Clinical and demographic characteristics are detailed in Table 1.

Pre-infarct angina (PIA) was diagnosed if a patient had arm, jaw, or chest pain in the preceding eight days before the diagnosis of STEMI. Total ischemic time and door-to-balloon time were the time elapsed from symptom onset (the time when chest pain became more intense and sustained) and presentation to the hospital or the passage of the coronary guidewire, respectively. Peripheral arterial disease (PAD) was considered if the patient had peripheral claudication and established aorto-iliac or peripheral disease.

Clinical follow-up was performed by record-linkage and ascertained by electronic records to check for the occurrence of a MACCE comprising death (any cause), a cerebrovascular accident (brain imaging was mandatory), new myocardial infarction in any vessel, or target lesion revascularization (TLR — new intervention on target lesion due to angina or ischemia), during the first year after the index STEMI. Patients having any of the aforementioned MACCE were censored. The study was approved by the hospital ethics committee (2019.128[108-DEFI/112-CE]), and the informed consent for the studied cohort was waived due to the retrospective nature of the analysis. The database was anonymized.

Statistical analysis

To identify the independent predictors of NI we ran a stepwise multivariable logistic regression that included variables with a P <0.1 in the univariable analysis. Cox proportional hazard models were used to identify predictors of MACCE during the follow-up, variables with a P <0.1 on univariable analyses were included in multivariable equations. The presence of possible interactions between NI and all the other variables was tested. Statistical analysis was conducted using Statistical Package for Social Sciences (SPSS version 25.0) and a two-tailed P <0.05 was considered significant for all tests.

RESULTS

From January 2008 to December 2017, of the 1150 STEMI consecutive patients screened, 12 were excluded for presenting an infection at the time of admission and 7 for developing an infection <48 hours after admission. From the 1131 patients included in the study, 126 (11.1%) developed a NI, mostly of respiratory (50.8%) and urinary (39.7%) tract origin (Figure 1). The median time until the diagnosis of NI was 3 days (IQR, 2–6).

Patients who developed a NI were older, more often men, non-smokers, and had more comorbidities. They also had lower hemoglobin, lower creatinine clearance, and lower systolic blood pressure on admission, as well as a higher Killip class during the hospital stay and a higher TIMI score for STEMI on arrival (Table 1). NI was also related to the use of an intra-aortic balloon pump (IABP), whereas utilization of a radial approach and staged PCI for multivessel coronary disease were more prevalent in patients without NI. Length of hospital stay was significantly longer in patients with NI (median 6 vs 12 days). There were no other significant differences between groups.

Predictors of NI

Diabetic patients on insulin therapy were approximately 3 times more likely to develop in-hospital infection (odds ratio [OR], 3.40; 95% confidence interval [CI], 1.53–7.56; P = 0.003), however, this association was not significant for non-insulin treated diabetes (OR, 1.25; 95% CI, 0.73–2.15; P = 0.41). Other predictors of NI were PAD (OR, 2.74; 95% CI, 1.52–4.95; P = 0.001) and the need for an IABP (OR, 3.09; 95% CI, 1.12–8.47; P = 0.03). NI was also statistically more prevalent in older patients (OR, 1.05 per year of age; 95% CI, 1.02–1.07; P <0.001), those with lower systolic blood pressure on admission (OR, 0.99 per mm Hg rise; 95% CI, 0.98–1.00, P = 0.002) and those who had a higher peak creatine-kinase (CK) activity (OR, 1.12 per unit rise; 95% CI, 1.03–1.22; P = 0.01). On the contrary, PIA was negatively related to NI (OR, 0.56; 95% CI, 0.33–0.95; P = 0.03).

Impact of NI on outcomes

As observed on the Kaplan-Meier survival curve (Figure 2), in a 1-year follow-up, the occurrence of MACCE was more than twice as common in the NI group: 47 (37.3%) vs 193 (14.6%), log-rank P <0.001, driven by a larger difference in the first month after STEMI. The statistically significant difference between patients with and without NI is consistent for all composite events of MACCE, except target lesion revascularization (Supplementary material).

Table 3 shows the proportional hazard Cox analysis for predictors of MACCE at 1-year follow-up.

