Making Sense of Your First BPCI Reconciliation

Submitted by jonpearce on Sun, 2015-12-06 10:31

BPCI participants who started in April 2015 are now eagerly awaiting their first reconciliation, which is scheduled to be released by January 1. Here's an overview of what you'll get with that first reconciliation.

How reconciliation works

First, let's take a brief look at how reconciliation works. In the reconciliation process CMS computes the final (as of the date of the reconciliation) cost of the episodes in each DRG in  which you’re participating, and compares those costs to the then-current target rates for those DRGs. If your aggregate costs are lower than the targets, CMS owes you the difference. If those costs exceed the targets, you’ll be writing CMS a check. Reconciliations are computed individually for each DRG, but are aggregated across all DRGs and participation periods being reconciled at that point in time.

This process occurs on a quarterly basis and reconciles episodes that occurred three quarters in the past. The table below shows the reconciliation schedule for participation quarters in 2014.



First Episode Starts

Last Episode Starts

Last Episode Ends


2014 Q1





2014 Q2





2014 Q3





2014 Q4





In addition, each performance quarter is reconciled three more times, called “true-ups”, to account for additional claims that are paid after the initial reconciliation as well as for other factors.  These true-ups occur for each of the three quarters following the initial reconciliation, so the 2015Q2 performance period will be reconciled again in April, July and September 2016. The third true-up (which is the fourth reconciliation) is final – it’s the only time that the actual reconciliation amount is known for certain. For more information about the true-up process and how it can affect the reconciliation amounts, refer to our blog on “Unravelling the Mysteries of BPCI Reconciliation”,

How you receive the reconciliation data

The actual reconciliation files will be released to your CMS Electronic File Transfer (EFT) portal. If you've activated your EFT account you can download the reconciliation files in the same way as you download the monthly data. Otherwise your analytics consultant or convener should send you those files.

There are two different sets of files in the reconciliation data. The first is a set of Excel worksheets showing your actual reconciliation results. The second is a set of claims data for the performance period and for the then-current baseline period from which the targets are derived.

The Reconciliation Spreadsheet

In its upper left corner the reconciliation spreadsheet shows a summary of the Net Payment Reconciliation Amount (NPRA) due to or from CMS for each initial reconciliation, as well as the subsequent true-ups for each performance period. If this is your first reconciliation there will only be a single performance period, with no subsequent true-ups. As more quarters are reconciled, each reconciliation and true-up will, and will be aggregated into a single NPRA amount. This summary shows the “uncapped” NPRA, along with several calculations that “cap” the NPRA at 20% of the target price.  These calculations are primarily implemented for “conveners” who have amassed large groups of BPCI participants for whom they share financial risk, and their purpose appears to be to limit the amount that these conveners can earn from the aggregate of those NPRAs. For most participants this “capping” is a non-issue.

The lower part of the reconciliation shows the detail for each DRG. Here you'll see the episode counts, average episode costs, target amounts and NPRA for each DRG. The sum of the NPRAs for each DRG is the total amount due to or from CMS for this performance period.

The columns of this spreadsheet section show the derivation of the NPRA. Columns A thru D show the DRG, episode length and risk track for the participation period being reconciled. Columns E thru J show the calculation of the target rate, which originates from each participant’s base period cost but needs to be adjusted by the appropriate trend factor and further adjusted by the local wage index. Column K shows the number of episodes in the performance period, while column L shows the total target amount (the per-episode target multiplied by the number of episodes). Column M shows the total actual episode cost for each DRG, and column N applies the appropriate Winsorization (outlier adjustment) factors to adjust episode costs that exceed the outlier limits. Column O subtracts the adjusted episode cost from the target to get the NPRA for each DRG, which is summed for the total for all DRGs.

In subsequent reconciliations the true-ups from previous performance periods will be shown below the calculations for the current performance period. These results will be combined to give the net amount due to or from CMS for this reconciliation.

The reconciliation data files

The reconciliation data files provide the detail behind the spreadsheets. The data is provided in the same format as is the monthly claims data, with the exception that the episode cost amounts are calculated on the Episodes file; those amounts are not calculated in the monthly data. This data is provided for the periods being reconciled, as well as the current calculation of the baseline period (see “Unravelling the Mysteries of BPCI Reconciliation” for reasons that the base period episode costs can change between reconciliations). This claims data provides significant insights into the reconciliation process, and why performance may have changed from the latest monthly reports provided by your convener or analytics consultant. Differences in episode counts, and in claims costs between the monthly reporting period and the reconciliation period, can be determined by comparing the respective claims files for those periods. In this way, the reasons for all differences between monthly results and reconciliation results should be identifiable.

Checking BPCI calculations

One very useful aspect of the reconciliation files is that it allows checking the accuracy of your analytics consultant’s programming. In the monthly data the episode costs are not computed – the analytics programs must compute those amounts by implementing the logic described in the BPCI documentation. But using the reconciliation data allows verification of the accuracy of those calculations against CMS’s calculation for each episode, so review of validation reports from your analytics consultant can provide comfort that their calculations are correct.

Ongoing analyses

As you continue in BPCI you’ll start seeing the true-ups for performance periods that have been already been reconciled, and you’ll probably be curious as to why a true-up changed the NPRA from the previously-calculated amount. These changes can occur for two reasons – changes in episode costs and changes in targets. Fortunately the reconciliation data provided with each true-up provides the information needed to dig into these factors and identify the reasons for changes. Your analytics consultant may provide a comprehensive set of reconciliation analytics reports, or you can prepare them yourself. In some cases we’ve found significant differences in claims cost between true-ups (large claims were paid that increased the cost of episodes and thereby reduced the NPRA), as well as changes in targets resulting from application of episode precedence rules or changes in program policies. Understanding these changes are important to assessing the accuracy of each NPRA, and determining its validity.