 Payment amount Calculation Current payment rates Service in physician office Payment to physician $68.97 Work/PLI ($35.33) + nonfacility PE ($33.64) Service in OPD Payment to physician 49.27 Work/PLI ($35.33) + facility PE ($13.94) Payment to hospital 75.13 Hospital outpatient department rate ($75.13) Total payment $124.40 Policy that aligns rates across settings Service in OPD Payment to physician 49.27 Work/PLI ($35.33) + facility PE ($13.94) Payment to hospital 19.70 Nonfacility PE ($33.64) - facility PE ($13.94) Total payment Note: $68.97 OPD (hospital outpatient department), E&M (evaluation and management), PLI (professional liability insurance), PE (practice expense). The Current Procedural Terminology code for this visit is 99213. Payments include both program spending and beneficiary cost sharing. Source: MedPAC analysis of payment rates in the 2011 physician fee schedule and outpatient prospective payment system. RATIONALE 3-2 Hospitals have been acquiring physician practices and employing physicians at an increasing rate. As more physicians become employed by hospitals, E&M office visits will shift from being billed as physician office services to being billed as OPD services. When hospitals bill for E&M office visits as OPD services, there are negative consequences for the Medicare program, beneficiaries, and the efficiency of the health care system: o Medicare currently pays higher rates for care in existing OPD clinics. If the movement toward OPD billing continues, spending would increase by an additional $2 billion annually by 2020 if the OPD share of E&M visits grows at its current rate. o Beneficiary cost sharing is substantially higher when E&M office visits are billed as OPD visits, and beneficiaries' Part B premiums increase as services shift to OPDs due to higher OPD rates. In addition, beneficiaries can be confused when they receive two coinsurance bills for a single E&M office visit. o When hospitals convert physician office buildings to OPD status, they spend money to comply with the life safety codes and take on the cost of generating additional bills for the hospital's facility payment. For E&M office visits, these additional expenditures result in higher Medicare payments but fail to create clear benefits for patients. To improve the efficiency of the health care system, Medicare should be discouraging, not encouraging, expenditures by health care providers that do not benefit patients. Setting the payment rates for E&M office visits provided in OPDs equal to the difference between the nonfacility practice expense rate and the facility practice expense in the PFS would result in payment rates that are equal whether an E&M office visit is provided in an OPD or in a freestanding practice. This practice would reduce the negative effects on the Medicare program, beneficiaries, and the health care system's efficiency. Reducing OPPS rates for E&M office visits would reduce overall and outpatient Medicare revenue for most hospitals. If this recommendation were fully implemented, we estimate that hospital overall Medicare revenue would be 0.6 percent lower under this policy than it otherwise would be, and outpatient revenue would be 2.8 percent lower (Table 3-11, p. 77). However, it is prudent to allow time for hospitals to adjust to the lower rates for E&M Report to the Congress: Medicare Payment Policy | March 2012 75 Payments for primary care services T he process through which CMS reviews the accuracy of the physician fee schedule's relative values has problems that led to primary care services (such as evaluation and management services) becoming undervalued over time and other services becoming overvalued (Medicare Payment Advisory Commission 2010b). These concerns led the Commission to make a series of recommendations to improve the process for identifying and correcting misvalued services (Medicare Payment Advisory Commission 2006). As a result of our recommendations, greater scrutiny of misvalued services, and changes to the methodology and data used to calculate practice expense values, payment rates for primary care services have increased in recent years (Medicare Payment Advisory Commission 2011c). In addition, the Commission recommended an adjustment to raise payments for selected primary care services furnished by primary care practitioners, which was adopted by the Congress (Medicare Payment Advisory Commission 2008b). Nevertheless, the Commission is still concerned that many fee schedule services are overvalued and that resources should be reallocated to other services, including primary care. For example, the relative value units (RVUs) for practitioner work are largely a function of estimates of the time it takes a practitioner to perform each service. The current time estimates rely visits; therefore, we recommend that this policy be phased in over three years. During the phase-in, one-third of the adjustment would occur in the first year, two-thirds in the second year, and payments would be fully adjusted in the third