Our study confirmed a high incidence of secondary infection in septic patients (31.0%). Urinary and deep venous catheterization could bring higher risk of developing secondary infection. Secondary infection also affected the prognosis, which featured poor survival at later period (> 15st day after admission). Expected prolonged in-hospital LOS was 4.63 days in secondary infection group. Our study also found that immunosuppression led to higher risk of secondary infection.
A recent meta-analysis revealed that lower respiratory tract infection was the most common nosocomial infection in general hospital and pseudomonas aeruginosa, Escherichia coli, Acinetobacter baumannii and Klebsiella pneumoniae were among most common pathogens[25]. Similar findings were discovered in our study. The high pathogenicity of such Gram-negative bacteria was due to drug resistant and invasive procedures which serves as approach of the invasion of pathogens[26]. Increased susceptibility of secondary infection could also be pathogen-specific[27,28]. However, it’s possible that some of the pathogenic microorganisms were not identified due to limited testing technologies.
We found higher APACHE II and SOFA scores on admission in patients with secondary infection, which was similar to previous studies[3,4,29]. Although illness severity was not found to be independent risk factor of secondary infection. We could not rule out the possibility that the more severely ill patients died mostly at the very early period of disease before developing secondary infections, which might influence the true association between the risk and illness severity.
It’s widely acknowledged that catheter indwelling was a major cause of nosocomial infection and compromised the safety of in-hospital patients[30-32]. We found that urinary catheterization was the independent risk factor of secondary infection. A study revealed that the onset of catheter-associated urinary tract infection in critical ill patients was mostly affected by duration of urinary catheterization and the presence of another site of nosocomial infection, which was confirmed by our study that many cases of secondary infections in urinary tract were subsequent to secondary infections at other sites[33]. Deep venous catheterization was also a common operation in ICU setting. Our finding was consistent with the study by van Vught et al. that it was also an independent risk factor[4]. Also, mechanical ventilation and blood transfusion was found to be associated with the risk of sepsis in univariate analysis. The need for mechanical ventilation of critical ill patients incurred high prevalence of ventilator-associated pneumonia, which accounted for nearly half of nosocomial infections[34]. It was found to be an independent risk factor of developing secondary infection in some studies[3,4]. But the result of multivariate analysis of our study was not consistent with previous findings. As we did not tell apart invasive and non-invasive ventilation and the duration of ventilation was not recorded, the result should be interpreted cautiously. The association between blood transfusion and secondary infection of sepsis could be induced by transfusion-related immune modulation (TRIM) caused by the altered immune function[35-38]. However, transfusion was not found to be an independent risk factor in our study and it could be affected by type and quantity of transfusion.
Secondary infection was observed to prolong the time of hospitalization under a multistate model, which we thought was an result of the complexity of disease requiring longer in-hospital treatment caused by secondary infection. When it comes to mortality, secondary infection was an independent risk factor of in-hospital death in this study. Survival analysis demonstrated that patients who developed secondary infection had worse outcome after first 15 days than those did not. Before 15 days, secondary infection group even had better survival. That could be explained by that patients who were severely sick died earlier than they developed secondary infections. This was consistent with the previous concept that the mortality of patients who survived that early period was more affected by secondary infection[13]. A re-increased microbiological burden revealed by positive blood cultures at later phase of sepsis (> 15 days) was observed in the study by Otto et al., which was indicative of secondary infection and the risk of poor outcomes[39]. As suggested by van Vught LA et al., septic patients without secondary infection had lower mortality at day 60 when compared to overall septic patients.4 However, Goldenberg et al. addressed that secondary infection was not the main cause of death in sepsis as they found only a small portion (14%) of septic patients died with an evidence of secondary infection. Some studies found that mitochondrial dysfunction, microvascular leak or even activity of daily living could serve as causes of death of sepsis[5,27].
Immune status of septic patients and its underlying mechanism has been widely studied. Innate immune function was altered after sepsis as a dysfunction of neutrophils, monocytes, dendritic cells and myeloid-derived suppressor cells (MDSCs) and cause altered first-line of defense, inhibition of T cell proliferation, altered inflammatory response and incomplete activation of T cells[8]. Adaptive immune function was also altered as sepsis affects the effector functions and phenotypes of T cells, B cells and innate-type lymphocytes[8]. HLA-DR was one of the makers reflecting the function of both innate and adaptive immune function and lower expression indicates immunosuppression[8]. In this study, HLA-DR expression was lower in secondary infection group at day 3. Also, It’s noteworthy that we found a dramatic increase of IL-10 level from day 1 to day 7 in secondary infection group. IL-10 was an anti-inflammatory cytokine and elevated level reflected the down-regulation of inflammation process. It might generate MDSCs and enhances the immunosuppression during sepsis[20,40]. Thus, our study again illustrated the immunosuppression in secondary infection group. In the early stage of disease, we observed a more severe inflammatory response in secondary infection group presented as a higher IL-6 and IL-8 level, which were both pro-inflammatory cytokines. This confirmed that hyper inflammation would happen during the early phase of disease in septic patients with secondary infection and higher pro- and anti- inflammatory process might exist at the same time[21,23]. Those enlightened us that the methods of identification of entering the phase of immunosuppression and the therapies that boost immunity could be potentially beneficial to the prevention of secondary infections of septic patients[13,41].
This study had some limitations. First, the sample size was relatively small as a single-center study. Second, some clinical data such the use of antibiotics, the exact dose of glucocorticoids, length of the use of mechanical ventilation and urinary catheterization were not documented due to the limited medical record. However, the influence of therapy plan cannot be ignored. Third, data of HLA-DR expression and serum cytokine levels of many patients were not available as a retrospective study. Thus the association between immunosuppression and secondary infection was worthy of further research with larger sample size.