Primary percutaneous coronary intervention in acute ST elevation myocardial infarction: Determinants of outcome

INTRODUCTION — Patients with symptoms suggestive of an acute myocardial infarction (MI) and those who have electrocardiographic (ECG) evidence of an acute MI manifested by ST elevations (>1 mm in two contiguous leads or true posterior infarct after nitroglycerin to rule out coronary vasospasm) that is considered to represent ischemia should undergo reperfusion therapy with either primary percutaneous coronary intervention (PPCI) or fibrinolytic therapy, unless contraindicated. Patients with typical symptoms in the presence of a new or presumably new left bundle branch block are also considered eligible. (See "Electrocardiogram in the diagnosis of myocardial ischemia and infarction" and "Electrocardiographic diagnosis of myocardial infarction in the presence of bundle branch block or a paced rhythm".)

Coronary reperfusion with PPCI or fibrinolytic therapy improves outcomes in patients with acute ST-elevation MI (STEMI), an MI with a new or presumably new left bundle branch block, or a true posterior MI. If performed in a timely fashion, PPCI is the reperfusion therapy of choice, compared with fibrinolysis, because it achieves a higher rate of TIMI 3 flow (more than 90 percent) (table 1), does not carry the risk of intracranial hemorrhage, and is associated with improved outcomes.

The time to onset of reperfusion therapy is a critical determinant of outcome with both PPCI and fibrinolysis [1].

This topic will discuss the impact of the following factors on outcomes in patients who undergo PPCI:

●The time between symptom onset and PCI, also referred to as treatment delay

●The time from hospital arrival to PCI (eg, door-to-balloon time or door-to-balloon delay)

●The time from first contact with the health care system

●Early versus late presentation

●The risk category of the patient

●Physician and hospital factors

●Importance of TIMI 3 flow (patent artery)

Periprocedural management issues and the outcomes after PPCI are discussed separately.

The clinical trials demonstrating benefit of PPCI compared with fibrinolytic therapy and the selection of a reperfusion strategy and the possible role of PCI after fibrinolysis are discussed separately. (See "Acute ST-elevation myocardial infarction: Selecting a reperfusion strategy", section on 'Primary PCI preferred to fibrinolysis' and "Percutaneous coronary intervention after fibrinolysis for acute ST-elevation myocardial infarction".)

DEFINITIONS — The total time between the onset of symptoms and primary percutaneous coronary intervention (PPCI) or fibrinolytic therapy, termed the treatment delay, has the following components [2]:

●Patient delay – This is the time between onset of symptoms and the call to the emergency medical system (EMS). This metric may be challenging to quantify, as patients may not be able to recall the exact time of symptom onset or symptoms may be intermittent.

●Prehospital system delay – For patients taken to a percutaneous coronary intervention center (PCI center), this is the time between the EMS call and the arrival at the PCI center and is equivalent to the time of EMS evaluation, prehospital electrocardiogram obtainment (if any), and transportation to the hospital. For patients transferred from a local hospital to a PCI center, this is sum of the time between the EMS call and arrival at the local hospital, the time at the local hospital, and the time between hospitals (second transfer time).

●Door-to-balloon delay – This is the time between arrival at the PCI center and PPCI. Time of PPCI refers to the time of coronary intervention with a balloon, stent, or other intervention, not to be confused with the time of arrival in the catheterization laboratory or the time of vascular access.

●Door-to-needle delay – This is the time between arrival at the hospital and administration of thrombolytic therapy.

●System delay – This is the sum of the prehospital system and door-to-balloon delays.

In this schema, patient delay is not easily corrected by the healthcare system, while strategies to shorten other delays are potentially modifiable. Such an effort will be useful if longer time intervals are associated with worse outcomes.

TIME FROM SYMPTOM ONSET — With regard to fibrinolytic therapy, the benefit (myocardial salvage and functional improvement) is greatest when fibrinolytics are given within the first two to three hours after the onset of symptoms, particularly within the first 70 minutes (figure 1) [3,4]. However, many patients with STEMI present late after the development of symptoms. In different registries of patients, the time from symptom onset to hospital presentation was ≥4 hours in 50 percent (a value that changed little from 1987 to 2000) [5], more than six hours in 40 percent [6], and more than 12 hours in 9 to 31 percent [7,8]. Data from an observational study found a lower percentage of late presenters (32 percent) in patients undergoing primary percutaneous coronary intervention (PPCI) [1]. Delay is greatest in females, older adults, and in those with symptoms that occur between 6 PM and 6 AM [5,6,9].

