The Facility Level Implosion: What's Really Behind the Numbers?

Jim Strafford , CEDC, MCS-P

Principal

Published November 26, 2013

The Centers for Medicare and Medicaid Services' (CMS) proposed assault on emergency department (ED) and outpatient facility levels has been well documented in Perspectives on the Acute Care Continuum. CMS claims its proposed move to a single facility level will simplify claims management and coding, while also deterring some of the "upcoding" CMS claimed hospitals were undertaking.

However, recent increases in acuity levels coded by EDs nationwide — and the costs associated with that increase — are certainly additional factors.

The fact that CMS failed to develop guidelines for ED facility coding — or adopt sound guidelines created by private industry (Lynx) and professional societies (ACEP) — only added to the pressure facing hospitals. Coincidentally, we at Healthcare Administrative Partners (HAP) have been tracking trends in ED leveling for several years; and there is no denying that the acuity of ED levels (99281-5) has increased on both the physician and facility side during this time.

When the "coding implosion bomb" was dropped by CMS, HAP personnel were completing research for a white paper analyzing ED facility coding trends. Surprisingly, our data pointed to a different conclusion than the one drawn by CMS.

Rather than grabbing for a bigger piece of the payment pie, hospital EDs that had historically undercoded their acuity levels now appeared to be correcting those errors. Their overall acuities were moving upward to become more in line with national benchmarks.

What's more, the trend worked both ways. Hospitals that had perhaps coded too high in the past were now adjusting their acuity mix downward toward national benchmarks.

To us, this suggests that information technology wasn't encouraging facility level "upcoding." Instead, access to better data and decision-making tools appeared to encourage appropriate coding. And as a result, the system was slowly moderating itself.

Underreporting of Acuity by Hospital EDs

From publically reported Medicare data, we selected outlier hospital EDs for review. Our sample included 92 "underachievers" whose facility coding was well below national acuity benchmarks.

This table shows the shift in acuity mix for our underachieving hospital EDs in 2009 and 2011 (the two most recent years with Medicare data available). Percentages represent the average proportion of cases coded at that level across all 92 underachieving facilities:

99282

(Low-moderate acuity)

99284

(Moderate-high acuity)

99285

(High acuity)

2009

45%

13%

2011

36%

27%

Keep in mind that most Medicare patients are older and more likely to report to the ED with high-acuity conditions. Given that 99282 is the second-lowest ED level (and usually reserved for minor injuries and rashes), it is very likely that hospitals that were coding it 45 percent of the time (and in some cases as often as 64 percent) were undercoding and could be eligible for refunds from Medicare.

Conversely, level 99284 represents a relatively high-intensity service (although patients at this acuity level are seldom admitted), and we should expect it to be the majority code for the Medicare population. However, in 2009, our underachieving hospitals were coding 99284 only 13 percent of the time. The increase to 36 percent in 2011 probably represents an increase in coding accuracy.

Increases in coding of 99285 (the highest hospital ED acuity level excluding Critical Care) were well documented by media and government watchdogs. "Overcoding" of 99285 by both physicians and facilities is currently under OIG and CMS scrutiny. However, when we look at historical data, it appears that these EDs have been undercoding 99285 for many years; and even the shift upward in 2011 is relatively low when you consider that ED physicians code 50 percent of their cases at level 99285.

Also, there is an important distinction between emergency department physician and facility coding. When Medicare patients are admitted to the hospital from the ED, their facility payments are bundled into the hospital DRG bill. So for the same hospital, facility coding of 99285 should be less frequent than physician coding at that level.

Overreporting of Acuity by Hospital EDs

Our study also looked at 57 EDs that had very high acuity metrics compared to national benchmarks:

	99285 (High acuity)
2009	45%
2011	35%

This "overachieving" sample included 10 hospitals that coded over 50 percent of visits at the highest level (99285).

The change here, while not as dramatic as for the underachieving group, is still significant. In 2011, 68.2 percent of the overachieving hospitals coded 99285 less frequently than in 2009. What's more, nine out of the 10 hospitals that coded 99285 more than half the time in 2009 dropped below 40 percent by 2011.

Interestingly, one of the EDs cited in the overbilling reporting, Baylor Medical Center, actually decreased coding of 99285 from 53 to 33 percent during this period.

Conclusion

Our findings support the idea that ED levels have increased. However, among hospitals that have historically undercoded, this shift appeared appropriate (and in fact brought them in line with CMS guidelines). Correction among "overcoders" was less dramatic, but still suggests that ED facility level coding was moderating toward the high middle.

Despite a lack of guidelines, the industry appeared to be righting itself.

It does not appear to us that CMS has a strong case to collapse the five ED levels into a single facility code. Such a move would have an adverse effect on hospital EDs serving a higher acuity patient mix — including trauma centers and academic institutions.

Instead of collapsing to a single level, we believe that CMS should instead implement workable facility coding guidelines. Under that scenario, we believe the ED levels coded would continue to moderate.

To read HAP's complete white paper on facility level coding, please contact Jim Strafford.

Partnering to improve patient lives

Vituity branding orange wave pattern background