Demystifying Hospital Readmissions Penalties

Financial Leadership Council Research Note August 1, 2016 Demystifying Hospital Readmissions Penalties Commonly Asked Questions from Hospital CFOs Highlights Readmissions penalties will likely get larger and more difficult to avoid. The progressive expansion of conditions included under the Hospital Readmissions Reduction Program (HRRP) makes escaping penalties increasingly difficult. A sixth condition, CABG, has been rolled into the program for Fiscal Year 2017 penalties. Readmissions forecasting is difficult and arguably unnecessary. Numerous factors make accurate readmissions projections extremely challenging. Nevertheless, many hospital executives continue to chase the perfect projection. We recommend recent condition-specific excess ratios as a better prognosticator for actionable analyses. There is substantial financial upside for cutting readmissions. Our analysis indicates no financial downside to reducing readmissions for a sub-optimally performing condition, even when the balance between readmissions penalty and contribution margin per case is considered. Background The Hospital Readmissions Reduction Program (HRRP) is a mandatory CMS initiative that incentivizes acute care providers to reduce readmissions by placing up to 3% of a hospital s inpatient Medicare feefor-service revenue at risk in any given year. Over 75% of hospitals will receive a readmissions penalty in FY 2017, with average penalties topping $200,000 per hospital, and the worst performing decile seeing payment reductions in excess of $1M. CMS expects the program will reduce payments to hospitals by over $500M in FY 2017. Understandably, these penalties occupy significant mindshare for hospital CFOs. This research note addresses some frequently asked questions from hospital finance leaders across the past 12 months as they seek to better understand the interplay between the program methodology, the level of penalty administered, and strategies to prevent future readmissions-related payment reductions. Key insights Overwhelming majority of hospitals would financially benefit from cutting readmissions Analysis of recent CMS data indicates that less than 2% of hospitals earn more profit than penalty from their excess readmitted cases. The data indicates that aggressive readmissions reduction targets will likely lead to improvement without causing outsized financial loss. Don t overlook the analytical efficiency gains provided by publicly available data Data lags, incomplete data, and the inherent complexities in calculating readmissions penalties often render a true forecast impossible. However, recent public releases of condition-level readmissions data allows hospitals to perform meaningful analytics and anticipate potential future penalty amounts. To reduce readmissions penalties, you must target the right cases Not all readmissions are equal. In order to reduce penalties, hospitals must inflect sub-optimal performance for the specific conditions measured under the program. Readmissions reductions beyond the scope of the program, while likely a worthy goal, will not impact readmissions penalties in any way. 2016 Advisory Board All Rights Reserved 33442 1

Why should I care about readmissions penalties? CFOs often ask us if they should be concerned about the Hospital Readmissions Reduction Program (HRRP). For most hospitals, the answer is yes. Let s look at some reasons why: 1. The penalties are substantial, and almost everyone gets one. In 2016, almost 80% of hospitals received a readmissions penalty of some kind, while 1 in 6 received a hefty penalty of 1% or more. On average, hospitals were penalized more than $150K per facility. HRRP Penalty Results Penalty Amount Number of Proportion of Total Estimated Average Penalty Received A Penalty 2,625 78.3% ($159,399) 0.5% Penalty or More 1,152 34.3% ($289,928) 1% Penalty or More 511 15.2% ($395,019) 2% Penalty or More 112 3.3% ($527,038) 3% Penalty (Max) 38 1.1% ($309,271) * * Lower dollar amount due to greater proportion of small hospitals. 