Background
Aneurysmal subarachnoid hemorrhage (SAH) is an acute cerebrovascular disease with devastating consequences, including high mortality and long-term functional impairment among survivors [
1]. Several brain-specific mechanisms of injuries secondary to SAH have been associated with worse functional outcomes, including early injuries, vasospasms and delayed cerebral ischemia (DCI), neuroinflammation or impaired cerebral autoregulation [
2‐
6]. However, there are few options for treating or preventing such complications of SAH. On the other hand, systemic (and potentially preventable) complications such as nosocomial infections, sepsis and organ dysfunction may also alter the course of the disease and worsen survival as well as functional capacity [
7]. The incidence of systemic inflammatory response syndrome (SIRS) after SAH may be up to 83% of patients, and the incidence of sepsis varies between 10 and 20%, with a strong association with poor outcomes [
8,
9].
Approximately 30% of patients with SAH will develop nosocomial infection during the course of hospitalization [
10]. However, most of the available data are derived from retrospective studies or cohort studies with imprecise definitions of infections or sepsis. The diagnosis of sepsis is challenging in patients with SAH due to the high incidence of early SIRS and the lack of reliable biomarkers that can be used to differentiate SIRS and sepsis [
8,
9]. Therefore, there is a risk of sepsis misdiagnosis, which can potentially lead to the under- or overtreatment of SAH patients. Considering the previously exposed, the objective of this investigation was to define the incidence of sepsis, diagnosed prospectively with the Sepsis-3 criteria, and its impact on mortality and functional outcomes of patients with SAH.
Discussion
In this study, we present the results of a prospective cohort study of SAH patients with functional outcomes (measured by the mRS score) and mortality collected during a maximum of 12 months of follow-up. We prospectively investigated the incidence of sepsis during the first 14 days of the hospital stay, with an incidence of 28%. Sepsis was independently associated with a threefold increase in poor functional outcomes and long-term (using the last data point for each patient) mortality (52.5% for septic patients and 16% for nonseptic patients—log-rank test p < 0.0001).
Aneurysmal SAH is a severe cerebrovascular event that may lead to lifelong disabilities. Mortality can be as high as 15% prior to hospital admission and may reach 40 to 45% in the 30 days after the bleed [
3,
18]. After the initial ictus, a number of complications leading to secondary brain injury can further impact outcomes. Sepsis, a complex syndrome where a dysregulated inflammatory response to an infection leads to organ dysfunction, is one of these complications [
7].
Regardless of adjustment, in the multivariable analysis, for known neurological factors that affect outcomes (poor grade, hydrocephalus and DCI), sepsis was still independently associated with an unfavorable outcome in our cohort. Some of these neurological factors, such as poor clinical presentation, depend on the initial brain injury—such as a loss of consciousness, seizures and the bleed volume [
14]. Others, such as DCI, remain poorly understood and have limited therapeutic options. Sepsis, on the other hand, may be prevented through measures that effectively reduce nosocomial infections. Moreover, accurate and early diagnosis of sepsis combined with an aggressive treatment for organ dysfunction may potentially improve outcomes for SAH patients [
3,
19].
Although some studies have shown an association between sepsis and unfavorable outcomes after SAH [
20‐
23], none used the current definition of sepsis (Sepsis-3.0). The previous definitions, which included the SIRS criteria, were less specific and did not clearly discriminate sepsis from a systemic inflammatory response due to noninfectious causes, such as an SAH. Two retrospective studies [
24,
25] found associations between the SIRS burden and worse outcomes in SAH, with SIRS incidence ranging from 69 to 85%. Although present in 82% of our patients, SIRS was not associated with a poor outcome after adjusting for confounders. SIRS could be related to the severity of SAH presentation. Sepsis, on the other hand, could play a role in brain dysfunction developed after SAH, working as a second attack on a vulnerable brain. This phenomenon has been shown in other populations where sepsis leads to brain dysfunction, cognitive impairment, muscular loss and exacerbated end-organ dysfunction [
19,
26].
The incidence of sepsis is higher than previously reported [
8,
9]. The incidence of infection was also higher than previous reports (38%), with 45% of those infections presenting as pneumonia. This study is the first large cohort study of SAH patients with a sepsis diagnosis from low- and middle-income countries, which may explain the differences from the current literature, which mostly comprises European and North American data. Recently, a multicenter study from Brazil [
27] has evaluated the national prevalence of sepsis on a single day and reported an incidence of 16.7% and a prevalence of 25% (meaning 30% of patients were septic on the day of the study). The in-hospital mortality rate was 56%. In contrast, the PRISM meta-analysis [
28], which reported data from the USA, the UK and Australasia, showed a sepsis-related mortality rate of 25%. Again, data on sepsis, with the use of new criteria, is scarce in low- and middle-income countries and almost nonexistent in the neurocritical population, which may limit such comparisons.
Our study has limitations. First, our institution is a reference center, and the majority of patients are admitted 24 h after ictus. The delay in transfer may introduce selection bias, as patients with a very severe disease presentation, such as evidence of intracranial hypertension, may die before reaching our ICU. Also, that those patients who died early were excluded from our sample may be a reason for the overestimation in the incidence of sepsis in this cohort. Additionally, being a convenience sample, our cohort may be underpowered. Second, follow-up data were acquired for 136 patients (95 patients at 12 months, 28 patients at 6 months and 13 patients at 3 months), with 13 patients lost to follow-up. Third, sepsis diagnosis remains difficult in these patients, even with the new criteria. Confounders such as vasoactive drugs for induced hypertension, neurogenic pulmonary edema and stress-induced myocardial depression may simulate sepsis-induced organ dysfunction. In this population, it is even more difficult due to the neurological impairment due to the bleeding itself, which can elevate the SOFA score (specifically, the neurological SOFA) without sepsis, confounding the diagnosis. To avoid such bias, our sepsis diagnosis was validated by an external specialist (an infectious diseases’ specialist), and for the neurological SOFA, only the rise in the score was considered (not the baseline, if already altered) and when not due to SAH complications (such as DCI or hydrocephalus for instance). Finally, we do not have data on all of the patients’ brain images after discharge, so we cannot explore the hypothesis of sepsis worsening brain damage in more detail—however, this was already addressed by other authors [
26] in previous studies.
Though the data indicate a relationship between sepsis and worse outcomes in this cohort, it is still scarce to define a cause–effect association with certainty, especially since the septic group was clearly more severe than the nonseptic group on presentation. However, the multivariate analysis still shows sepsis as an independent variable of worse outcomes. Thus, our study in fact serves to formulate and empower this hypothesis—further studies, with larger cohorts, should be analyzed in order to better clarify the matter.
That aside, a number of potential opportunities arise from a better understanding of the impact of sepsis in patients with SAH. Future studies should aim to develop novel and more accurate methods to improve sepsis diagnosis, such as early molecular detection of bacteria in the blood or monitoring of blood biomarkers (e.g., procalcitonin and C-reactive protein) [
9,
10]. Additionally, strategies that aim to reduce early pneumonia, such as the preemptive administration of antibiotics in comatose patients, may potentially reduce the incidence of sepsis and benefit poor grade SAH patients [
29].
Sepsis mortality has dropped in recent years in the general ICU population, mainly due to early recognition and effective treatment [
30]. Our study shows an association between sepsis and outcomes in SAH; though it is underpowered to determine a cause–effect relationship, it still provides data to create this hypothesis. Based on our results, we believe that nosocomial infection prevention, improved diagnosis and optimal management of infection and sepsis in SAH patients may have a major impact on patients’ outcomes, especially in low- and middle-income countries.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.