Using Data to Develop CJR Strategies - 5 Steps to Take Now

About 800 hospitals are staring down the barrel of the Medicare Comprehensive Care for Joint Replacement (CJR) payment program, which will be implemented on April 1. Those hospitals that have gone through the somewhat-arduous process of signing up to receive their claims data and downloading it from the CMS portal now have the opportunity to use that data to begin to understand their CJR episodes of care and to design a strategy for success under the program. The steps below outline the beginning of a data-driven strategy for CJR.

Assess your episode landscape

Most hospitals’ primary CJR cost-saving[1] strategy will involve managing post-acute institutional costs. Information about the non-acute care settings’ costs and utilization had not been available prior to CMS releasing data for its myriad alternative payment programs, so viewing the CJR data may be a hospital’s first glimpse of what actually happens after patients leave the acute care setting.

Every hospital has a unique episode and post-acute care (PAC) landscape.  The landscape is defined by the referral patterns for CJR patients to PAC settings after discharge, and the relative costs of those PAC services.  (A small portion of PAC is for readmissions, which we’ll cover later.)

An example of a report[2] depicting a hospital’s episode landscape is shown below.  The landscape report arrays the costs of each episode from least to most costly, and color codes the separate care components: index admission, PAC provider, readmissions, etc.  Many factors become apparent when the data is reviewed in this format: first, the index admission cost is relatively consistent, reflecting the fact that DRG payments for these services are the same amount within each DRG. Next, the contributions of skilled nursing (green), home health (yellow), inpatient rehabilitation (light blue), and readmissions (red) are the most significant drivers of cost variation moving from the lower-cost episodes on the left to the higher-cost episodes on the right. In the example below, there are very few, if any, cost savings to be gleaned from the episodes in the left half of the graph.  Remember, the Medicare program does not benefit from a hospital’s reduced internal costs (e.g. standardizing implants) because the DRG price/payment is not influenced by those factors.   

Of particular interest in this example is that the hospital refers a large percentage of its CJR patients directly to home, with only outpatient rehab and physician care.  This hospital has little opportunity to create savings for the Medicare program for that lower-cost half of its patient population. Since it already has a relatively low percentage of patients utilizing post-acute institutional services, this hospital’s strategy will need to focus on reducing the length of stay of those patients who utilize SNF services. This may be difficult, though, because the CJR patients receiving SNF services may be the higher-acuity patients whose costs are more difficult to reduce.

One other notable factor here is that because of the low PAC use this hospital likely has an average episode cost that is below the regional average.  Since the initial targets are developed as 2/3 from the hospital-specific baseline and 1/3 from the regional average, the hospital’s average episode cost is initially compared to a target developed primarily from its own low-cost baseline, where it may be difficult to achieve savings. As the baseline transitions to a higher regional proportion the targets may increase.   The transition to 100% regional average targets can’t happen soon enough for this hospital. 

Compare the landscape above to the one below.  Here, all CJR patients are referred to a PAC provider, either institutional or home health, and about 70% are referred to SNF, often in conjunction with inpatient rehabilitation services. This hospital has a lot of opportunity to reassess its PAC referrals and to work with its clinical team establish care pathways and protocols.  This hospital has the potential to achieve significant savings under the CJR program.

Track your PAC utilization

As highlighted in the previous section, the critical component for care management in CJR episodes is PAC utilization. While there will always be some patients whose conditions require skilled nursing or inpatient rehabilitation care, many others can be discharged directly to home.  Once you have assessed the CJR care landscape for your hospital, it is helpful to track the PAC utilization trends – both overall and by operating physician.  The overall landscape and trends of this care can be presented in a single metric showing the percentage of episodes in which the patient is discharged directly home. Many BPCI participants in the major joint replacement episodes develop goals and strategies around this metric, so monitoring it on an ongoing basis can provide critical feedback as to the effectiveness of care management strategies.  Since the CJR program does differentiate and establish separate targets for DRGs 469 and 470, with and without hip fractures, it is important to use the same stratification when evaluating this metric as well.

Further analysis of PAC trends can be performed by examining each PAC setting separately. In the example below, SNF utilization decreased steadily throughout 2013, only to increase significantly during the middle of 2014. Inpatient rehabilitation, which was absent in Qtr 3 2014, increased in Qtr 4 2014.

Compare your episode cost trends to target

By examining the hospital’s CJR cost trends throughout the base period, you will get a better indication of the direction and magnitude of potential performance.  If the quarterly trend in the baseline period is flat or increasing, it may forewarn of potential losses for the participation period – particularly if there are no significant interventions already underway. Average episode costs likely will vary from quarter to quarter, especially for lower-volume hospitals, so assessing and understanding this volatility will prepare you for the degree to which financial results during the CJR performance period will also vary when reviewed over a short term. For this reason it is important to look at a longer trend, not just one or two quarters.  These variations will mask the effects of care management programs, and may inhibit assessment of their effectiveness.

The CJR program differs from BPCI in that targets are prospectively established, and are therefore known in advance of the participation period. In addition, CJR targets are not determined solely from historical costs; instead they are based on an increasing percentage of regional episode average cost. Therefore, participants may already be in a surplus or deficit position at the beginning of the CJR program. Evaluating the performance to these targets, and monitoring it on a consistent basis, allows participants to evaluate their financial results at the start of the program and throughout the participation period.

Review your readmission exposure

For most hospitals, addressing readmissions as a CJR strategy is secondary to evaluating PAC provider utilization. However, for some hospitals, readmission costs for CJR episodes may be significant. Reviewing a detailed readmission report, such as the one below, will indicate whether readmissions are infrequent, and occurring in normally-expected categories, or whether there is a significant incidence of preventable readmissions such as septicemia or joint surgery revisions.

Begin designing your preferred PAC network

It is extremely important for a successful care redesign and PAC strategy to develop a network of preferred providers that will partner with you to provide appropriate levels of cost-effective care.  While the primary objective of a PAC partnership is to reduce the cost of care, forward-thinking PAC providers will realize that these arrangements can increase patient volume, and that substitution of lower-cost settings (i.e., SNF for rehab) can be mutually beneficial to the hospital and PAC provider.  The report below uses the CJR claims data to identify a hospital’s most frequent PAC referrals as well as the most cost effective PAC providers. Reviewing the average episode costs and readmission rates by provider can identify appropriate candidates for a preferred PAC network.


The CJR program provides participants with previously-unavailable data about the post-acute services provided to their patients. This data will be invaluable in designing strategies for cost reductions while improving the quality of care to patients. Use this data to:

  • Understand the use of post-acute services by your patients
  • Review the trends in utilization and costs of PAC providers
  • Compare your historical episode costs to the current targets to estimate the starting point of your financial performance
  • Assess the degree to which readmissions are an issue, and
  • Start identifying providers on which to build your post-acute network.

Throughout the CJR program close monitoring of the quarterly (which will hopefully become monthly) data will allow hospitals to monitor their costs relative to targets, design and implement post-acute network management strategies, and manage readmissions. Understanding what your data tells you is the key to success in bundled payment programs.


[1] In this article “cost” refers to the costs to the Medicare program, which consist of all fee-for-service payments for covered Part A and Part B services to patients during the 90-day episode. We do not address providers’ internal costs of doing business in this article.

[2] A comprehensive analytics tool set such as CJR360 must be used to develop these types of reports.