2. The cumulative impact of penalties adds up quickly year-to-year. Since the program s inception in FY 2013, almost 90% of hospitals have received a readmissions penalty at least once, 75% of hospitals have received a penalty in consecutive years, and more than half have received a penalty every year under the program. Simply put, consecutive penalties can add up quickly the majority of hospitals have already been penalized at least $100K, while hundreds have crossed the $1M threshold. Cumulative HRRP Penalties FY 2013-2016 Cumulative Penalty Number of Proportion of Total $100K+ 1,790 55.3% $200K+ 1,379 42.6% $500K+ 748 23.1% $1M+ 373 11.5% $2M+ 115 3.6% 3. Per hospital penalties are likely to increase as the program expands. The HRRP has steadily expanded since its introduction. The program now includes 5 conditions: AMI, COPD, Heart Failure, Pneumonia, and Total Hip/Knee Arthroplasty. CABG will also be added to the roster starting in FY 2017. Under the program, penalty amounts for each condition are calculated independently, so exemplary performance on one condition cannot by used to offset poor performance on another. Because of this, the introduction of new conditions only increases the likelihood that hospitals will incur a penalty. We saw this occur in FY 2015 when COPD and THA/TKA were added to the program: 20% more hospitals received a penalty and penalty amounts nearly doubled. 2016 Advisory Board All Rights Reserved 33442 2

Medicare Raw Readmission Rates vs. Penalties FY 2013 0.49% 0.47% 78.3% 77.6% Medicare 30-Day Readmission Rate 0.27% 65.1% 65.4% 0.25% Average Penalty Receiving penalty FY13 FY14 FY15 FY16 FY17 THA/TKA, COPD Introduced CABG To Be Introduced Source: Advisory Board analysis of CMS Readmissions Tables, Hospital Compare. The story will likely be the same in FY 2017, when CMS adds CABG to the condition roster and expands the pneumonia cohort definition. Current projections indicate these changes will yield more than $100M in additional penalties, or about $40K per hospital on average. Given this trend, future expansions of the program with similar impacts cannot be ruled out. 4. One year of poor performance can trigger up to three years of penalties. The separation between hospitals with strong readmissions performance and those receiving large penalties is surprisingly slim: a few readmissions either way can make all the difference. Excess Cases vs. Average Penalty Total FY16 Excess Cases (All Conditions) Number of Average FY16 Penalty 1 727 ($22,806) 2-5 554 ($70,966) 6-10 422 ($139,472) 11-20 356 ($261,521) 21-30 153 ($397,206) 30+ 178 ($818,692) As CMS uses 3-year rolling timeframes to determine penalties for each fiscal year, the data indicates that a handful of excess cases over the course of three years, or just a couple cases per year, can earn hospitals hundreds of thousands of dollars in penalties. The corollary effect is that a few unexpected readmissions in any single year can earn hospitals three years of penalties down the line. 5. Readmissions play a broader role under other payment programs. The impact of readmissions extends far beyond penalties earned under HRRP. Readmissions inflate episodic costs, negatively impacting hospitals Medicare Spending per Beneficiary (MSPB), a performance measure currently worth 25% of overall inpatient Value-Based Purchasing (VBP) scores. Higher episodic costs also have consequences under emerging bundled payment models, such as the Comprehensive Care for Join Replacement (CJR) program and the recently proposed Episodic Payment Models (EPMS) for CABG, AMI, and hip fracture, or any of the total cost of care models. 2016 Advisory Board All Rights Reserved 33442 3

Can we forecast our future readmissions performance? CFOs often ask us if it is possible to accurately project future readmissions penalties. Most requests have been for penalty estimates in advance of the public release of CMS s official penalty data, but well after the relevant performance periods have been completed. More recently, we ve also received a number of requests for true readmissions forecasts, which involves anticipating condition-specific readmission case volumes and penalties before the data collection periods have closed. Calculating penalties ahead of time While calculating penalties ahead of CMS s official public release is possible from a mathematical standpoint, there are several reasons why this has not been feasible to date. 1. Calculating readmissions penalties is inherently complex. To calculate penalties ahead of time, CMS s readmissions methodology needs to be replicated. Detailed, enormous, and expensive datasets need to first be acquired, then multiple years of data (million of claims) need to be aggregated. Eligible cases have to be identified, index admissions and readmissions need to be accurately flagged, and CMS s complex risk adjustment methodology needs to be reverse-engineered. 2. You d have to calculate readmissions for everyone. Because each hospital s penalty is measured relative to national performance, in order to estimate one hospital s penalty, performance needs to calculated for every hospital nationwide. 