PCI use drops in states with public reporting
Medicare patients admitted to the hospital with acute MI are less likely to receive PCI in states that require public reporting of PCI outcomes, according to a study published Oct. 10 in the Journal of the American Medical Association. The data raised red flags about possible underuse, the lead author told Cardiovascular Business in an interview.

Karen E. Joynt, MD, MPH, of the Harvard School of Public Health in Cambridge, Mass., and colleagues collected data from fee-for-service Medicare patients admitted to hospitals between 2002 and 2010 with a diagnosis of acute MI. The authors compared PCI and mortality rates from states that require public reporting of PCI outcomes (New York, Massachusetts, and Pennsylvania) with states in the region that do not have public reporting requirements (Maine, Vermont, New Hampshire, Connecticut, Rhode Island, Maryland and Delaware).

Massachusetts initiated a reporting requirement in 2005; Joynt and colleagues also performed a longitudinal analysis of changes in PCI and mortality in Massachusetts over time.

The authors analyzed the records of 97,802 patients with a primary diagnosis of acute MI (49,660 patients from reporting states and 48,142 from nonreporting states).  Using ICD-9 codes, they identified three subgroups of patients:
  • Non-ST-segment elevation MI (NSTEMI);
  • ST-segment elevation MI (STEMI) ; and
  • Cardiogenic shock or cardiac arrest.

The authors also divided the sample into age groups; those between the ages of 65 and 74, and those aged 75 or older.  There were similar proportions of each subgroup in the reporting and nonreporting states.

Joynt and colleagues found that in 2010, 37.7 percent of patients admitted with a primary diagnosis of acute MI received PCI in reporting states, vs. 42.7 percent in nonreporting states. The difference was most marked among the sickest patients:  61.8 percent of the STEMI group in reporting states received PCI vs. 68.0 percent in nonreporting states, and 41.5 percent of cardiogenic shock/cardiac arrest group received PCI in reporting states, compared with 46.7 percent in nonreporting states.

In the pre-reporting period in Massachusetts (prior to 2005), 40.6 percent of acute MI patients received PCI vs. 41.8 percent of patients in other nonreporting states. Beginning in 1996, the rate began to decrease, again with the most pronounced decrease among the most seriously ill patients.

Before reporting, 59.7 percent of the sample patients in Massachusetts received cardiac catheterization compared with 63.2 percent in nonreporting states; after reporting began in Massachusetts, 59 percent of the Massachusetts sample received cardiac catheterization vs. 67percent in nonreporting states, "suggesting that some decisions not to proceed took place prior to seeing the patients' coronary anatomy," according to the authors.

The study found no statistically significant differences in 30-day mortality among patients in reporting states vs.  nonreporting states. However, there was "a trend, not statistically significant, toward increasing mortality among patients in the reporting states," said lead author Joynt in an interview with Cardiovascular Business.  "The STEMI group saw the biggest reduction in PCI and that's also where we saw the biggest mortality bump. STEMI patients had 35 percent higher odds of mortality in reporting states. 

"As opposed to overuse of PCI in the stable outpatient population, our concern is underuse in patients who could benefit most from this life-saving procedure," Joynt said.

In an accompanying editorial, Mauro Moscucci, MD, MBA, of University of Miami Miller School of Medicine in Miami, acknowledged that denial of care to high-risk patients was a concern when outcomes are subject to public reporting. But he suggested that better recognition of medical futility may account for at least part of the decline in PCI procedures in reporting states.

Regarding the lack of significant differences in mortality despite the decline in PCI, Moscucci introduced the possibility that "risk adjustment requires optimal coding of comorbid conditions and is subject to gaming through upcoding, thus leading to observed outcomes that are better than predicted." 

Moscucci noted that in 2010, Massachusetts began allowing exceptions to reporting in cases where the patient meets criteria that deems him or her as being at "exceptional risk." In her interview, Joynt called that policy "a step in the right direction."

Joynt pointed out, "It's all observational data. It's hard to say that (the decrease in PCI and the trend toward higher mortality in reporting states) is all causal, but the results raise a concern.

"The take-home message of the study is, let's look at the principles behind the public reporting policies—transparency, quality improvement, accountability. Let's push these principles forward by tweaking the policies to make them more fair and the reporting more accurate."