In-hospital prognosis and long-term mortality of STEMI in a reperfusion network. “Head to head” analisys: invasive reperfusion vs optimal medical therapy

Study design

This is a prospective hospital register of STEMI patients with a long-term vital status follow up. All STEMI patients aged >18 admitted in the Coronary Care Unit of a University hospital from January 2002 to December 2013 were prospectively and consecutively included. The study was designed and implemented in accordance with Guidelines for Good Clinical Practice and with the ethical principles laid down in the Declaration of Helsinki. All participants gave their written consent to participate in the study. The study was approved by our institution Ethics Committee, the CEIC-Parc de Salut Mar with reference number 2012/4806/I.

Variables of interest and STEMI management

Demographic variables and comorbidities such as history of hypertension, diabetes, hypercholesterolemia, smoking, and previous angina were prospectively collected. Clinical characteristics of the event were recorded, including ischemia times, AMI location and complications such as the development of pulmonary edema or cardiogenic shock or the presence of malignant arrhythmias. In addition, information about the management of the acute event, including medical treatments during the hospital stay, reperfusion therapy (including both thrombolysis or PPCI), type of reperfusion therapy (thrombolysis or PPCI) and invasive procedures (coronary angiography, mechanical ventilation), was also collected.

Patients’ care followed the current clinical practice guidelines for STEMI patients at the time of the study [15–17], but reperfusion therapy was applied according to the STEMI Code instruction [4] as a regional Reperfusion Network. In our study there was no standard care for patients. All treatments were performed under the physicians’ medical criteria depending on clinical patients’ situation.

STEMI Reperfusion Network was initiated in Catalonia in June 2009 with the purpose of reaching an optimal reperfusion therapy with PPCI [4]. Before the establishment of the STEMI Network (June 1st, 2009–pre-Network period), PPCI was performed in our hospital in STEMI patients only during working hours; thrombolytic therapy during on-duty time. After June 2009 (post-Network period), PPCI was the elective reperfusion therapy in STEMI patients. During working hours (8 am-20 pm), PPCI was performed in our hospital and patients first admitted in our hospital during on-duty time were transferred to another PPCI capable centre near our institution. Depending on the period of admission, patients were classified in two groups: pre-Network (January 2002 to May 2009) and post-Network (June 2009 to December 2013).

Events of interest

Events of interest were defined as in-hospital and 2-year mortality. In order to identify long-term fatal cases, we accessed the National Death Registry. This is an exhaustive and mandatory official database which collects individual data of all the deceased in Spain from 1987 up to now. This database, promoted by the Spanish Health Ministry, provides public institutions (healthcare administrations, research centers) with information regarding vital status and date of death, although it does not indicate the specific cause of death. We linked our data with the National Death Registry. We assumed that study participants who did not appear in this registry were alive at the end of the follow-up.

Statistical analysis

In the comparison of study groups (pre and post-Code), analysis of variance or Kruskall-Wallis test were used for continuous variables and the Chi-square test for categorical variables. Unconditional logistical regression and Cox regression were used to determine the association between comorbidities, reperfusion therapy and in-hospital or long-term mortality, with adjustment to the identified confusing variables. We tested the interaction between the use of reperfusion therapy, medical therapy and period of admission in order to evaluate in-hospital prognosis as well as after 2 years. P values lower than 0.05 were considered statistically significant.

In order to evaluate a mortality time trend along the period, day of admission was also incorporated in the Cox regression model as a spline term to accommodate a possible non-linear effect.

STEMI management

Reperfusion therapy increased in the post-Network period (89.2% vs. 64.4%). Among those treated with reperfusion there was an important increase in the use of PPCI (99% in the post-Network vs. 43.9% in the pre-Network period) with a subsequent decrease in the use of thrombolytics (1% vs. 56.1%). The changes in the reperfusion therapy strategy were associated with a slight increase in the ischemia time: median time from pain onset to reperfusion performance was 165 min (105–325 min) vs. 186 min (130–284 min) in pre- and post-Network periods respectively, p < 0.001. There were no changes in time from pain onset to monitoring: pre-Code 130 (60–258 min) vs. post-Network 90 (45–201 min), p = 0.254.

In-hospital prognosis

However, a significant decrease trend in in-hospital mortality was observed in the period analyzed (Fig. 1). Therefore, in order to analyze the effect of the reperfusion network on in-hospital mortality and the potential variables that could be involved in this effect, we included this mortality decreasing trend in the multivariate models. We observed a trend towards a decrease in in-hospital mortality in the post-Network period compared to the pre-Network period even when the model was adjusted by age, sex (Model 1), comorbidities (Model 2), severity (Model 3) and reperfusion (Model 4) with ORs ranged between 0.45 and 0.59 (Table 4). Noteworthy, when the model was further adjusted by optimal medical therapy (aspirin, statins, beta-blockers and angiotensin converting enzyme inhibitors), the benefit of the post-Network period disappeared totally (Table 4).

