Hospital Diagnostic Coding Trends for Heart Failure: No Observed Gaming Effect

January 16, 2019

An analysis of American hospitals showed little evidence that recent changes in diagnostic coding patterns for patients hospitalized with heart failure (HF) have affected a key performance measure that influences hospital rankings.

Coding for a principal diagnosis in some critically ill patients partially shifted from acute HF to acute respiratory failure (ARF), with HF coded as a secondary diagnosis, from 2006 to 2014.

However, the small but significant trend was not correlated with changes in HF risk-standardized mortality rate (RSMR), which — unlike RSMR for ARF — is a hospital performance metric.

"Theoretically, hospitals could take their sickest heart failure patients and say they have acute respiratory failure instead of heart failure as the principal diagnosis. And then those patients are no longer tracked by traditional quality measures," Allan J. Walkey, MD, MSc, Boston University School of Medicine, told | Medscape Cardiology. "And that's what we found."

Moreover, hospital mortality was more than four times greater in patients hospitalized with HF but given a principal diagnosis of ARF, compared to a principal diagnosis of HF.

But there was no sign the hospitals were trying to game the system to improve their rankings; if they were, Walkey observed, they did it poorly. Code shifting was "fairly rare," and the analysis didn't show evidence that the practice led to a decline in RSMR for HF.

"It doesn't appear that particular mechanism is affecting hospital rankings," at least through 2014, said Walkey, lead author on the analysis published online January 7 in the Journal of Cardiac Failure.

He and his colleagues identified 1,368,816 hospitalizations of adults with HF at 646 hospitals in administrative claims data from Premier, which is said to cover 20% of hospitals in the United States.

Of the total group, 97.6% had been assigned a principal diagnosis of HF using International Classification of Diseases, 9th Edition, Clinical Modification (ICD-9-CM) codes for diagnoses and procedures. The remaining 2.4% had been assigned a principal-diagnosis code for ARF, with HF as a secondary diagnosis.

The prevalence of a principal diagnosis of ARF rose from 0.4% of hospitalizations with HF in 2006 to 3.4% in 2014.

Of note, those with principal diagnosis of ARF and a secondary diagnosis of HF were in a number of ways more critically ill than those with principal diagnosis of HF. They were more likely to require ventilatory support, for example; noninvasive ventilation was used in 39.2% and 15.2% of cases, respectively, and invasive mechanical ventilation in 73.4% and 14.0%, respectively.

Hospital mortality was significantly higher for patients with a principal diagnosis of ARF and a secondary diagnosis of HF than it was for those with a principal diagnosis of HF (11.7% vs 2.6%; < .001). Risk-adjustment did not substantially change those numbers, Walkey said.

Over the study period, hospital mortality for patients with a principal diagnosis of HF fell by 2.9% annually, whereas mortality for those with a principal diagnosis of either HF or ARF fell by only 1.7% per year (P = .01).

Code shifting from HF to ARF for the principal diagnoses was not significantly correlated with the temporal change in hospital RSMR (= .47). And hospital RSMR was not significantly correlated with the proportion of HF hospitalizations that were coded with HF as the principal diagnosis (P = .24), the group reports.

Walkey pointed out a few caveats to the analysis. For example, adjustments to derive RSMR accounted only for baseline features and medical history; in-hospital events were not included, as they are for RSMR used by the Centers for Medicare and Medicaid Services. And the findings cannot be extrapolated to hospital comparisons by 30-day postdischarge mortality.

Walkey and the other authors report no conflicts of interest.

J Card Failure. Published online January 7, 2019. Abstract

Follow Steve Stiles on Twitter: @SteveStiles2. For more from | Medscape Cardiology, follow us on Twitter and Facebook.





Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: