For most readers the results of this analysis will be amazing but true.
EDITOR’S NOTE: Longtime RACmonitor contributor and R1 Physician Advisory Services Vice President Dr. Ronald Hirsch recently pointed site Publisher Chuck Buck in the direction of something that, frankly, wowed him: a PhD thesis from a young Columbia University student on the topic of the costs of Medicare audits. The following article highlights some of her ideas.
Since 2011, hospitals around the country have been facing audits by the Recovery Audit Contractors (RACs). The onslaught of short-stay audits led to the development of the two-midnight rule by the Centers for Medicare & Medicaid Services (CMS), to clarify which patients are appropriate for inpatient admission and which should be treated as inpatients. While CMS has repeatedly stated that this is a complex medical decision to be made by the physician, they have allowed auditors to second-guess things and deny payment for inpatient admissions when the auditors do not agree.
While it is clear that hospitals have expended resources to ensure that patients are placed in the right status, the glaring question is: to what extent should the government monitor for wasteful expenditure through programs like RACs? Monitoring can save government programs money, but it can also be costly to implement and comply with, and in the context of healthcare spending, this could hurt patients’ health as well. In her recent economics PhD thesis, Maggie Shi, a student at Columbia University, studied this in the context of the RAC program. Combining novel administrative data with hospital claims, cost data, and IT adoption data, Shi isolated variation in monitoring across hospitals and across patients to understand its effect on hospital behavior and patient outcomes.
The RAC program was contracted to four firms operating in four separate regions, Shi noted. She analyzed audit intensity and found that it varied across these firms, as illustrated in the sharp contrasts in audit rates across the region borders (demarcated in red in Fig. 1).
Shi compared hospitals on the “low-audit” side of the border to their neighbors on the “high-audit” side, before and after the major expansion in auditing in 2011. Increased auditing deters hospital admissions – in particular, it deters short stays, which Medicare considers more likely to be medically unnecessary. The vast majority of savings comes from deterring future admissions, rather than reclaiming payments on prior ones. But these savings also come at a cost: increased auditing causes hospitals to incur higher administration costs, and to install IT specifically to address issues raised by audits.
Shi wondered if the savings from deterred admissions may also harm patient health, if these admissions were actually necessary. She studied this by using variation in audit likelihood across patients due to the two-midnight rule, which indicated that patients whose entire stay in the hospital crossed two midnights could not be audited by RACs for medical necessity. Because this rule counts midnights, audit likelihood increased discontinuously for patients who stepped foot in the hospital immediately after midnight, since they had just “missed” their first midnight. Comparing patients who arrived at a hospital’s emergency department before and after midnight, she found that once the two-midnight rule came into effect, hospitals were less likely to admit just-after-midnight patients. However, these patients were no more likely to revisit the hospital within 30 days (see Fig. 2). The null effect on revisits, even among patients facing a 25-percent reduction in admission likelihood, suggests that the admissions deterred by monitoring are unnecessary.
If there is no apparent effect on patient care, is there an effect on cost? Shi estimates that an audit rate increase of 1 percent (relative to a mean of 2.2 percent) for a hospital’s inpatient stays results in $2.08 million in government savings, $87,000 in government monitoring costs, and $451,000 in hospital costs (see Fig. 3).
When asked to comment on these findings, Dr. Ronald Hirsch, a RACmonitor editorial board member, stated that “the mathematical modeling used by Shi in this study are beyond the comprehension of most, but her conclusions are sound. Audits save money for Medicare, not only by recouping improper payments, but also by deterring admissions and care that may lead to a claim that has a higher potential of being audited. In other words, hospitals are spending a great deal of money to help save the Trust Fund, with no benefit to themselves.”
Hirsch also pointed to a recent study published in the Journal of the American Medical Association (JAMA) Internal Medicine showing that the CMS Oncology Care Model program, designed to improve outcomes and lower the cost of caring for patients with cancer, led to a reduction in Medicare spending – but the “savings” accrued by the participants did not cover the costs of administering the program, and there was no net improvement in quality or patient experience. “Once again, providers are literally paying the cost of protecting the Trust Fund,” Hirsch said.
Maggie Shi will be a special guest during the Dec. 6 live edition of Monitor Mondays at 10 Eastern. Shi will discuss her thesis; be sure to tune in.