Initial studies primarily evaluated outcomes with percutaneous transluminal coronary angioplasty (PTCA) without stenting, which is no longer the standard procedure. The data were conflicting as to whether there is [10-13] or is not [14,15] a significant relationship between mortality and the time from symptom onset to reperfusion similar to that seen with fibrinolytic therapy with PTCA.

Although outcomes are improved with PTCA plus stenting compared to PTCA alone, published data are limited on the prognostic importance of the time from symptom onset to reperfusion with stenting and the evidence is inconclusive:

●Two small cardiovascular magnetic resonance studies, which evaluated the relationship between the time from symptom onset to PPCI with stenting and MI size, came to differing conclusions [16,17].

●The best available data come from a study of 6209 patients in a Denmark registry who underwent PCI with stenting between 2002 and 2008 [2]. In the multivariable analysis, there was no significant relationship between onset of symptoms and PPCI (treatment delay) and long-term mortality. However, components of the treatment delay were significantly related to outcome. The impact of the components of treatment delay on mortality is discussed below. (See 'Time from contact with the health care system (system delay)' below.)

Possible explanations for the absence of clear evidence to support a relationship between better outcomes and shorter time between symptom onset and PCI include the fact that patients are poor informants about the onset of symptoms, the period of intermittent thrombotic obstruction of the artery associated with stuttering of symptoms before total occlusion (lack of clarity as to what percent of the time is unstable angina), the improved effectiveness of PPCI versus fibrinolytics in late presenters (see below), the use of more aggressive antithrombotic therapies in the era of stenting, and patients with later presentation being less ill (survivor cohort effect) [2].

The interaction between time from symptom onset to PPCI and the interval between hospital arrival and PPCI (door-to-balloon time) is discussed below. (See 'Early versus late presentation' below.)

Late PCI to open an occluded artery — In contrast to fibrinolysis, revascularization after 12 hours with PCI could be beneficial in the 9 to 31 percent of patients with STEMI who present more than 12 hours after the onset of symptoms [7,8]. (See "Acute ST-elevation myocardial infarction: The use of fibrinolytic therapy", section on 'Timing'.)

It has been estimated that, in the absence of reperfusion therapy, the infarct-related artery is still occluded at 12 to 24 hours in 65 percent of these patients [18]. The establishment of TIMI 3 flow, as discussed below, may be of benefit in these patients. (See 'Prognosis after primary PCI' below.) The randomized trials that have evaluated routine late PCI have included patients at different time periods after symptom onset, ranging from more than 12 hours to up to 28 days. Some have demonstrated an improvement in left ventricular function with intervention, but none have demonstrated a significant benefit on hard clinical outcomes. The role of late PCI to open an occluded artery is discussed in detail elsewhere. (See "Acute ST-elevation myocardial infarction: Selecting a reperfusion strategy", section on 'Late presentation'.)

TIME FROM HOSPITAL ARRIVAL (DOOR-TO-BALLOON TIME) — In contrast to the uncertain relationship between mortality and the time from symptom onset to balloon inflation, the time from hospital arrival to balloon inflation (the door-to-balloon time [D2B time]) has been relatively well studied and is predictive of in-hospital mortality [1,10,14,19-23]. Evidence for a detrimental effect of longer D2B times on outcomes comes from both observational studies and randomized trials:

●In a report of 29,222 ST elevation myocardial infarction (STEMI) patients in the NRMI-3 and 4 registries (1999 to 2002) who were treated with percutaneous coronary intervention (PCI) within six hours of presentation, longer D2B times were significantly associated with increased in-hospital mortality (3.0, 4.2, 5.7, and 7.4 percent for D2B time of ≤90 minutes, 91 to 120 minutes, 121 to 150 minutes, and >150 minutes, respectively) [21]. Patients with D2B times >90 minutes had a significant increase in mortality compared to those with D2B times ≤90 minutes (odds ratio 1.42, after adjusting for patient characteristics).

