Study population and patient characteristics
63 inpatients with ≥ one P. aeruginosa blood culture positive were enrolled retrospectively during this study period. Eleven of them were excluded (five cases had incomplete information, five cases infected with other bacteria, one case missed his TTPs). Therefore, 52 cases were analyzed in this study finally (Figure 1).
Median age of these patients was 1.79 (0.43-9.0) years. Median weight was 11 (7.00-27.00) kg, and the male account for 61.5% (32/52). The average of hospitalization stay was 22.52 (9.05-38.3) days. The most common underlying disease were immunosuppression (50.0%, 26/52), followed by neutropenia (46.2%, 24/52), and hypoalbuminemia (42.3%, 22/52). The common complications were pneumonia (50%, 26/52) and meningitis (9.6%, 5/52). The primary origins of infection were respiratory tract infection (42.3%, 22/52), skin and soft tissue infection (15.4%, 8/52), vascular-catheter related infection (15.4%, 8/52), and primary infection (13.5%, 7/52). 20 (38.5%) patients were admitted to intensive care unit. 22 (42.3%) patients were nosocomial. The median of Pitt bacteremia scores was 1.5 (1-4.00). 31 (59.5%) patients were given antibiotic prior to the blood culture, while 16 (30.8%, 16/31) patients had received inappropriate antimicrobial therapy within 24 hours after blood culture. There were 4 (7.7%, 4/52) patients infected with multiple resistant bacteria. The in-hospital mortality was 23.1% (12/52), and 28.8% (15/52) patients had sepsis shock. More details of clinical characteristics are shown in Table 1.
TTP of P. aeruginosa bacteremia in children
Median TTP was 18.74 h (IQR 16.14-20.77). The bar chart is shown in Figure 2. The optimal cut-off of TTP as a surrogate marker was evaluated by ROC analysis. The optimal point for TTP was 17.87 h with 75.0% sensitivity and 72.5% specificity (AUC 0.77, 95%CI 0.604-0.935), indicating a moderate predicting capability (Figure 3). Therefore 18 h was selected as the standard cut-off. The cases were divided into early TTP (TTP≤18 h) and late TTP group (TTP>18 h). The Kaplan–Meier survival curve of patients with the 2 TTP groups is shown in Figure 4 and Figure 5.
Comparison of clinical characteristics between early and late TTP groups
Table 2 shows the characteristics of early and late TTP groups. Early TTP group patients had significant higher in-hospital mortality (42.9% vs 9.7%, P=0.014), higher incidence of sepsis shock (52.4% vs 12.9%, P=0.006), higher Pitt bacteremia scores (3.00 vs 1.00, P=0.046) and more intensive care unit admission (61.9% vs 22.6%, P=0.008). There were more immunosuppression patients in late TTP group as compared to early TTP group (64.5% vs 28.6%, P=0.023). Four MDR were all detected in late TTP group patients. The demographic characteristics, other underling conditions, the complications, origins of infection, nosocomial infection, antibiotic given before blood culture, and length of hospitalization stay were with no remarkable differences (Table 2).
Comparison of clinical characteristics between the survival and the non-survival groups
The median TTP in fatal group was 15.19 (IQR 11.21-18.24) hours, obviously shorter than 19.42 (IQR 16.92-20.97) hours in survival group (P=0.005). Pitt scores in non-survival group were 4.50 (IQR 1.25-7.25), significantly higher than that in survival group [1.00 (IQR 1.00-3.75)]. The incidence of sepsis shock was remarkably higher in non-survival group when compared to survival group patients (58.3% vs 20.0%, P=0.025). More patients had hypoalbuminemia among survival group than fatal group (75.0% vs 32.5%, P=0.023). No significant differences were detected in other clinical characteristics (Table 3).
Risk factors of in-hospital mortality
Univariate analysis revealed that early TTP, Pitt bacteremia scores≥4 and hypoalbuminemia were associated with in-hospital mortality. Multivariate analysis showed early TTP (OR 5.88; 95%CI 1.21-21.96) and Pitt bacteremia scores≥4 (OR 4.95; 95%CI 1.26-27.50) were independently correlated with in-hospital mortality (Table 4).
Risk factors of septic shock
Univariate analysis also indicated that early TTP, Pitt bacteremia scores≥4, hypoalbuminemia and intensive care unit admission were correlated with sepsis shock. Multivariate analysis showed early TTP (OR 6.30; 95%CI 1.18-33.77), Pitt bacteremia scores ≥4 (OR 8.15; 95%CI 1.53-43.32), hypoalbuminemia (OR 6.46; 95% CI 1.19-33.19) were independently associated with sepsis shock (Table 5).