year. One benefit of the phase-in is to delay full implementation of the policy until after Medicare starts paying hospitals a portion of their uncompensated care costs in 2014. Starting in 2014, a portion of funds currently distributed as DSH payments will start to be distributed to compensate hospitals for a share of their uncompensated care costs (charity care and bad debts). To the extent that a hospital is serving an above-average share of uninsured and underinsured individuals, it will receive a larger share of the payments from the uncompensated care pool. If the 76 primarily on surveys conducted by physician specialty societies that have a financial stake in the process. Research for CMS and for the Assistant Secretary for Planning and Evaluation of the Department of Health and Human Services has shown that the time estimates are likely too high for some services (Medicare Payment Advisory Commission 2011c). In a recent letter to the Congress on the sustainable growth rate system, the Commission recommended that the Congress direct the Secretary to regularly collect data--including service volume and work time--to establish more accurate work and practice expense values (Medicare Payment Advisory Commission 2011a). To help assess whether Medicare's fees are adequate for efficient care delivery, the data should be collected from a cohort of efficient practices rather than from a sample of all practices. The Commission also recommended that the Congress direct the Secretary to identify overpriced fee schedule services and reduce their RVUs accordingly. These reductions should be budget neutral within the fee schedule, which would redistribute payments from overpriced to underpriced services. In addition, the Congress should set an annual numeric goal for RVU reductions of at least 1 percent of fee schedule spending. See Appendix B for a full description of these recommendations. ? objective is to assist hospitals serving the poor, paying a portion of their uncompensated care costs is a far better targeting of Medicare dollars than making high payments to all hospitals for E&M visits. To evaluate the effect of this policy on the adequacy of Medicare payments for different categories of hospitals, we focus on its effect on overall Medicare revenue. However, for completeness, we also present the effect as a share of outpatient revenue only (Table 3-11). We find that the aggregate effect of this policy on hospitals' overall Medicare revenue will be 0.6 percent or less, but the effect will vary widely by hospital. As a category, major teaching hospitals would have the largest loss of Medicare revenue (1.1 percent). For-profit hospitals would have the smallest loss (0.2 percent overall Medicare revenue). More than 10 Hospital inpatient and outpatient services: Assessing payment adequacy and updating payments TABLE 3-11 Hospital group Reduction in Medicare revenue from equalizing OPPS and PFS rates for E&M office visits varies widely among hospitals Percent reduction in overall Medicare revenue All hospitals 0.6% Percent reduction in outpatient Medicare revenue 2.8% Urban 0.6 2.7 Rural 0.7 2.8 6.1 Major teaching 1.1 Other teaching 0.4 2.2 Nonteaching 0.4 2.0 Nonprofit 0.6 2.8 For profit 0.2 1.0 Government 1.0 4.3 Ranking of percent revenue loss 5th percentile 0.0 0.0 10th percentile 0.0 0.0 Median 0.1 0.6 90th percentile 1.2 6.9 95th percentile 2.6 8.5 Note: OPPS (outpatient prospective payment system), PFS (physician fee schedule), E&M (evaluation and management). The reduction may be smaller to the extent hospitals shift patients to other types of clinics such as rural health clinics once payment rates for hospital-based clinics decline. Source: MedPAC analysis of 2010 cost reports and 2009 outpatient claims. percent of all hospitals would lose no Medicare revenue, and 5 percent would lose at least 2.6 percent of overall Medicare revenue. Moreover, reductions in revenue would be smaller if hospitals convert some of their outpatient clinics to rural health clinics or federally qualified health centers, which receive payments above traditional physician office rates due to serving populations that appear to be underserved. In addition, hospitals may choose to start operating the physician practices that they own as freestanding clinics, which would result in cost savings for the hospitals due to lower billing and overhead costs. We are concerned that some of the hospitals losing the most Medicare revenue provide ambulatory physician services to many low-income members of their communities. Large reductions in Medicare revenue for these hospitals may adversely affect access to ambulatory physician services in these low-income populations. Therefore, during the three-year phase-in, we recommend that revenue losses from this policy be limited to 2 percent of overall Medicare revenue for hospitals that serve a relatively large share of low-income patients. Specifically, we recommend that during the phase-in, losses be limited to 2 percent of the hospital's overall Medicare