STEMI Transfer and Treatment: What Works, What Doesn't

STEMI transfer and treatment in Quebec

Shelley Wood

October 26, 2011

October 26, 2011 (Vancouver, British Columbia) — A pitched effort by the Quebec Ministry of Health to understand how STEMI care is managed — and possibly mismanaged — in that province should serve as launching pad for discussing what works and what doesn't in the management of these patients. So says Dr Laurie Lambert (Institute for National Excellence in Health and Social Services [INESS]), who presented a series of detailed analyses looking at six months' worth of Quebec data here at the Canadian Cardiovascular Congress 2011.

The analyses were undertaken by the publicly funded INESS and were based on chart reviews of all patients admitted with a diagnosis of AMI between October 2008 and March 2009.

"We had very little idea of, in Quebec, how we were treating patients," Lambert told heartwire. "What were our delay-to-treatment times? So this research was funded to really provide a portrait of STEMI care in Quebec."

Dr Laurie Lambert

There are a total of 82 hospitals that treat at least 30 AMI patients per year in the province of Quebec, including 13 PCI centers, five of which have no on-site surgery. The first of the analyses Lambert presented looked at variations in care across the 13 PCI hospitals, finding wide variations in the choice of first reperfusion device, choice of radial vs femoral artery, and time to treatment. The majority of primary-PCI procedures are performed in patients who are transferred from another hospital. A key point here, Lambert said, is there really is no standardized approach to the practice of primary PCI across Quebec hospitals.

Prehospital ECGs and alerts improve care

Another analysis looked at the impact of prehospital ECGs and prehospital "alerts." Echoing findings reported in the US and elsewhere, Lambert et al's analysis showed that patients whose hospital admission had been preceded by an ECG obtained in the ambulance were more likely to get fibrinolysis, less likely to get "no reperfusion treatment," less likely to be transferred for primary PCI, and more likely to be admitted directly at a PCI hospital. Delays from door to treatment were significantly reduced when ECGs were transmitted from the ambulance. For emergency-response teams without the capacity to transmit ECGs to a receiving hospital, a system of "alerts" in which ambulance crews informed hospitals about suspected STEMIs also was associated with reduced door-to-reperfusion delays.

A third analysis compared hospitals with surgery on site and hospitals with no on-site surgery and found that delays from door-to-treatment, as well as in-hospital mortality rates, were no different between hospitals. Where differences emerged was in regions with both types of hospital: here, both ambulances as well as patients making their own way to the hospital tended to favor the hospitals with surgery on site.

To heartwire, Lambert agreed that this issue is highly "politically charged," in large part because hospitals already performing PCIs are reluctant to lose patients to new hospitals. In the discussion following her presentation, Dr Stéphane Rinfret (Laval University, Quebec, QC) observed that the findings speak to the need for policy makers to take into account "existing corridors" when planning new PCI centers, since it's hard to reroute the system.

"These corridors of service are strong," Lambert agreed to heartwire. "That's what they do, that's what they've always done, and that's harder to change."

Decision drivers

Yet another analysis explored the decision not to perform PCI in a STEMI patient sent for the procedure. In Quebec, 13% of women and 10% of men are sent for primary PCI but don't get it, Lambert noted. Not surprising, the rate of non-PCI performance was twice as high at hospitals that had surgery on-site, and among those not sent for PCI, referral to CABG was cited as the reason in 28% of patients. Women and older patients were more likely to not end up getting PCI.

Further analyses showed that nonperformance of PCI was more likely when door-to-cath lab times exceeded 90 minutes and among patients who were transferred for PCI from a non-PCI hospital. Of note, in-hospital mortality was three times higher when PCI was not performed than when it was.

The data leave several important questions unanswered, Lambert observed. Door-to-device time is "highly competitive," she said, "but those data are considering only the people who successfully have an angioplasty, and all the processes of care prior to PCI tend to be shorter if PCI is performed. But we don't know which causes which: is it because you took longer that the patient did not have a PCI? Or is it because they are sicker that it takes longer?"

Researchers and policy makers should keep these patients who did not get PCI in mind, Lambert said, when trying to gauge delays and clinical outcomes within a system of STEMI care, "regardless of whether or not they subsequently undergo PCI."

Can information lead to action?

In an interview with heartwire, Lambert stressed that all of these analyses provide answers to one part of a larger puzzle. While no hospital names were provided in the data presented at the meeting, all 82 hospitals in Quebec have been provided with all of these results, so they can see how they compare with other institutions, although each hospital will see unblinded data only for its own institution.

"This lets them know where they rank, according to type of hospital, region, etc, and we've sent this as well to the health agencies and to the ministry [of health]. What we're hoping is that we can initiate change, but that's something very difficult to do in a complex system."

Lambert points to the pay-for-performance approach adopted in the US, where hospital performance is publically reported: a strategy some physicians and hospitals have criticized since it doesn't take into account specific circumstances. For example, with something like door-to-balloon delays, "sometimes when you take a long time with a patient, there are good reasons why," she said. But is pay for performance the future? Lambert says, "I hope not."

"But this is what we need to research to find out — does this lead to improvements? One way to prompt change is publicizing the hospitals' performance, naming them publicly, but when you do that, you may see hospitals start trying to game the system, so it's really a complex issue. And I think it would be great to have more discussion about this, and for people across Canada to discuss . . . more openly what's working, what's not."

Lambert had no conflicts of interest.