3. The raw data simply hasn t been available. To date, CMS hasn t released datasets with enough detail until well after the final penalties have been announced. For example, CY 2015 claims-level data, which would be needed to estimate FY 2017 readmissions penalties, are not typically available until October of 2016, several months after the final readmissions data has already been released. CMS recently announced it will make claims data available to the public on a quarterly basis. If this release schedule ameliorates current time lags, readmissions penalties could hypothetically be estimated in advance of the final rule each year, assuming calculation challenges can be overcome. Projecting future readmissions performance To create a true readmissions forecast, hospitals need to resolve the issues covered above, but also anticipate readmission events before they happen. This is problematic. As detailed previously, most organizations receive a readmissions penalty based on a handful of excess cases. Such a slim margin for error dramatically increases the chances of penalty projections being off, potentially by hundreds of thousands of dollars. Further complicating matters, because each hospital s penalty is calculated relative to national performance, creating a readmissions forecast requires the accurate projection of readmissions across all hospitals. How can we anticipate or examine future penalties short of a full forecast? There is a far simpler method for anticipating your hospital s future readmissions penalties look at your performance from last year. Based on our analysis of readmissions statistics released by CMS, there are a few simple, yet powerful indicators that can be calculated from the data that should allow hospitals to anticipate whether they are going to receive a penalty and understand its potential magnitude. Will my hospital receive a penalty next year? Over the past four years, 90% of facilities that received a readmissions penalty in one year ( base year ) also received a penalty the following year. This relationship holds true at the condition level: the majority of hospitals received repeat penalties for poor performance on the same conditions (table on next page). 2016 Advisory Board All Rights Reserved 33442 4

Readmissions Penalty Recurrence Rate FY 2013-2016, by Condition Readmissions Penalty in Base Year Readmissions Penalty in Subsequent Year Recurrence Rate Overall 7,038 6,371 90.5% AMI 3,120 2,294 73.5% HF 4,364 3,407 78.1% PN 4,200 3,233 77.0% COPD 1,356 1,039 76.6% HIP-KNEE 722 525 72.7% Source: Advisory Board analysis of CMS Readmissions Tables, Hospital Compare. Another simple, but strong penalty indicator is a hospital s volume of excess readmitted cases from the prior year. Per the table below, unless the hospital had two of fewer excess cases last year, they are almost guaranteed to receive a penalty the coming year. Probability and Magnitude of Penalty Based on Excess Cases in Previous Year Total FY 2015 Excess Cases Number of Received Penalty in Average FY16 Penalty (%) Average FY16 Penalty ($) 0 697 32.3% -0.28% ($30,802) 1-2 512 84.4% -0.49% ($73,761) 3-5 466 94.8% -0.65% ($141,030) 6-10 363 96.8% -0.81% ($224,095) 11-30 150 99.7% -1.02% ($335,798) 30+ 208 100.0% -1.31% ($721,477) Source: Advisory Board analysis of CMS Readmissions Tables, Hospital Compare. How big will my penalty be next year? Looking back at the historical data, there is a clear correlation between excess cases one year and the magnitude of the penalty received the following year, both from a percentage and from a dollars standpoint. While this picture will obviously differ hospital to hospital, because the historical data has such a strong correlation with penalties earned the following year, it is clearly valuable, especially given the availability and reliability of the data compared to any readmissions forecast. Should we reduce readmissions? Won t that cut into my hospital s revenue? A common question we ve received is whether it makes financial sense for hospitals to reduce readmissions. This idea considers readmissions in a vacuum and appears to stem from the belief that lowering readmissions-related case volumes will decrease the available contribution margin, and thus reduce overall facility revenues. Our analysis of readmissions data indicates that the financial benefit from reduced readmissions almost always outweighs the sacrificed contribution margin when that condition has an excess ratio greater than one. Before discussing our findings, it s important to note the broader implications of reducing avoidable readmissions. The conditions in the HRRP were chosen due to the avoidable nature of a plurality of readmissions by each condition-type, often resulting from inadequate coordination of post-hospitalization care or follow up. In other words, the ultimate goal of HRRPs is to improve patient care. Excellent readmissions performance will also positively impact outcomes in other programs, namely, hospital inpatient VBP, mandatory bundled payments, and any of the ACO-related programs. 