revenue for hospitals with a DSH percentage that is at or above the median for all hospitals (a stop-loss provision). A hospital's DSH percentage is the sum of the percentage of Medicare inpatient days that are for patients who are eligible for supplemental security income and the percentage of total inpatient days that are for patients who participate in Medicaid. For 2010, the median DSH percentage among all PPS hospitals was about 25 percent. Assuming no change in hospitals' operations under a fully implemented policy, we estimate that about 4 percent of hospitals would qualify for the stop loss discussed above. We find that the profile of these hospitals is mixed, but they do have some different characteristics from other hospitals. The hospitals qualifying for the stop loss are more likely to be government owned, more likely to have major teaching status, have a higher percentage of Report to the Congress: Medicare Payment Policy | March 2012 77 Medicaid patients, and have a lower all-payer margin than all other hospitals. However, hospitals that qualify for the stop loss also have a higher overall Medicare margin, probably because of relatively high payments for their DSH and teaching status. IMPLICATIONS 3-2 Spending o This recommendation would reduce Medicare program spending by between $250 million and $750 million in 2013 and by between $1 billion and $5 billion over 5 years. The spending implication of this recommendation is based on Medicare spending projections that were made prior to a sequester, as the recommendation was developed and voted on before the sequester was triggered and became current law. If a Medicare sequester does occur, it will change the spending implication of the recommendation. Beneficiary and provider o Beneficiaries would see reductions in Medicare cost sharing and in Part B premiums due to lower outpatient spending.22 However, because this recommendation would reduce payment rates for E&M office visits provided in OPDs, we need to monitor beneficiaries' access to these services. RECOMMENDATION 3-3 The Secretary of Health and Human Services should conduct a study by January 2015 to examine whether access to ambulatory physician and other health professionals' services for low-income patients would be impaired by setting outpatient evaluation and management payment rates equal to those paid in physician offices. If access will be impaired, the Secretary should recommend actions to protect access. RATIONALE 3-3 In some communities, OPDs serve as a primary source of ambulatory physician services for the low-income population. Some of these safety-net hospitals are among those that would lose the most from equal payments for E&M office visits across OPDs and physician offices. To ensure that access to ambulatory physician and other health professional services is maintained for lowincome patients that rely on these safety-net hospitals, the Secretary should study whether equal payments across OPDs and physician offices for E&M office visits impair access of low-income patients to those services. If the Secretary finds access problems, actions should be undertaken to protect access. IMPLICATIONS 3-3 Spending Ensuring access to ambulatory physician and other professional services among vulnerable populations Although we have included a phase-in that has a stop loss as part of our recommendation for setting Medicare payments for E&M office visits equal across freestanding practices and OPDs, we believe more investigation is needed on the potential effects this policy could have on access to ambulatory physician and other professional services among low-income populations. 78 o This recommendation would have no effect on program spending. Beneficiary and provider o This recommendation may help identify problems beneficiaries are having with regard to accessing ambulatory physician services. ? Hospital inpatient and outpatient services: Assessing payment adequacy and updating payments Endnotes 1 Outpatient service volume is measured by counting the number of separately payable Healthcare Common Procedure Coding System (HCPCS) codes. HCPCS definitions can change over time, as can the HCPCS codes that are paid separately and the ones that are bundled, which can have some effect on annual changes in volume. 8 Medicare spending varies in part because of the factors Medicare uses to account for differing wages, payment rates, and health status. We adjust for those factors to arrive at service use. A discussion of our methods to compute regional variation in service use is available at: http://www.medpac. gov/documents/Dec09_RegionalVariation_report.pdf. 2 The data on visits to hospital-based practices come from outpatient claims files. Data on visits to freestanding physician offices come from physicians' Medicare claims. 9 Figure 3-8 shows a smaller difference between the urban and rural margins than the 2001 report because the figure excludes margin data for any hospital that has become a CAH. 3 Occupancy reflects both Medicare and non-Medicare patients. Because occupancy is declining, we can infer that the decline in Medicare days per beneficiary is not due to a lack of capacity. 