Analytics for Bundled Payments - Overview of the Analysis

Submitted by jonpearce on Tue, 2012-01-10 08:56

This article is an excerpt of Singletrack Analytics' "Analytics for Bundled Payment" white paper.  To request a copy, complete the information on the Contact Us page.

Overview of The Analysis Process

The analysis process for developing the bundled payment proposal is depicted below.  The contracting organization must first identify physicians who will be involved in the care of the patients in the DRGs involved in the demonstration.  Next, the DRGs to be included in the demonstration must be identified and analyzed.  From this analysis, care plans will be developed that target areas of cost that can be reduced or eliminated.  These care plans will form the basis of the budgets to be used in the development of the episode bundled pricing proposal to CMS.

 

Identifying MS-DRGs for Bundling

Contracting organizations are not required to include all MS-DRGs in their bundled payment proposals (except in Model 1 where all inpatient admissions are included).  Instead, they will select MS-DRGs that are most appropriate for bundling in their particular situation.  Critical to success will be the active participation of physicians who treat patients in the selected DRGs, since they will be the primary managers of cost.  Therefore, a primary consideration for selection will be the degree of cooperation that can be expected from those physicians. 

Previous CMS bundled payment initiatives have involved surgical DRGs (cardiac surgery in the Heart Bypass demonstration; orthopedics and cardiac surgery in the Acute Care Episodes (ACE) demonstration), so the initial outreach to medical staff members may be to the surgeons.  However, medical specialists may find an attractive opportunity to work with patients over the longer post-discharge period that is an option in the MBPI, and to develop pathways to improve patient health while reducing cost, yet not suffer a loss of fee for service payment.  In this way, the MBPI can function as a more controlled accountable care organization, involving only patients who have been hospital inpatients, yet providing an opportunity for physicians to develop some population-based care management measures that may be useful in future more aggressive payment models.  Since CMS is clearly interested in expanding bundled payment to medical cases, and particularly those involving chronic diseases, involvement of medical specialists can be important to this initiative.

Primary characteristics of potentially attractive DRGs are:

  • Relatively high volume, as indicated by the number of admissions
  • Relatively high cost (to CMS), based on the payments made to the providers
  • High variations in cost, possibly indicating a lack of consistent treatment protocols, and
  • Variations having common characteristics (patient complications, attending physician, etc.) are particularly attractive since the reasons for the variations can be identified.

Computing the volume and cost of the selected DRGs is straightforward and involves counting the admissions and summing the Medicare payment amount for each DRG.   Quantifying the variation in payment is slightly more complex, but can be computed as the coefficient of variation (CV) of the cost for the DRG (the standard deviation divided by the mean).  This expresses the percentage of variation in cost from the average cost that would be expected to encompass 66% of the cost if the costs were normally distributed, (which they won’t be, but we’ll get into the charge distributions later). 

This means that the DRGs having the greatest opportunity for cost savings will be those that are high-volume and high-cost with a high CV. 

Analyzing MS-DRG candidates for bundling

Once potentially-attractive candidates for bundling have been identified, they should be reviewed more closely by a team focused on each DRG candidate.  These teams should be composed of physicians of the specialty in which the DRG candidates lie, along with clinical and operational staff members who have knowledge of how those cases flow through the treatment process, including post-acute care.  These teams should review each of the cases, particularly those involving readmissions, high costs or unusual levels of physician or post-acute provider utilization.  The team’s goal is to identify areas in which changes to patient care could have maintained quality while resulting in lower levels of utilization or cost. 

The result of this process will be a group of DRGs that have been reviewed by the respective physicians and clinical managers and for which clinical plans of action have been developed to create cost savings. 

Developing proposed budgets for each MS-DRG candidate

At this point the clinical action plans for each DRG must be quantified to develop a proposed budget for the bundled payment initiative.  These budgets are also used in determination of the discounts to be given to CMS in the bundled payment proposal.  Budgets are developed by pricing the revised utilization amounts developed by the clinical team using the appropriate Medicare payment rates for the respective provider types.  These budgets are compared to the historical actual payment amounts to show the estimated savings that could result from the implementation of the clinical plans of action.  This analysis should also indicate the payment reductions that will result for each provider type, since those reductions must be considered in the allocation of savings.

The savings projections from the DRG budgets will form the basis of the proposed discount to CMS.  While CMS has specified minimum discount amounts required[1], larger discounts may be necessary for proposals to be acceptable to CMS.

In addition to understanding the cost drivers, it will be important to develop parameters for further defining the bundling episode beyond the DRG definition.  While CMS is unlikely to allow significant redefinition of the episodes, it is to allow the exclusion of certain types of cases or costs.  For example, it’s likely that CMS will allow applicants to define and exclude readmissions that are clearly unrelated to the primary admission.  Similarly, unrelated physician services may also be excludable.  It is incumbent upon the applicant, however, to define these exclusions and to provide persuasive arguments as to why they should be applied.  This analysis is important, since all costs not specifically excluded will be charged against the bundled payment budget.

The next installment of this article series describes the data to be provided to organizations applying for the bundled payment initiative.  To request a copy of the entire white paper, complete the information on the Contact Us page.

[1] Minimum discount of 3% for organizations proposing a 30-89 day post-discharge episode length; 2% for organizations proposing a post-discharge length of 90 days or longer.