A recurring theme in our blog posts is the need for payment policy to reflect a clearly defined purpose. Start with what you want to achieve and work backwards to a policy that delivers it. The ongoing saga of the CMS 2-Midnight rule is an excellent case study for this principle.
First, some history: CMS was/is faced with two related but distinct challenges. The first is the inexorable rise in observation stays, particularly those stays which exceed 48 hours. CMS observed an increase in the average duration of observation cases exceeding 48 hours from three percent in 2006 to eight percent in 2011. A subsequent OIG study put the rate at 11 percent in 2012. Extended observation creates a knock-on effect for beneficiaries, both for personal liability and qualification for post-acute care benefits. Hospitals blamed the increase upon aggressive Recovery Audit Contractors and rules that only paid for ancillary tests for inpatient admissions deemed unnecessary after review. This is not a determination that the service was not medically necessary, but rather that it need not have required an inpatient stay.
To help lessen the use of observation for lengthy stays, CMS moved to allow payment for those services provided in the inpatient admission should the claim be reclassified to outpatient, excluding the cost of the bed. This move fits neatly with the growing internal pressure placed upon CMS by a raft of successful claim appeals. Another distinct challenge facing CMS is the prevalence of short stay (one day or zero day) cases being paid at full DRG rate. Citing the findings of their CERT contractor that short stay claims had an improper payment rate of 36.1 percent, a policy was required to reduce current levels of payment for short stay cases. And so the 2-Midnight rule was crafted requiring physicians to anticipate the need for at least 48 hours of inpatient care before generating an order to admit.
The hospital industry and physicians have pushed back hard on these regulations and after repeated delays will have their day in court. While it is likely that CMS will prevail, the agency has at least indicated a willingness to entertain other policy approaches in the FY 2015 proposed rule (P806). While most of the comments received by CMS opposing the 2-Midnight rule focus upon the availability of a physician crystal ball, differences in interpretation and efficiency standards (if a hospital has a weekend service available, will it be penalized for discharging quicker?), while still leaving beneficiaries in financial limbo after audits. We thought we would stress something more basic. The 2-Midnight rule is both an overly complex response to a payment issue and it gets in the way of the fundamental operation of the inpatient prospective payment system (IPPS).
Each year hospitals reduce the average length of stay (LOS) of inpatient cases, spurred on by the additional profit levels within DRGs and the ability to reallocate inpatient capacity. Take for instance pacemaker insertion. In 1976 the average LOS for a Pacemaker insertion was 12 days (Thompson, Averill, & Fetter, 1979); by 2013 it had dropped to an average of three days with yet more procedures performed on an outpatient basis (an average of three days means the DRG includes cases with both fewer and more than a three day LOS). If cases with an LOS below two days are subject to added scrutiny and likely reclassification as outpatient claims then, even if the CMS gamble that patients will not be held longer by hospitals to meet the threshold holds true, the incentive to discharge more quickly has been removed. This is not an efficiency incentive.
One alternative that maintains incentives with little administrative overhead is to apply a low-cost policy. Specifically, apply a limit upon the maximum profit any case may generate within a DRG — a concept similar to the operation of an outlier threshold to limit loss. A fixed maximum profit level (payment less cost) would be established for each DRG as a percentage of the base payment for the DRG. This would be constructed in conjunction with a global maximum to address DRGs with particularly high payment levels. For example, a DRG with a base payment of $10,000, after establishing a profit cap of 25 percent, would be paid cost plus $2,500 if costs were less than $7,500. Cases with cost over $7,500 would be paid the DRG rate. A global maximum, say $15,000, could be applied such that no single case could make more than $15,000 relative to cost regardless of the base DRG payment. Shorter stay cases would have higher margins per day, encouraging LOS reduction, but the differential between outpatient and inpatient rates would be reduced while maintaining pressure upon other providers to achieve the ever lower efficiency benchmark.
Such a policy could easily replace other administratively burdensome and emotive policies such as those governing acute and post-acute transfers. Moreover, it can rebalance the windfall DRG profit achieved when patients die or leave against medical advice. Reframing the policy solution to promote the use of lower cost settings and efficient treatment improves significantly upon an administratively burdensome fixation upon short stay cases without regard to underlying IPPS incentives.