10 The MDH and SCH payments tend to increase payments toward a hospital's historical level of costs, which increases the hospitals' inpatient margin to zero. The result is SCH inpatient margins were 2.8 percent and MDH inpatient margins were -1.7 percent in 2010. The SCH add-on tends to be higher than the MDH add-on for two reasons: First, it adjusts all inpatient payments, while the MDH payment is a blend of 75 percent based on historical costs and 25 percent based on PPS rates; second, the SCH payments are based on a base year of 2006 or earlier and the MDH payments are based on a base year of 2002 or earlier. The more recent base year is more advantageous. For more details see the text box (pp. 67-70) on recent changes in payment rules. 4 The share of hospitals and their affiliates providing each service was calculated as the percentage of hospitals indicating availability of the services within the hospital, network, system, or joint venture. 5 The services included in the overall margin are Medicare acute inpatient, outpatient, graduate medical education, Medicare SNF (including swing beds), Medicare home health care, Medicare inpatient psychiatric, and Medicare inpatient rehabilitation. 6 In 2009 there was a substantial difference between the forecasted market basket used to set payment updates, projected to increase by 3.6 percent, and the actual increase of 2.6 percent, measured after the year is completed. Payment updates were set based on the forecasted market basket increase. Inpatient cost growth per discharge was roughly in between the actual and forecasted increase in the market basket. On a case-mix-adjusted basis, outpatient costs grew at underlying input prices. 7 Another common measure of hospitals' financial pressure is "days cash on hand." However, we find wide differences in this metric not just due to pressure but also due to financing choices among hospital systems. For-profit hospitals routinely have less cash on hand than nonprofits. This situation reflects differences in nonprofit and for-profit choices with respect to using available cash for investments or to pay down debt. It may in part reflect the fact that income on investments is taxable to for profits and not taxable to nonprofits. The measure is further confounded by the large numbers of hospitals that hold cash off their balance sheet in foundations. 11 Our update recommendations focus on inpatient operating payment rates and payment rates for outpatient services (which encompass both operating and capital costs of outpatient services). The Secretary of Health and Human Services makes a separate evaluation of updates to per discharge payment rates for inpatient capital costs. 12 Hospitals located in counties with relatively low levels of spending will receive a share of the fixed $150 million reserved for 2011 and $250 million reserved for FY 2012 based on their relative proportion of IPPS operating payments. PPACA set the two-year payment total at $400 million. 13 Base DRG payments reflect the sum of the hospital's wage index and cost of living adjusted operating and capital payment rates multiplied by the DRG relative weight for the (affected) MS--DRG(s). Base DRG payments do not include payments for the indirect costs of graduate medical education, service to a DSH share of low-income patients, outlier payments, or additional payments, such as those under the SCH and MDH programs. 14 The American Recovery and Reinvestment Act of 2009 mandates that EHR payments also be made to hospitals through the Medicaid program. Report to the Congress: Medicare Payment Policy | March 2012 79 15 That is, the Commission recommends that the payment rates for 2013 be increased by 1 percent from the 2012 rates. The Congress would have to override other existing statutory provisions to achieve this result. 16 The payment rates in the physician fee schedule have three parts: physician's work, practice expense, and professional liability insurance. Of the three, only practice expense differs when a service is provided in an office or a hospital-based entity. 17 In the PPS, the coinsurance rate for some services is above 20 percent. This rate is a result of a policy that CMS implemented when it launched the OPPS. In the cost-based payment system that preceded the OPPS, the coinsurance rate for most services was above 20 percent and averaged nearly 50 percent. When CMS launched the OPPS, the agency determined a dollar-denominated national coinsurance amount for each APC that occurred under the cost-based payment system. If the national coinsurance amount for an APC was above 20 percent of the APC's payment rate, CMS kept the national coinsurance amount frozen over time, while it updated the APC's payment rate annually by the hospital market basket. As the payment rates increased, the frozen national coinsurance amounts became smaller fractions of the payment rates. For each APC, CMS maintains this policy until the national coinsurance amount is 20 percent of the payment rate. After that, the national coinsurance amount is increased each year at the same rate as the payment rate. Currently, about two-thirds of the services covered under the OPPS have coinsurance rates of 20 percent, while the remaining services are above 20 percent. 