	Pre-network N = 670	Post-network N = 598	P value
		OR (CI 95%)
Model 1	1	0.45 (0.15; 1.37)	0.160
Model 2	1	0.48 (0.16; 1.49)	0.207
Model 3	1	0.59 (0.18; 1.95)	0.385
Model 4a	1	0.57 (0.16; 1.97)	0.375
Model 4b	1	0.50 (0.16; 1.59)	0.239
Model 5	1	1.19 (0.30; 4.76)	0.805
Model 6	1	1.14 (0.32; 4.08)	0.840

Model 1: Adjusted by age and sex

Model 2: Model 1 plus hypertension, diabetes and smoke

Model 3: Model 2 plus Killip grade III-IV at admission

Model 4a: Model 3 plus reperfusion (including both thrombolysis or PPCI)

Model 4b: Model 2 plus reperfusion (including both thrombolysis or PPCI)

Model 5: Model 4a plus medical therapy (Aspirin, beta-blocker, ACE-inhibitors and statins)

Model 6: Model 2 plus medical therapy (Aspirin, beta-blocker, ACE-inhibitors and statins)

Long-term mortality

There was no difference in 2-year mortality among acute phase survivors between the two analyzed periods (10% pre-Network vs. 8.5% post-Network, p = 0.467). Kaplan Maier curves with cumulative 2-year mortality rates are shown in Fig. 2.

Reperfusion therapy and ischemia times

In our series, the STEMI Network had an important impact on the reperfusion therapy rate and strategy. Reperfusion therapy was performed in nearly 90% of all STEMI patients, more than other national registries [18] and similar to the best European countries in STEMI acute phase reperfusion [19] like Czech Republic, a small country with a huge and experienced AMI network. Moreover, we also observed a change in the reperfusion strategy that was almost exclusively based on PPCI in the post-Network period, similar to what occurred in the Czech Republic registry [19].

Reperfusion therapy strategy is one of the most important factors to improve AMI prognosis, but time to reperfusion therapy is a main pillar too [20]. In our study we report an important increase in reperfusion rate and in the use of PPCI instead of thrombolysis, with a slight increase in time from pain onset to reperfusion performance (21 min). Our global ischemia time is similar to that reported in other European countries [21] or better than that reported in other recent studies [19]. Although some registries have confirmed the benefit of minimizing total ischemia time to improve in-hospital and long term prognosis [20], other studies have recently shown that short variations in ischemia time were not enough to modify in-hospital STEMI prognosis [22].

In-hospital mortality

The establishment of the STEMI Network was associated with an in-hospital mortality decrease among the lowest compared to most developed European countries [23]. When we tried to identify the variables which caused that decrease, reperfusion therapy was not the main factor. The use of evidence-based medical therapy could be another potential explanation as has been reported in other recent STEMI registers [18]. The use of aspirin, beta-blockers, ACE inhibitors and statins in our series increased and could be considered at least as optimal as those prescribed in the Courage Trial [12], a study that proved a similar benefit of this optimal medical therapy compared to angioplasty in stable patients with angina. Both the introduction of this medical therapy in the multivariate model (model 5 and 6; Table 4) and the decrease of the benefit in the post-Code period imply that optical medical therapy could be one of the main variables related to the lower mortality of STEMI patients in the post-Code period. This fact suggests that optimal medical therapy could be as important as both the observed reperfusion therapy rate increase and the PPCI reperfusion strategy. These findings need careful validation in bigger observational studies.

Long-term mortality

There were no differences in 2-year mortality rate among survivors to the STEMI acute phase between both periods, although our 2-year mortality is similar to the 1-year mortality rate reported in other recent studies [24]. Unfortunately, we had only information related to the vital status but not to the cause of death in fatal cases. No information about other prognosis variables, such as re-infarction or the need of revascularization was available. However, the main cause of death in the first year after an AMI is mainly related to cardiovascular events, as we can see in recent registries or important randomized studies [24, 25].

On the other hand, the benefits of PPCI vs thrombolytic therapy out of the acute phase could be more related to a lower re-infarction rate or to the need of revascularization than to prevent cardiovascular death. The reduction in long-term mortality could be due to optimal control of cardiovascular risk factors and the use of evidence-based medical therapy [26], especially in high risk patients [27].