2016 Advisory Board All Rights Reserved 33442 5

Focus on the right set of readmissions Not all readmissions can be considered equal for the purposes of the HRRP. To minimize penalties, hospitals need to focus on patients that fall into the diagnostic categories that CMS has identified (AMI, HF, PN, COPD, THA/TKA, and CABG starting in FY 2017). Readmissions for all other patients do not contribute to penalties under the HRRP. When looking at cases for patients with those conditions, it is important to compare the right values. Penalty dollars come right off the top of hospital earnings. To offset a dollar of penalty, a hospital has to earn an additional dollar of contribution profit. We have found that many of our members are mistakenly comparing penalty amounts to potential revenue loss when trying to decide whether to cut cases, which considerably skews the calculation. With that in mind, there is a simple rule of thumb for determining the financial efficacy of lower readmissions. If excess readmissions are earning your hospital more in penalties than they are contributing to your profit margin, then it makes financial sense to improve your readmissions performance on those conditions. National Statistics for Cases with HRRP Conditions Condition Avg. Penalty (per excess case) Avg. Cont. Profit (per case) Avg. Net Loss (per excess case) Overall -$21,371 $4,046 -$17,325 AMI -$27,584 $5,389 -$22,195 COPD -$11,245 $3,568 -$7,677 HF -$12,167 $3,640 -$8,527 PNE -$12,530 $3,556 -$8,974 THA/TKA -$98,598 $4,722 -$93,876 The table above illustrates that the penalty dollars saved by cutting excess readmissions almost always outweigh potential contribution profit losses. By our estimates, only a handful of organizations across the country have per-case profit margins large enough to justify keeping those readmissions, from a strictly analytical standpoint. with Greater Average Contribution Profit than Penalty per Case Condition Receiving Penalty w/ Cont. Profit > Penalty per Case % w/ Cont. Profit > Penalty per Case AMI 1,071 8 0.7% COPD 1,403 12 0.9% HF 1,458 22 1.5% PN 1,462 24 1.6% THA/TKA 1,241 0 0.0% This is an important finding for hospital financial leaders that may be pondering whether sacrificing readmissions-related revenue in the chase for lower readmissions rate results in a net financial loss. The data indicates that, in almost all situations, reducing excess readmissions promises greater financial upside, in line with the goal of the program. In an extreme situation, our modeling indicates that pushing readmissions up to the 3% penalty cap and continuing to readmit to out-earn the penalties is almost unattainable, as the number of cases required to offset such a large penalty is virtually impossible to attain. 2016 Advisory Board All Rights Reserved 33442 6

My organization wants to cut readmissions. What should our target be? Strictly from a financial perspective, it is ideal for a hospital to reduce readmissions to the point where they have zero excess cases, but no further. Under this scenario, the hospital would avoid penalties completely while still receiving the revenue and contribution profit associated with the remaining cases. For most hospitals, such accurate targeting is difficult, if not impossible, due to the inherent unpredictability of readmissions. So which is the better option: reduce readmissions carefully and keep a few excess cases, or aggressively cut readmissions and lose out on additional case revenue and contribution profit? In the example below, we model out the financial impact of hospitals playing it safe reducing their readmissions volumes by twice the amount needed to avoid a penalty. By doing, so they would avoid a readmissions penalty, but would potentially lose twice as much contribution profit. Modeled Impact of Reducing Readmissions by Twice Excess Case Volume Condition Excess Readmissions Readmission Case Reduction