18 The outpatient office services are represented by the following CPT codes: 99201, 99202, 99203, 99204, 99205, 99211, 99212, 99213, 99214, and 99215. 80 19 The most obvious feature of standby capacity for a hospital is the emergency department (ED). In the OPPS, CMS has established two broad categories of APCs for payment of ED visits, Type A and Type B. A Type A ED is an "organized hospital-based facility for the provision of unscheduled episodic services to patients who present for immediate medical attention. The facility must be available 24 hours a day." CMS indicates that a Type B facility has less stringent criteria than a Type A facility, but its (lengthy) definition indicates that it is available for emergency care on an urgent basis. 20 The arguments for higher costs in OPDs than in physician offices that we discuss in this chapter are similar to those discussed for higher costs in OPDs than in ASCs discussed in Chapter 5. In particular, OPDs face higher costs than ASCs because of greater regulatory burdens and higher patient complexity. 21 For clinic and emergency department visits, CMS has instructed hospitals to develop internal guidelines for reporting the appropriate visit level. Although this procedure gives hospitals some leeway in how they code E&M office visits, CMS has advised hospitals to follow the intent of the descriptions for these CPT codes. 22 Because beneficiaries' Part B premiums are based on total Part B spending (including OPD spending), the new E&M policy will reduce Part B premiums. The rate of reduction will be slowed by the transition policies, which act to slow the financial impact of the policy on OPD spending. The policy will also act to reduce beneficiaries' direct cost-sharing burden due to lower prices for E&M visits on which the 20 percent cost sharing is based. The speed at which cost sharing is reduced will be slowed by the three-year transition. However, the 2 percent stop-loss provision would not directly affect cost sharing because it will be an adjustment to overall payments, not an adjustment to payment rates from which the beneficiaries' 20 percent cost sharing is derived. Hospital inpatient and outpatient services: Assessing payment adequacy and updating payments References Agency for Healthcare Research and Quality, Department of Health and Human Services. 2007a. AHRQ quality indicators. Guide to inpatient quality indicators: Quality of care in hospitals--Volume, mortality, and utilization, Version 3.1. Rockville, MD: AHRQ. March 12. http://www.qualityindicators. ahrq.gov/iqi_download.htm. Fitch Ratings. 2011. 2011 median ratios for nonprofit hospitals and healthcare systems. August 25. Agency for Healthcare Research and Quality, Department of Health and Human Services. 2007b. AHRQ quality indicators: Guide to patient safety indicators, Version 3.1. Rockville, MD: AHRQ. March 12. http://www.qualityindicators.ahrq.gov/psi_ download.htm. Ginsburg, P. B. 2010. Wide variation in hospital and physician payment rates. Evidence of provider market power. Research brief no. 16. Washington, DC: Center for Studying Health System Change. Agency for Healthcare Research and Quality, Department of Health and Human Services. 2009a. Agency for Healthcare Research and Quality AHRQ Quality Indicators (AHRQ QI): Guidance on using the AHRQ QI for hospital-level comparative reporting. Rockville, MD: AHRQ. June 30. Agency for Healthcare Research and Quality, Department of Health and Human Services. 2009b. Guidance on using the AHRQ QI for hospital-level comparative reporting. Rockville, MD: AHRQ. June. BDC Advisors. 2010. Trends in physician-hospital integration. Presentation prepared by staff from BDC Advisors. Bureau of Labor Statistics, Department of Labor. 2011. Consumer price index--all urban consumers. http://www.bls.gov/cpi/. Carpenter, D. 2011. 2011 building report: Shifting priorities. Health Facilities Management, February. Centers for Medicare & Medicaid Services, Department of Health and Human Services. 2010. Medicare program; hospital inpatient prospective payment systems for acute care hospitals and the long term care hospital prospective payment system changes and FY2011 rates; provider agreements and supplier approvals; and hospital conditions of participation for rehabilitation and respiratory care services; Medicaid program: accreditation for provider of inpatient psychiatric services. Final rule. Federal Register 75, no. 157 (August 16): 50042-50677. Centers for Medicare & Medicaid Services, Department of Health and Human Services. 