●In a systematic evaluation of 1440 STEMI patients who received primary PCI (PPCI) during 2006 to 2007 in Quebec, Canada, D2B times exceeded 90 minutes in 68 percent; death or readmission for MI or heart failure at one year occurred in 13.6 percent [1]. Mortality at 30 days and one year occurred significantly less frequently in those with D2B times ≤90 minutes compared to those with longer D2B times (3.4 versus 6.1 percent; odds ratio [OR] 1.87, 95% CI 1.02-3.41 and 5.5 versus 9.1 percent; OR 1.71, 95% CI 1.06-2.76, respectively).

●In an analysis of 4548 patients enrolled in the CADILLAC and HORIZONS-AMI trials, short D2B times (≤90 minutes) were associated with a significantly lower mortality rate at one year compared to longer times (3.1 versus 4.3 percent; hazard ratio 0.72, 95% CI 0.52-0.99) [23]. (See "Acute ST-elevation myocardial infarction: Management of anticoagulation", section on 'Primary percutaneous coronary intervention'.)

Despite the evidence supporting the concept that longer D2B times lead to worse outcomes, improvement in D2B times has not lead to improvement in survival rates. In a study of nearly 100,000 STEMI patients in the United States CathPCI Registry of the National Cardiovascular Data Registry who underwent primary PCI between July 2005 and June 2009, median D2B times fell from 83 to 67 minutes, comparing the first to the last 12 months of the time period [24]. Despite this significant improvement, there was no change in risk-adjusted in-hospital mortality (5.0 versus 4.7 percent, respectively; p = 0.34).

While there is a direct relationship between increasing mortality and longer D2B time, and while we support efforts to support D2B time, it may be difficult to demonstrate further improvements in survival with shorter times, particularly if they are on the order of magnitude of minutes. This is in part related to the impact of total ischemic time on mortality. In addition, the results of studies with longer follow-up may be illuminating.

Outcomes in patient subgroups — The issue of whether the improvement in survival with short D2B times is delivered equally to all patients has been studied, and two subgroups of patients appear to benefit the most: those who present early after symptom onset and those at high risk.

Early versus late presentation — With regard to early versus late presentation, it appears that short D2B time is associated with improved outcome in early presenters more than in late presenters:

●In a study of 2322 patients who underwent PPCI comparing short versus long D2B time (<2 versus >2 hours) with either balloon angioplasty or stenting, there was a significant decrease in seven-year mortality in patients presenting ≤3 hours from symptom onset (15.0 versus 24.7 percent), but not in patients presenting after three hours (18.5 versus 21.1 percent) [25].

●In an observational study of 4548 patients who underwent PPCI with stenting in the CADILLAC and HORIZONS-AMI trials, short (≤90 minutes) compared to long D2B times were associated with a significantly lower one-year mortality rate in patients with early presentation (≤90 minutes versus ≥90 minutes), but not those with later presentation (1.9 versus 3.8 percent, hazard ratio 0.86, 95% CI 0.26-0.93 and 4.0 versus 4.6 percent, hazard ratio 0.86, 95% CI 0.58-1.28, respectively) [23]. (See "Acute ST-elevation myocardial infarction: Management of anticoagulation", section on 'Classification of anticoagulant agents'.)

Patient risk category — With regard to risk category, it appears that high-risk patients benefit more from short D2B time:

●In the study of 2322 patients discussed above, there was a significant decrease in seven-year mortality with shorter D2B times (<2 versus ≥2 hours) in high-risk (21.5 versus 32.5 percent) but not low-risk patients (9.2 versus 10.8 percent). High risk was defined as one or more of Killip class 3 or 4 heart failure (table 2), age >70 years, or anterior MI.

●In the analysis of 4548 patients discussed above, short compared to long D2B times showed a trend toward lower one-year mortality in both high- and low-risk groups (5.7 versus 7.4 and 1.1 versus 1.6 percent, respectively) [23]. In patients presenting early (≤90 min), the hazard ratios for mortality rate in patients with short versus long D2B times were identical for high- versus low-risk patients, but the absolute mortality rate differences were greater in high-risk patients (3.3 versus 0.7 percent). In patients presenting late (>90 minutes), mortality was similar with short and long D2B times in both high- and low-risk patients.

The findings above suggest that short D2B times are critically important in patients who present early, especially in high-risk patients. The impact of PCI-related delay on decision making in the individual patient is discussed elsewhere. (See "Acute ST-elevation myocardial infarction: Selecting a reperfusion strategy", section on 'Fibrinolysis'.)

