As hospitals billed Medicare for the most complex care, which indicates that a patient has at least one other major secondary illness like sepsis, the number of stays billed at each of the other severity levels decreased. While the average length of stay decreased for highest-severity codes, the average length of all stays combined remained largely the same, the OIG found.
The government watchdog group offered an example with a hospital that billed Medicare for heart failure and shock with a major complication. Though the mean length of stay for that case is four days, it lasted only two. If Medicare was billed for only a minor complication, the payment would have been closer to $6,200 instead of $9,100.
Collectively, Medicare paid hospitals approximately $14.5 billion for stays that lasted a relatively short amount of time. That is $4.9 billion more than it would have paid if these cases had been billed at the next lower severity level.
Moreover, 54% of the bills for the most complex treatment reached that level because of only one major complication, which is a red flag for potential upcoding, investigators said.
In one case, a hospital submitted a diagnosis for pneumonia and 24 secondary diagnoses, 23 of which were either minor complications or not complications at all. There is a $2,800 difference between the highest severity and next lowest severity code.
"Billions of taxpayer dollars are tied up in billing and coding decisions, and there is significant variation among hospitals that bill at the highest severity," Bryan said, noting that nearly half of the $109.8 billion Medicare spent for inpatient hospital stays in FY 2019 was for stays billed at the highest severity level. "To us, these are all indications that stays with certain characteristics are subject to inappropriate billing practices such as upcoding and merit targeted oversight on CMS' part."
The American Hospital Association said it is reviewing the data brief.
Providers have argued that they had "under-coded" for years and the recent increase in coding severity reflects patients' acuity and complexity more accurately. They also argue that length of stay is not an adequate indicator of patient acuity given that technology and treatment pathways have expedited care.
In response to a 2019 report from the Massachusetts Health Policy Commission that saw similar trends in coding and length of stay, the Massachusetts Hospital Association said that the expansion of electronic health records and the implementation of ICD-10 has allowed healthcare providers to "more accurately and granularly capture patient acuity and previously under-reported conditions."
But upcoding practices precede the advent of EHRs and the emphasis on outpatient care. Prior to the EMR, many emergency department charts were hand-written and had poor documentation, often resulting in "presumptive" coding that was based on the service provided without regard for documentation, research shows.
Throughout the 2000s, the Medicare Payment Advisory Commission and OIG warned that the Medicare Advantage program and evaluation and management services were vulnerable to upcoding. Meanwhile, several healthcare providers reached federal settlements over upcoding allegations.
In 2009, two healthcare entities paid more than $10 million to settle allegations that they fraudulently billed Medicare for more complex evaluation and management services. In 2019, Sutter Health agreed to pay $30 million to settle allegations that the Sacramento, Calif.-based health system submitted inflated diagnosis codes for Medicare Advantage beneficiaries. Prime Healthcare also paid $65 million in 2018 to resolve upcoding allegations.
Emergency services have been another area of concern. Emergency departments in Texas and Oklahoma overbilled for $45.14 of $100 paid for ED services because coding was "higher than reasonable and necessary to adequately care for the patient's needs," a 2012 Medicare audit found.
Envision Healthcare Corp.'s EmCare agreed to pay $29.8 million in 2017 to resolve claims that it admitted patients to the hospital when they should've gone to an outpatient facility.
Similar to the most recent OIG report, the Massachusetts Health Policy Commission found that coding severity had gone up in recent years while length of stay and critical care treatment levels had decreased.
"This is a known phenomenon," David Auerbach, senior director of research and cost trends at the commission, said during an HPC board meeting in 2019 while discussing the report. "There are industries and consultants who have formed to take advantage of these higher payments and higher severity levels."
EHRs have made it easier for providers to add more diagnoses to cases, which increases the severity. Meanwhile, hospitals continue to invest in revenue cycle specialists who focus on "coding optimization."
CMS has tended to respond to "coding creep" by coming up with adjustments in payment rates to offset it, said Paul Ginsburg, Leonard D. Schaeffer chair in health policy studies at the Brookings Institution and director of the USC-Brookings Schaeffer Initiative for Health Policy.
"That works on the average, but the incentive for aggressive coding remains," he said. "Some that have less resources to invest in coding come out behind."
Baylor Scott & White won a False Claims Act case in 2019, when a federal court ruled that the health system's alleged upcoding scheme was consistent with the government's own "encouragement" of hospitals to use the billing codes to glean as much reimbursement as possible from Medicare.
The court cited a CMS regulation that stated that the agency "does not believe there is anything inappropriate, unethical or otherwise wrong with hospitals taking full advantage of coding opportunities to maximize Medicare payment that is supported by documentation in the medical record."
As courts continue to sift through the gray area between "coding optimization" and fraud, the OIG recommended that CMS conduct targeted reviews of highest severity stays and the hospitals that frequently bill for them, particularly when they only list one major complication.
While CMS said that it wouldn't do targeted investigations until there was conclusive evidence linking billing changes and upcoding, the agency said it would share the findings with its contractors and keep an eye out for red flags.
"If the documentation is there, that is the key—all Medicare is getting is the code," Bryan said. "It will take going into the medical record to see what is behind that code."