2011. Medicare program; hospital inpatient prospective payment systems for acute care hospitals and the long-term care hospital prospective payment system changes and FY2012 rates; hospitals' FTE resident caps for graduate medical education payment. Final rule. Federal Register 76, no. 160 (August 18): 51476-51846. Flex Monitoring Team, Rural Research Centers. 2011. Count of critical access hospitals as of March 31, 2011. http://www. flexmonitoring.org. Ginsburg, P. B., Center for Studying Health System Change. 2011a. Health care industry consolidation. Testimony before the Subcommittee on Health, Ways and Means Committee, U.S. House of Representatives. 112th Cong., 1st sess. September 9. Ginsburg, P. B. 2011b. Hospital employment of physicians. Presentation at the National Health Policy Forum, March 25. Healthcare Financial Management Association. 2011. The new era for hospital-physician alignment. Westchester, IL: HFMA. IHS Global Insight. 2011. Healthcare cost review: Third quarter 2011. Washington, DC: IHS Global Insight. Jha, A. K., E. J. Orav, and A. M. Epstein. 2010. The effect of financial incentives on hospitals that serve poor patients. Annals of Internal Medicine 153, no. 5 (September 7): 299-306. Kocher, R., and N. R. Sahni. 2011. Hospitals' race to employ physicians--The logic behind a money-losing proposition. New England Journal of Medicine 364, no. 19 (May 12): 1790-1793. Landrigan, C. P., G. J. Parry, C. B. Bones, et al. 2010. Temporal trends in rates of patient harm resulting from medical care. New England Journal of Medicine 363, no. 22 (November 25): 2124- 2134. Medicare Payment Advisory Commission. 2001. Report to the Congress: Medicare in rural America. Washington, DC: MedPAC. Medicare Payment Advisory Commission. 2006. Report to the Congress: Medicare payment policy. Washington, DC: MedPAC. Medicare Payment Advisory Commission. 2007. Report to the Congress: Promoting greater efficiency in Medicare. Washington, DC: MedPAC. Department of Health and Human Services. 2010. Hospital Compare database. http://www.hospitalcompare.hhs.gov/. Report to the Congress: Medicare Payment Policy | March 2012 81 Medicare Payment Advisory Commission. 2008a. Department of Health and Human Services report to Congress: Plan to implement a Medicare hospital value-based purchasing program (November 21, 2007). Comment letter to Chairmen and Ranking Members of Committee on Finance, U.S. Senate and Committee on Ways and Means, U.S. House of Representatives. April 10. Medicare Payment Advisory Commission. 2011c. Report to the Congress: Medicare payment policy. Washington, DC: MedPAC. Medicare Payment Advisory Commission. 2008b. Report to the Congress: Reforming the delivery system. Washington, DC: MedPAC. O'Malley, A. S., A. M. Bond, and R. A. Berenson. 2011. Hospital employment of physicians: Better quality, higher costs? Issue brief no. 136. Washington, DC: Center for Studying Health System Change. Medicare Payment Advisory Commission. 2010a. Comment letter to CMS on the proposed rule entitled: Medicare program; proposed changes to the hospital inpatient prospective payment systems for acute care hospitals and the long-term care hospital prospective payment system and proposed fiscal year 2011 rates; effective date of provider agreements and supplier approvals; and hospital conditions of participation for rehabilitation and respiratory care services Medicaid program: accreditation requirements for providers of inpatient psychiatric services for individuals under age 21. May 27. http://www.medpac.gov/ documents/IPPS_2011_MedPAC_Comment.pdf. Medicare Payment Advisory Commission. 2010b. Report to the Congress: Medicare payment policy. Washington, DC: MedPAC. Moody's Investors Service. 2011. Not-for-profit healthcare medians show resiliency against industry headwinds but challenges still support negative outlook. August 30. Robeznieks, A. 2010. Building momentum: Construction & Design Survey shows new signs of life for health projects. Modern Healthcare, March 15. Robeznieks, A. 2011. Re-building: Annual Construction & Design Survey shows continuing resurgence in new projects, including work restarted post-recession. Modern Healthcare, March 14. Robinson, J. 2011. Hospitals respond to Medicare payment shortfalls by both shifting costs and cutting them, based on market concentration. Health Affairs 30, no. 7 (July): 1265-1271. Medicare Payment Advisory Commission. 2011a. Moving forward from the sustainable growth rate (SGR) system. Letter to the Congress. October 14. Medicare Payment Advisory Commission. 2011b. Report to the Congress: Medicare and the health care delivery system. Washington, DC: MedPAC. 82 Rosen, A. K., K. M. Itani, M. Cevasco, et al. 2012. Validating the patient safety indicators in the Veterans Health Administration: Do they accurately identify true safety events? Medical Care 50, no. 1 (January): 74-85. Standard & Poor's. 2011. U.S. not-for-profit stand-alone hospital ratios showed improvement in 2010, but significant pressures are on the horizon. August 2. Hospital inpatient and outpatient services: Assessing payment adequacy and updating payments