TIME FROM CONTACT WITH THE HEALTH CARE SYSTEM (SYSTEM DELAY) — The door-to-balloon delay represents only a portion of the total time that the STEMI patient is managed by the healthcare system. This longer time interval, referred to as system delay, is the sum of the prehospital system and door-to-balloon delays. For patients first brought to a hospital without percutaneous coronary intervention (PCI) capability, the system delay has multiple components. (See 'Definitions' above.)

The associations between healthcare delays and mortality were evaluated in a study of 6209 patients with STEMI from 2002 to 2006 who were included in large, public medical databases in Western Denmark [2]. Approximately one-third of these patients were field-triaged to a PCI center and two-thirds were admitted to a local hospital and then transferred to a PCI center. In this cohort, the median follow-up time was 3.4 years and the cumulative one-year mortality was 9.3 percent. The following findings were noted:

●In the univariable analysis of the components of delay (system, prehospital system, door-to-balloon, treatment, and patient), system delay had the strongest association with mortality (hazard ratios [HR] 1.22, 1.19, 1.13, 1.054, and 1.042, respectively). All hazard ratios were significant.

●For system delays of 0 to 60 minutes, 61 to 120 minutes, 121 to 180 minutes, and 181 to 360 minutes, the long-term cumulative mortality was 15.4, 23.3, 28.1, and 30.8 percent, respectively.

●In the multivariable analysis, treatment delay and patient delay were not associated with mortality, but system delay was independently associated with mortality (HR 1.10, 95% CI 1.01-1.16 per one-hour delay). The main components of system delay, prehospital system delay, and door-to-balloon delay were similarly associated with mortality (HR 1.10, 95% CI 1.02-1.18 and 1.14, 95% CI 1.05-1.24, respectively).

While this study demonstrates the negative impact of increasing system delay on mortality, it does not allow any firm conclusion regarding the time at which an alternative strategy using a fibrinolytic agent would be preferable. (See 'Transfer from a non-PCI center' below.)

We believe that the earlier a patient presents (and thus a higher baseline risk), the less door-to-balloon delay that is acceptable. Thus, fibrinolytic therapy is an important option for patients who present within the first three hours and who are at a low risk of bleeding.

Direct transfer from the field — STEMI patients diagnosed with a prehospital ECG should be transferred directly to the catheterization laboratory of a PCI-capable hospital (bypassing the emergency department) if the following two criteria are met: The patient is hemodynamically stable, and the patient is received by personnel (including one physician) qualified to care for critically ill patients. Transfer to the catheterization laboratory should occur in less than 120 minutes, if possible, in stable patients.

As pointed out by the authors of the Western Denmark study discussed above, one implication of worse outcomes with longer door-to-balloon times is that mortality may be improved with direct transfer to a PCI hospital compared with evaluation and treatment at a closer hospital without PCI capability. The feasibility and impact of this strategy on door-to-balloon delay was evaluated in a nonrandomized comparison of 344 patients with chest pain of less than 12 hours duration and ST-segment elevation characteristic of acute MI in 2005 and 2006 [26]. In this study, 135 patients were referred directly from the field by paramedics trained in ECG interpretation and 209 patients were referred directly from regional emergency departments. The median door-to-balloon time, defined as the time between arrival at the first hospital to first balloon inflation, was significantly shorter in patients referred from the field (69 versus 123 minutes) and the percent of patients with door-to-balloon times of less than 90 minutes was significantly higher (80 versus 12).

For patients with STEMI diagnosed with a prehospital ECG who are transported to a PCI-capable hospital, one way to shorten the time to reperfusion is to bypass the emergency department (ED) and take the patient directly to the catheterization laboratory. The efficacy of this approach was addressed in a study of 12,158 STEMI patients, 10.5 percent of whom bypassed the ED [27]. The time from first medical contact to device activation was shorter in these individuals compared to those who went to the ED (68 versus 88 minutes; p<0.0001). There was a trend toward lower mortality (2.7 versus 4.1 percent; adjusted odds ratio 0.69, 95% CI 0.45-1.03).