DISCUSSION

Our study reveals that 11.1% of STEMI patients had a NI during the hospital stay, a prevalence lower than reported by some studies in mixed populations undergoing PCI (from 16% to nearly 30%) [6, 10, 11], but higher than others (from 2.4 to 5%) [2, 3, 12]. This may reflect the inhomogeneous definitions of hospital-acquired infection and various indications for PCI (from stable disease to STEMI).

Consistent with most studies [6, 10, 12], pulmonary and urinary tract infections were the most frequent NI site. Even though primary PCI carries vascular invasiveness, the incidence of bloodstream infections was low. As expected, we observed a prolonged hospital stay in infected patients compared to the non-infected group.

Age, diabetes on insulin therapy, PAD, insertion of an IABP, low blood pressure, and high peak CK were identified as factors that favor infection. This comes as no surprise, since it may reflect patients with larger infarcts, requiring more invasiveness, as well as those who are more prone to infections (the elderly, the diabetics, and those with established PAD). Insulin therapy most likely works as a marker of diabetes progression, signaling patients with aggravated immune, vascular, and neurological dysfunction, rather than representing a direct result of prior antidiabetic therapy on NI risk [13–15].

The insertion of an IABP can understandably lead to an increase in bacteremia, wound, respiratory, and urinary tract infections as these patients frequently require a prolonged stay in intensive care units. IABP’s complications have a considerable discrepancy of prevalence reported in the literature (0.9% to 7%) [16–19]. Also, IABP’s correlation to NI likely reflects STEMI’s severity (hemodynamic instability or cardiogenic shock), rather than an effect of the IABP itself. According to previous reports, a femoral rather than radial approach is associated, not only with a higher rate of bleeding and vascular complications, but morbidity and mortality as well [20–22]. However, the correlation with NI reported in our study may likely be related to the operator’s preference in patients who arrive unstable to the catheterization laboratory, and so the reasoning behind the cause-and-effect relationship to predict infection may be the same as for the IABP.

PIA was a protective characteristic. It is likely related to the smaller infarct size caused by preconditioning which limits the reperfusion-injury phenomenon [23, 24], rather than having a direct influence on the development of a NI.

Our analysis also showed a significant interaction between infection and smoking, seemingly not related to the infection site (namely, respiratory or urinary). Since smoking contributes both to the development of infection and cardiovascular disease in the long term [25, 26], and mortality from infection was also reported to be higher in smokers [27], this signals a tendency for a synergic effect between infection and smoking on MACCE. However, it is also reasonable to speculate that a nosocomial infection is more a sign of an underlying lung dysfunction then aggravated by heart insufficiency translated in mid-term events, than a causal risk factor per se.

Limitations

The 48 hour cutpoint used for the definition of nosocomial infection is debatable, however, it is widely accepted and utilized in the literature. A major limitation of our study is the relationship between common risk factors to predict adverse events and the risk of a NI. The relative impact of NI in follow-up is, therefore, difficult to filter, despite the confounding variables incorporated in the multivariate equation and interaction analysis. Another limitation is the well-known limitation of a retrospective analysis, with its inherent bias to assume a cause-and-effect relationship between NI and outcomes. Lastly, being a single-center cohort study, the results may not be representative of all patients with STEMI undergoing PCI.

CONCLUSION

Our study determined that NI is a relatively common complication of STEMI (11.1%), with most risk factors that predict NI also being related to mid-term adverse events. NI does not constitute an independent predictor of MACCE, however, its occurrence during the first year was more than two times higher in smokers who complicate with a NI.

Supplementary material

Supplementary material is available at https://journals.viamedica.pl/kardiologia_polska.

Article information

Conflict of interest: None declared.

Open access: This article is available in open access under Creative Common Attribution-Non-Commercial-No Derivatives 4.0 International (CC BY-NC-ND 4.0) license, allowing to download articles and share them with others as long as they credit the authors and the publisher, but without permission to change them in any way or use them commercially. For commercial use, please contact the journal office at kardiologiapolska@ptkardio.pl.

How to cite: Santos M, Oliveira M, Vieira S, et al. Predictors and mid-term outcomes of nosocomial infection in ST-elevation myocardial infarction patients treated by primary angioplasty. Kardiol Pol. 2021; 79(9): 988–994, doi: 10.33963/KP.a2021.0058.

REFERENCES