Net Financial Impact Net % of Penalty Recouped Overall 10-20 +$165,441 80.0% AMI 3-6 +$19,799 77.6% COPD 4-8 +$14,636 69.0% HF 6-12 +$22,241 66.3% HIP-KNEE 3-6 +$65,602 87.3% PN 4-8 +$14,279 71.0% Even in this example of significant over-correction, overall profitability improves substantially. From a target-setting perspective, this data indicates that hospitals should set aggressive goals for reducing readmissions volumes. Even if you overcut, your margins will still likely improve. We have decreased our readmissions case volumes. Why hasn t our penalty gone down? As hospital leaders focus on improving their facilities readmissions performance, many are dismayed to find out that their penalties have not declined. There are several reasons why a hospital may reduce their overall readmissions volumes but still receive an equal, if not greater, penalty the following year: 1. Recent improvement hasn t been sufficient to offset past performance. The penalty a hospital receives in a given fiscal year is determined by readmissions to the facility 2-5 years prior. For example, penalties are based on readmissions between July 1, 2011 and June 30,, 2014. Thus, it is possible an organization could improve their readmission performance in a single year and not see any penalty reversal for several years. 2. National readmissions are improving faster than your hospital s rates. HRRP penalties are determined on a curve: Hospital performance is compared to national average readmission rates. It is therefore possible for a hospital to improve in absolute terms, yet not make up ground relative to the rest of the nation. In such a case, penalties may even increase from prior years. 3. New conditions are causing penalty spikes. The introduction of new conditions in FY 2015 (COPD and THA/TKA) and FY 2017 (CABG) may be spiking your readmissions penalty, even though your hospital has improved on other conditions. It is important to keep in mind that the program is essentially multiple separate penalties: one for each condition. As the roster has expanded, so has the potential for larger overall losses. 4. You can t see all your readmissions and you re not actually improving. On average, around 30% of patients are readmitted to a different hospital than they were initially admitted to. That subset of readmissions, called other site readmissions, has stayed relatively constant even as hospitals have reduced their readmission rates overall (table on next page). 2016 Advisory Board All Rights Reserved 33442 7

A reduction in same site readmissions essentially, those patients discharged from your hospital that are readmitted to your hospital is an inadequate measure of improved readmissions performance. As already discussed, a small number of cases may be sufficient to drive readmissions penalties, meaning that other site readmissions could contribute markedly to overall penalties. This is important to recognize, as many executives have expressed surprise that penalties have failed to abate despite seemingly successful readmissions reduction programs at their home base. Other Site Proportion of Total Readmissions CY 2011 2014 Condition CY 2011 CY 2012 CY 2013 CY 2014 Overall 30% 30% 31% 31% AMI 48% 48% 47% 47% CABG N/A N/A 29% 28% COPD 22% 23% 24% 24% HF 28% 29% 30% 30% PNE 27% 28% 29% 30% THA/TKA 36% 36% 36% 35% Methodology The analysis in this note incorporates final readmissions penalties (Table 15) from FY 2013-FY2016, released in conjunction with CMS s Final Inpatient Rule annually. FY 2017 Final Rates were not available at the time of this analysis. Modeling was performed using of excess ratios and condition-level volumes data from Hospital Compare. Analysis of same and other site data was completed using multiple years of claims level data via CMS Standard Analytical Files from CY 2011-CY2014. All financial detail incorporated in this note is based on FY 2015 MedPAR base operating rates. The information in this note is accurate as of its publishing date but is subject to change. If a change should occur, an updated note will be published. Contacts Alex Tallian Senior Analyst talliana@ Eric Fontana Managing Director fontanae@ Christopher Kerns Executive Director kernsc@ Related Resources To explore these topics in more depth, access the following related resources: Customized Readmissions Modeling Please email analytics@ Pay for Performance Map FY 2017 https://dag./2013_g_dag_gmap/home/mapview?mapname=p4p Value-Based Care Consulting https://www./consulting/value 2016 Advisory Board All Rights Reserved 33442 8