Transfer from a non-PCI center — The worse outcomes seen with longer door–to-balloon times are of particular concern for patients with acute MI who are initially evaluated at hospitals without on-site PCI capability. The time spent at the first hospital, the subsequent transfer time, and the time spent at the receiving hospital prior to arrival in the catheterization laboratory are sources of longer door-to-balloon times. (See 'Time from hospital arrival (door-to-balloon time)' above.) The frequency, magnitude, and clinical impact of delays occurring at each of these time intervals were examined in a prospective study of 2034 patients referred to a PCI center (Minneapolis, Minnesota, United States) for primary PCI [28]. The following findings were noted:

●Delays occurred most frequently in the ED at the referral hospital, with less delay seen at the PCI center or in transport (64, 16, and 13 percent, respectively).

●For the referral hospital, the most frequent reasons for delay were awaiting transport and ED delays (26 and 14 percent, respectively).

Diagnostic dilemmas, including nondiagnostic initial ECGs, led to the greatest delay, but these had limited or no impact on mortality. The door-in to door-out (DIDO) time is defined as the duration of time from arrival to discharge at the first (STEMI referral) hospital. The DIDO time is probably modifiable, as opposed to the non-modifiable delay caused by an additional ambulance ride. As a result, the DIDO time has become a performance measure of the United States Centers for Medicare and Medicaid Services [29]. The relationship between DIDO time and outcomes was evaluated in a study of nearly 15,000 patients who initially presented to a STEMI referral hospital and were subsequently transferred to a PCI facility; the patients were enrolled in the Acute Coronary Treatment and Intervention Outcomes Network (ACTION) Registry-Get With the Guidelines (GWTG) between 2007 and 2010 [30]. Exclusion criteria were fibrinolytic therapy, the need for rescue or facilitated PCI, or a planned pharmacoinvasive approach. (See "Diagnosis and management of failed fibrinolysis or threatened reocclusion in acute ST-elevation myocardial infarction" and "Acute ST-elevation myocardial infarction: Selecting a reperfusion strategy" and "Acute ST-elevation myocardial infarction: Selecting a reperfusion strategy", section on 'Fibrinolysis'.)

The following results were reported:

●The median DIDO time was 68 minutes, with only 11 percent of patients having a goal DIDO time of less than 30 minutes.

●Predictors of longer DIDO times included older age, female sex, off-hours presentation (any time other than 8 AM to 5 PM Monday through Friday), and non-emergency medical services presentation to the first hospital.

●Patients with DIDO times less than 30 minutes had significantly shorter door-to-balloon times (85 versus 127 minutes) and a lower in-hospital mortality rate (2.7 versus 5.9 percent).

Recommendations for reperfusion in patients in whom a significant delay in transfer is anticipated are discussed elsewhere. (See "Acute ST-elevation myocardial infarction: Selecting a reperfusion strategy", section on 'Fibrinolysis'.)

False positive cath lab activation — Strategies that may decrease door-to-balloon time in patients transferred for primary PCI allow cardiac catheterization laboratory activation by the practitioners at the referring hospital or in the field. A weakness of this strategy could be referral of patients who are not candidates for PCI by practitioners who are potentially less skilled in ECG interpretation.

The frequency of false-positive referrals and the associated diagnoses were evaluated in a prospective registry of 1335 patients referred from 30 hospitals in Minnesota and Wisconsin between 2003 and 2006 [31]. False positive was defined as no culprit coronary artery (present in 14 percent), negative biomarkers (11.2 percent), or both (9.2 percent). In those patients with negative biomarkers, the most common discharge diagnoses were early repolarization, nondiagnostic ECG, previous MI, left bundle branch block, and pericarditis.

RECOMMENDATIONS OF OTHERS — We generally agree with recommendations made by major guideline societies regarding the management of STEMI patients who undergo primary PCI [32-34].

HOSPITAL PERFORMANCE — The time from hospital arrival to primary percutaneous coronary intervention (PPCI) in the United States is decreasing slowly. In large surveys, the median door-to-balloon (D2B) time fell from 120 minutes from 1994 to 1995 to 108 minutes from 1999 to 2002 [35,36]; however, between 1999 and 2002, only 35 percent of patients met the recommended goal of less than 90 minutes [36]. The median D2B time was higher (180 minutes) among patients transferred for PPCI, as only 4.2 percent met the recommended goal [37].

The (United States) National Registry of Myocardial Infarction evaluated outcomes in over 900,000 ST elevation myocardial infarction (STEMI) patients who were eligible for reperfusion upon arrival to the hospital between 1990 and 2006 [38]. D2B time among nontransferred patients declined significantly from 111 minutes in 1994 to 79 minutes in 2006.

