Based on the existing scientific evidence, the main clinical practice guides recommend an early invasive strategy (EIS) in patients with medium-high risk non-ST-segment elevation acute coronary syndrome (NSTE ACS).1–5 However, some registry-based studies have documented a paradoxically lesser utilization of early invasive strategies in high risk patients.6–12 The ARIAM-SEMICYUC study offers an opportunity to re-examine this problem based on the recent Spanish data.

The objectives of the present study are: (1) to describe the use of EIS in relation to the baseline risk of patients with NSTE ACS admitted to intensive care units (ICUs); and (2) to identify the independent predictors of the application of EIS in patients admitted due to NSTE ACS.

The profile of the patients, stratified according to GRACE risk level, was consistent with that of other registries8,9,11 (Table 1). The high risk group was characterized by older age, a greater proportion of women, increased comorbidity (diabetes, previous infarction, renal failure), the existence of heart failure upon admission, a high baseline bleeding risk, and increased mortality.

Regarding the drug treatment administered from symptoms onset and during admission to the ICU (Table 2), the patients at high risk were less often treated with statins and beta-blockers, with no other significant differences among the three risk levels. Likewise, risk level was directly correlated to the use of noninvasive mechanical ventilation, with a negative association to the adoption of EIS—though the differences failed to reach statistical significance.

The variables associated to the implementation of EIS are summarized in Table 3. Basically, the group of patients with NSTE ACS subjected to early invasive treatment were younger, with a lesser prevalence of diabetes and of heart failure upon admission, a lesser bleeding risk, and were admitted to a hospital with the availability of a hemodynamics laboratory.

The logistic regression analysis (Table 4) identified the availability of a hemodynamics laboratory, the presence of ST-segment changes, the absence of diabetes, the absence of tachycardia upon admission, and the existence of low bleeding risk as independent predictors of the use of an invasive treatment strategy. Previous coronariography was not independently correlated to the adoption of EIS (adjusted OR 0.77, 95%CI 0.50–1.21), and did not act as an effect modifier of the association between the GRACE level and the implementation of EIS (interaction p-value=0.8425). On replacing heart rate with the Killip class, the presence of class >1 was found to be significantly associated to a lesser utilization of invasive strategies (OR 0.64, 95%CI 0.42–0.96).

The results of the present study confirm the inverse relationship between the risk of events and the use of an EIS in NSTE ACS, detected in previous studies.6,8–11,17,18 At first sight this is paradoxical, since it would indicate that the patients who could benefit most from such treatment are precisely the individuals in which it is least used.6–12

The reasons for this paradoxical situation have not been fully clarified. Given the greater proportion of women and elderly people in the high risk stratum, the hypothesis of lesser therapeutic effort could be considered in these individuals.7,17,19,20 However, in our study there were no other indications of inequalities in therapeutic effort according to the risk stratum involved (Table 2), and neither age nor gender were found to be independently associated to the implementation of EIS.

Given the increased frequency of prior coronariography in the patients at high risk, we must consider the possibility that improved prior knowledge of the coronary anatomy could have favored the use of a conservative strategy in these patients. However, in the multivariate analysis, antecedents of coronariography did not act as an effect modifier (interaction p-value=0.8425), and were not independently associated to the implementation of EIS (adjusted p-value=0.2501).

The dose-response association between bleeding risk and the adoption of EIS is an important finding in this study, suggesting that the lesser utilization of EIS in the patients with high GRACE scores is at least partly due to the increased bleeding risk of these patients. Form this perspective, the existence of a certain dissociation between the guides (designed in reference to ideal patients, with a single disease) and actual clinical practice (individual patients with comorbidities) could be understandable. However, this argument loses strength on considering the radial access—much less susceptible to bleeding phenomena than the femoral access.21–25

The scant utilization of an early invasive strategy in diabetic patients, reported in many studies,6,7,9,26 is more difficult to explain. Clinicians may presume the existence of multiple vessel disease not amenable to revascularization in diabetic individuals, and therefore an unfavorable risk-benefit ratio in such cases. However, this explanation does not fit well with the repeatedly demonstrated benefits derived from coronary intervention in diabetics.27–29

The association between tachycardia upon admission (an indicator of heart failure) and scant interventional practices is less well known. This association possible may be attributable to chance (type I error). However, the negative association between a Killip class of >1 and the implementation of an early invasive strategy suggests that the association is real. Alternatively, difficulties in transferring unstable patients (within or between hospital centers) to the hemodynamics laboratory, or limitations of therapeutic effort, may possibly contribute to lesser utilization of invasive strategies in these patients.

In conclusion, there is a tendency to concentrate the utilization of EIS among patients with NSTE ACS exhibiting scant comorbidity. This paradoxical situation is still far from being resolved. Clinicians therefore should carefully revise the risk–benefit relationship of interventionism in their patients,24,25 particularly among diabetics, patients with heart failure, and patients at high bleeding risk.30

The ARIAM registry is a project financed by the Sociedad Española de Medicina Intensiva, Crítica y Unidades Coronarias (SEMICYUC).

The authors declare no conflicts of interest.