Evaluating Your First BPCI Reconciliation

Most BPCI participants will undergo their first BPCI reconciliation in October, which will cover the first quarter of 2014. We’ve heard that some participants appear to be basing their ongoing participation in the BPCI program on these results. If a surplus occurs, they’ll stay in the program and may even add additional episode families. But if the results are negative, they’ll be jumping ship faster than the captain of the Costa Concordia.

Randomness Drives Quarterly Reconciliations

But making significant decisions based on a single reconciliation may be quite unwise, for several reasons. First, as we’ve mentioned in previous articles, quarterly average episode costs bounce around like a pogo stick, especially for low-volume DRGs. The table below shows the quarterly variation in average episode costs for four different DRGs. For the lower-volume cardiac valve procedures, quarterly variation could easily exceed 10%, and even for higher-volume DRGs it could exceed 5%. For DRG 216, a participant having its first reconciliation in quarter 3 of 2010 would be shocked to see an average episode cost of about $98,000, which would probably create a significant loss – only to find the exact opposite situation occur in the following quarter– if they stuck around long enough to find out. This may be one reason why CMS is allowing participants to delay the financial settlement following quarterly reconciliations for up to four quarters.

This means that the results of a single quarter are largely attributable to luck. If your financial results are good, you may have simply had a lower-intensity group of patients, while if the results are less favorable you may have had a few more readmissions or a couple of long SNF stays. It takes a longer period for variations to start indicating the actual effectiveness of the clinical initiatives that will cause BPCI programs to succeed.

In addition, the first reconciliation covers the initial period of BPCI participation. Most care management initiatives will not have taken effect at that time. Since reconciliation targets include the 2-3% discount to CMS, participants should expect to see an equivalent amount of loss in the reconciliation, all other factors being equal. Nobody should expect to see a surplus unless they’ve implemented clinical initiatives long enough in the past to be having a significant effect in the reconciliation period.

Alternative Approaches for Determining BPCI Results

Two approaches may be more effective at evaluating the long-term effectiveness of initiatives in BPCI programs. The first is to take a longer look at cost trends. The graph below shows the long-term trend in the average episode costs, indicated by the trendline on the graph. This shows that amidst the quarterly variation the longer-term trend is downward. Because of the lag in the reconciliation process, and the evaluation period allowed, participants may have several quarters of data available subsequent to the reconciliation quarter that may give longer-term guidance of the actual cost trends. Analyses of these trends may be more helpful than reviewing the results of a single reconciliation.

The other approach reflects the fact that long-term improvements in cost-effectiveness don’t occur by themselves. They occur because of effective care management initiatives that reduce readmissions and focus post-acute care on more cost-effective alternatives. The effectiveness of these activities may be measurable on their own. For example, many participants have found that SNFs often retain patients beyond the end of their need for skilled care in order to be paid for the maximum number of covered days. As a result, participating hospitals have begun to encourage referrals to SNFs whose lengths of stay are more based on medical necessity rather than financial gain. In the table below, SNF 65 has a higher average episode cost than other higher-volume SNFs, which may encourage the participating organization to either work with that SNF to reduce length of stay, or alternatively to attempt to direct patients to other SNFs. The effectiveness of this strategy can be measured by evaluating changes in utilization of “preferred” SNFs. 

Another process-based analysis is shown below. This participant created an initiative to decrease post-discharge use of inpatient rehab facilities and instead utilize SNFs. The graphs below show the results of that migration occurring in the first quarter of 2013, with IP rehab costs decreaseing while SNF costs increased, but to a lesser degree.

These process-oriented metrics may provide better guidance of ultimate financial success than individual quarterly reconciliations.

In any case, a single quarter’s financial results rarely gives a true picture of the long-term outlook for bundled payment participation. Participants should look for longer-term trends to materialize, or use other process-related metrics to assess success.