Studies of successful hospitals have identified a number of organizational strategies that contributed to minimizing treatment delay [39-42]. In a multivariate analysis of 365 hospitals, six strategies that reduced D2B time were identified [41]:

●Emergency medicine physician ability to activate the catheterization laboratory (mean reduction in D2B time, 8.2 minutes).

●Single call to a central page operator who activates the catheterization laboratory (13.8 minutes).

●Expectation that staff will arrive in the catheterization laboratory within 20 minutes after being paged compared to more than 30 minutes (19.3 minutes).

●Having an attending cardiologist always on site (14.6 minutes).

●Emergency department activation of the catheterization laboratory while the patient is en route (15.4 minutes).

●Regular audits of performance times.

●Standard and simple algorithms.

More robust improvement in D2B time was demonstrated in a subsequent prospective, single-center study of a strategy designed to improve D2B time [42]. The hospital intervention included mandated emergency department physician activation of the catheterization laboratory and immediate transfer of the patient to an immediately available catheterization laboratory by an in-house nursing transfer team. D2B times were compared between 60 STEMI patients cared for between October 2004 and August 2005 (preintervention) and 86 STEMI patients cared for between September 2005 and June 2006 (postintervention).

The following statistically-significant improvements were seen in the postintervention period:

●Median D2B time decreased overall (114 versus 76 minutes)

●D2B time during regular hours (84 versus 65 minutes)

●D2B time during off hours (124 versus 78 minutes)

●Treatment within 90 minutes increased from 28 to 71 percent

●Mean infarct size decreased (peak creatinine kinase 2623 versus 1517 IU/L)

Significant improvement in D2B time as a result of the ability to transmit the prehospital electrocardiogram has been demonstrated in two studies [43,44]. In a NRMI-4 report, the catheterization laboratory was activated as a consequence of the prehospital electrocardiogram in about 6 percent of over 56,000 patients with STEMI seen between 2000 and 2002 [45]. This interaction was associated with significant reductions in mean D2B time with PPCI (94 versus 110 minutes) and mean door-to-drug time with fibrinolytic therapy (24 versus 35 minutes).

The strong evidence supporting better outcomes of PCI with multiple performance improvement strategies has led to the inclusion of shorter D2B time as a quality indicator in practice guidelines and the development of pay for performance reimbursement by insurers to hospitals [44].

Nonsystem factors leading to delay — The factors discussed above have been referred to as system-related causes for delay in D2B time. In a study of nearly 83,000 patients with STEMI in the United States PCI Registry (2009 to 2011), nonsystem delays occurred in 14.7 percent [46]. Reasons included delays in providing consent, difficult vascular access, and difficulty in crossing the lesion. As expected, and consistent with an increase in D2B time, the in-hospital mortality was greater in those with compared to those without nonsystem delay (15.1 versus 2.5 percent; p<0.01) even after adjustment for baseline characteristics.

IMPORTANCE OF LOCAL EXPERTISE — An important question is whether the results from controlled clinical trials at major centers can be applied to community practice. The ability to perform emergent percutaneous coronary intervention (PCI) on a 24-hour basis requires a complex organization and the firm commitment of the catheterization laboratory staff. These logistic issues make primary PCI (PPCI) less available, especially at smaller, low-volume hospitals (table 3 and table 4). It is therefore not surprising that results following PPCI vary among different centers and different physicians.

Hospital and operator volume — Some of the variation in results among institutions can be attributed to differences in the annual volume of PPCI at an individual hospital [14,47-50]. This can be illustrated in a (United States) National Registry of Myocardial Infarction (NRMI) review of PPCI and fibrinolysis in over 62,000 patients from 446 acute-care hospitals in the United States [47]. PPCI was associated with reduced mortality compared to fibrinolysis in high-volume (≥49 procedures, 3.4 versus 5.4 percent) and intermediate-volume hospitals (17 to 48 procedures, 4.5 versus 5.9 percent); the degree of benefit was similar to that seen in the large randomized trials. In contrast, there was no mortality benefit with PPCI in low-volume hospitals (≤16 procedures, 6.2 versus 5.9 percent), but the total stroke rate was significantly reduced (0.4 versus 1.1 percent).

