Registry Finds Wide Variation in Exclusion of STEMI Patients from D2B Time Reporting

Download this article's Factoid (PDF & PPT for Gold Subscribers)

The percentage of patients with ST-segment elevation myocardial infarction (STEMI) who experience ‘non-system delays’ to primary percutaneous coronary intervention (PCI) and thus are excluded from door-to-balloon (D2B) time reporting varies widely from hospital to hospital, according to a large registry study published online April 17, 2014, ahead of print in the American Journal of Cardiology. The findings suggest a need for more explicit and uniformly applied exclusion rules to avoid potential ‘gaming’ of the system, the authors say.   

Investigators led by Michael C. Kontos, MD, of Virginia Commonwealth University (Richmond, VA), analyzed data from 43,909 STEMI patients admitted to 294 hospitals participating in the National Cardiovascular Data Registry’s (NCDR’s) ACTION Registry-Get With The Guidelines (GWTG) database from January 2007 through March 2011.

A median 9.2% of patients were excluded from D2B reporting due to non-system delays, but the percentage ranged widely among institutions from 0 to 68%. Hospitals were categorized according to tertiles of percentage of excluded patients:

  • Low (≤ 7%; n = 100)
  • Intermediate (7.1%-11.2%; n = 96)
  • High (> 11.2%; n = 98)

Hospitals in the low-exclusion group excluded 4.0% of patients compared with 9.1% in the intermediate group and 16.8% in the high-exclusion group. Absolute differences in age and risk factors among the 3 groups were small and probably not clinically significant, the authors say.

The most common reasons for patient exclusion were:

  • Cardiac arrest and/or need for intubation prior to PCI (38.4%)
  • Difficulty crossing the culprit lesion (19.8%)
  • Difficult vascular access (7.6%)
  • Delays in providing consent (5.4%)
  • Other (28.9%)

Differences in rates of exclusions related to difficult vascular access and delays in providing consent were relatively minor (< 2%) among all hospital groups. In contrast, hospitals in the low-exclusion group were more likely than the intermediate or high groups to exclude for cardiac arrest and need for intubation (42.2% vs 40.5% and 35.7%, respectively), while those in the high-exclusion group were more likely than the low or intermediate groups to exclude for difficulty in crossing the lesion (22.5% vs 17.0% and 17.4%, respectively).

There were no differences among the 3 exclusion groups in rates of hospital annual STEMI or primary PCI volume, bed size, location (rural vs non-rural), or teaching hospital status (academic vs nonacademic).

The overall D2B time was 65 minutes and decreased to 64 minutes with exclusion of patients with non-system delays. When all patients were included, D2B times were faster in low-exclusion compared with high-exclusion hospitals. Although after exclusions, due to differences in exclusion rates, D2B times were similar among the 3 groups with an absolute difference of only 1 minute.

Overall, 82.9% of patients had a D2B time ≤ 90 minutes. The percentage falling within that window increased to 86.2% after exclusion for non-system delays (P < .001). When all patients were included, hospitals in the low-exclusion group had the highest proportion of those with D2B times within 90 minutes When patients with non-system delays were excluded, hospitals in the high-exclusion group had the highest proportion within the recommended times (table 1).  

Table 1. Percentage of Patients with D2B Times ≤ 90 Minutes by Hospital Exclusion Rate


All Patients
(n = 43,909)

Low Exclusion
(n = 13,441)

Intermediate Exclusion
(n = 18,090)

High Exclusion
(n = 12,738)

P Value

Excluding Patients with Delay





< .001

Including Patients with Delay





< .001

Overall, more than 90% of STEMI patients with a D2B time within 90 minutes were seen at just over one-fifth (20.1%) of hospitals. But when patients with non-system delays were excluded, that percentage doubled (P < .001); similar upgrades were seen across the exclusion groups (all P < .001).

What Should Count as a ‘Non-System Delay’?

According to the authors, some factors that can lead to prolonged D2B times, such as the need to exclude serious comorbidities or stabilize a patient prior to intervention, are not under the control of the primary-PCI team and therefore can be considered appropriate, while others, such as difficult vascular access, relate more to the skills of the physicians and are questionable.

“The rationale for excluding patients undergoing primary PCI who have non-system delays is that it allows for more accurate comparisons between centers that have patient populations that can vary considerably in complexity and risk,” Dr. Kontos and colleagues explain. To enable consistent comparison and reporting among institutions, “exclusion rules should be explicit and applied uniformly,” they assert.

Importantly, the authors observe, the ability to meet certain metrics affects hospital reimbursement, primarily through the Centers for Medicare and Medicaid (CMS) ‘Pay for Performance’ program. The incentive to achieve prescribed D2B goals could lead hospitals to exclude some patients inappropriately, they say. Moreover, the high variation in exclusion rates among hospitals could give rise to suspicions of ‘gaming’ the system by taking advantage of questionable exclusions. 

The ‘Weasel Clause’ 

Hitinder S. Gurm, MD, of the University of Michigan Medical Center (Ann Arbor, MI), noted that in response to an NCDR revision of the exclusion policy, a Letter to the Editor published in the Journal of the American College of Cardiology (Ellis SJ et al. J Am Coll Cardiol. 2010;56:1763) termed reporting exclusions the “weasel clause.”

“I think that’s what it is,” Dr. Gurm told TCTMD in a telephone interview. Exclusions essentially allow hospitals to cherry-pick STEMI patients who arrive early and frequently have a good outcome and not report more complicated cases that often have a poor outcome, he indicated.

“The whole premise of the door-to-balloon time initiative was to reduce mortality and [in recent studies] we didn’t see it,” Dr. Gurm said. “The problem is that we created a measure that was not directly related to outcome, and we never followed up on that.

“In retrospect, I think the whole idea was somewhat flawed,” Dr. Gurm continued. “When we created these exclusions, which remove about 55% of the deaths [from reported statistics], there’s really no way we could have expected to see a major change [in mortality] because you are excluding the high-risk patients from scrutiny…. There is no incentive to look at those cases and ask, ‘What could we have done [better]?’”

The study authors suggest 2 possible solutions to the problem: eliminate all exclusions or audit exclusion criteria (eg, allowing exclusion of cardiogenic shock or other high-risk patients).

Dr. Gurm came down firmly on the side of elimination. “I think we need to look at all comers,” he said. “Then there can be no gaming.” Moreover, he added, when a performance measure like D2B time is being met by more than 90% of hospitals, “it’s not a quality target anymore and we need to move on. The focus needs to be on total ischemia time, educating the community, and [managing] patients upstream.”


Cotoni DA, Roe MT, Li S, et al. Frequency of non-system delays in ST elevation myocardial infarction patients undergoing primary percutaneous coronary intervention and implications for door-to-balloon time reporting (from the Mission: Lifeline program). Am J Cardiol. 2014;Epub ahead of print.



Related Stories:

Registry Finds Wide Variation in Exclusion of STEMI Patients from D2B Time Reporting

The percentage of patients with ST-segment elevation myocardial infarction (STEMI) who experience ‘non-system delays’ to primary percutaneous coronary intervention (PCI) and thus are excluded from door-to-balloon (D2B) time reporting varies
  • The study was supported by the American Heart Association.
  • Dr. Kontos reports serving as a consultant to or on the advisory board for Astellas, Prevencio, and Well Point.
  • Dr. Gurm reports receiving funding from the Agency for Healthcare Research and Quality.