Two additional findings were noted in a second NRMI analysis [48]:

●The mortality rate after PPCI was significantly lower in high- compared to lower-volume hospitals (relative risk 0.72). The absolute difference of approximately 2 per 100 patients treated was independent of the number of MIs seen at the hospitals.

●There was no significant relationship between the volume of fibrinolytic interventions and in-hospital mortality (7.0 versus 6.9 at the highest- versus lowest-volume hospitals).

The 2011 American College of Cardiology Foundation/American Heart Association/Society for Cardiovascular Angiography and Interventions guideline for PCI recommended that PPCI be performed by experienced operators (>75 elective PCI procedures per year, of which at least 11 are ideally for ST elevation MI [STEMI]) at high-volume centers (>400 PCI procedures per year, of which at least 36 are ideally for STEMI) [51,52]. The benefit of PPCI was considered less well-established for less-experienced operators in patients who are eligible for fibrinolysis.

PCI without on-site cardiac surgery — In the early days of PCI, it was thought that on-site cardiac surgery might be required to best treat complications, particularly abrupt closure. This concern has diminished with the use of stents, which has largely eliminated the problem of acute vessel closure in nearly all cases. The following three studies support the performance of PCI in hospitals without on-site backup:

●The outcome with these two strategies in hospitals without on-site cardiac surgery was addressed by the C-PORT trial in which 451 fibrinolysis-eligible patients with an acute MI seen within 12 hours of symptoms were randomly assigned to PCI or fibrinolysis [53]. Patients undergoing PPCI had a significantly lower incidence of the composite end point (death, recurrent MI, and stroke) at six months (12.4 versus 20 percent for fibrinolysis); the difference in outcome was primarily due to a lower rate of recurrent MI (5.3 versus 10.6 percent).

●The issue of whether primary PCI can be safely performed at hospitals without on-site cardiac surgery was addressed in a 2011 meta-analysis of 11 studies that included over 124,000 STEMI patients treated with PPCI that compared outcomes between hospitals without and with on-site cardiac surgery [54]. Comparing hospitals without to those with on-site surgery, there was no increase in the observed risk of in-hospital mortality (4.3 versus 7.2 percent, respectively; odds ratio [OR] 0.96, 95% CI 0.88-1.05 or emergency bypass 0.22 versus 1.03 percent; OR 0.53, 95% CI 0.35-0.79).

●In a multivariable analysis (published after the meta-analysis) of 384,013 patients who underwent PCI between 2006 and 2012, there was no difference in survival at 30 days, one year, or five years between those treated at facilities with off-site compared to on-site surgical back up (hazard ratios 0.87, 95% CI 0.71-1.06; 0.92, 95% CI 0.79-1.07; and 0.92, 95% CI 0.84-1.01, respectively) [55].

The 2011 American College of Cardiology Foundation (ACCF)/American Heart Association (AHA)/Society for Cardiovascular Angiography and Interventions (SCAI) guideline for PCI made a weak recommendation for PPCI at institutions without on-site cardiac surgery that met certain institutional and operator criteria (table 5) [51,52]. In 2014, the SCAI/ACC/AHA published an expert consensus document on PCI without on-site surgical backup. That document, while not making new broad recommendations, indicates that recommendations made in the 2011 guideline are reasonable [56]. In addition, it provides recommendations for facility and personnel requirements as well as requirements for off-site surgical backup and case selection in this setting (table 6A-D).

PROGNOSIS AFTER PRIMARY PCI — This section will present information regarding prognosis after primary percutaneous coronary intervention (PPCI) for STEMI. In a registry study of 2804 such patients, 30-day, one-year, and five-year all-cause (and cardiac) mortality rates were 7.9 (7.3), 11.4 (8.4), and 23.3 (13.8) percent, respectively [57]. In this study, the main causes of cardiac death within the first 30 days were cardiogenic shock and anoxic brain injury. After 30 days, causes of death were predominantly noncardiac, with malignancies and pulmonary diseases dominating.

Other risk factors are discussed below.

TIMI flow grade — Patients with normal blood flow in the infarcted artery at the end of the procedure have a better prognosis than those who do not. This issue is discussed separately. (See "Suboptimal reperfusion after primary percutaneous coronary intervention in acute ST-elevation myocardial infarction", section on 'Prognosis'.)

Electrocardiographic markers — Electrocardiographic markers can successfully predict outcome in the broad range of patients with ST elevation myocardial infarction. In the first prospective evaluation of electrocardiographic markers in patients undergoing PPCI, the prognostic utility of six different methods for evaluating the extent of ST-segment elevation resolution after PCI was assessed in 4866 patients enrolled in the APEX-AMI trial [58]. All six methods were successful in predicting outcomes of death or the composite of death, cardiogenic shock, or heart failure at 90 days. One of these, the measurement of the residual, absolute ST elevation in the single, most affected (worst) lead 30 minutes after PCI, performed as well as the more complex methods. Data from additional therapeutic trials are needed before this marker can be considered a reasonable surrogate for clinical outcomes. (See "Electrocardiogram in the prognosis of myocardial infarction or unstable angina", section on 'ECG for prognosis in STEMI'.)

Infarct size — Larger infarct size after PPCI is associated with worse outcomes. In a patient-level meta-analysis of 10 randomized trials of primary PCI in which infarct size was assessed with cardiac magnetic resonance imaging or technetium-99m sestamibi single-photon emission computed tomography within one month of STEMI, the following was noted [59]:

●Median infarct size (percent of left ventricular myocardial mass) was 17.9 percent.

●One-year estimates of all-cause mortality, reinfarction, and heart failure hospitalization were 2.2, 2.5, and 2.6 percent, respectively.

●There was a strong, graded response between infarct size and mortality and hospitalization for heart failure at one year.

Other risk factors — Other risk factors including Killip class (table 2), age, and the number of diseased vessels are also important. These parameters have been incorporated into three risk models that have been prospectively validated: the Zwolle risk index, the TIMI risk score, and the CADILLAC risk score (table 7) [60,61]. (See "Risk stratification after acute ST-elevation myocardial infarction".)

The general discussion of the prognosis after myocardial infarction is found elsewhere. (See "Prognosis after myocardial infarction".)

SUMMARY AND RECOMMENDATIONS

●Primary percutaneous coronary intervention (PPCI) with stenting, if performed in a timely manner, is associated with better outcomes than fibrinolysis. (See "Acute ST-elevation myocardial infarction: Selecting a reperfusion strategy".)

●The time from symptom onset to PCI has not been shown to be an important determinant of outcome. At the least, the benefit from PPCI is less dependent upon the time from symptom onset than is fibrinolysis (figure 1). (See 'Time from symptom onset' above.)

●The time from hospital arrival to PCI (door-to-balloon time) is an important determinant of benefit, with the best outcomes occurring when the time to PCI is 90 minutes or less. (See 'Time from hospital arrival (door-to-balloon time)' above.)

●Increasing system delay is associated with worse outcomes. (See 'Time from contact with the health care system (system delay)' above.)

●Patients who are transferred to a PCI center have better outcomes than those treated with fibrinolysis at the presenting hospital if PPCI is delivered according to guideline standards. Most of the benefit in these trials is due to a lower rate of reinfarction after PCI, which is unrelated to the time required for transfer. (See 'Transfer from a non-PCI center' above.)

●Destination protocols for emergency medical system that bypass non-PPCI-capable hospitals and thereby shorten system delays to PPCI have been associated with improved outcomes in ST elevation myocardial infarction. (See 'Hospital performance' above.)

●Hospitals should adapt strategies to reduce door-to-balloon times and thereby improve outcomes in STEMI patients treated with PPCI. (See 'Hospital performance' above.) All hospitals should meet American College of Cardiology Foundation/American Heart Association criteria for institutional and operator volumes in hospitals with or without on-site cardiac surgery. The door-to-balloon time should be less than two hours. (See 'Importance of local expertise' above.)

●Late PCI to open an occluded artery should be considered in patients with severe heart failure, hemodynamic or electrical instability, or persistent ischemic symptoms. (See 'Late PCI to open an occluded artery' above.)

●PPCI should not be performed in hospitals without on-site cardiac surgery unless they meet specific criteria, including having a proven plan for rapid transport to a cardiac surgery operating room in a nearby hospital and having appropriate hemodynamic support capability for transfer. (See 'PCI without on-site cardiac surgery' above.)

The discussion of the use of these observations to formulate a reperfusion strategy for the individual patient is found elsewhere. (See "Acute ST-elevation myocardial infarction: Selecting a reperfusion strategy".)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Duane Pinto, MD, MPH, who contributed to earlier versions of this topic review.