Introduction

Pseudomonas aeruginosa stands out as a significant cause of bacteraemia, particularly in nosocomial infections, posing a heightened risk to immunocompromised patients, especially those with neutropenia [1,2,3,4,5,6]. Bacteraemia associated with P. aeruginosa carries a notably high mortality rate [1,2,3,4,5,6,7,8,9,10,11]. In fact, among bacteraemias caused by various Gram-negative aerobic bacteria, it exhibits the highest mortality, comparable to that of Staphylococcus aureus bacteraemia [1, 4, 5].

Several factors common to any pathogen causing bacteraemia are also associated with poorer outcomes among patients with P. aeruginosa bacteraemia, such as advanced age, presence of comorbidities, and sepsis [2, 3, 6,7,8, 10, 11]. A key predictor of mortality is the use of inappropriate empirical antimicrobial treatment, associated to the increased capacity of resistance development of P. aeruginosa [2, 6, 10, 11]. Indeed, multidrug-resistant P. aeruginosa is associated with increased mortality compared to susceptible strains [6, 9, 10]. Despite this, previously overlooked factors in studies is the timely implementation of source control and the impact of infectious diseases (ID) consultation. While the beneficial roles of the aforementioned interventions have been established for S. aureus bacteraemia and candidemia [12,13,14], only two studies have evaluated the impact of ID consultation in P. aeruginosa bacteraemia [8, 11]. Both studies demonstrated its effectiveness, as it influenced both antimicrobial treatment and source control procedures, underscoring its potential for improving patient outcomes [8, 11].

Our objective was to determine predictors of mortality in patients with P. aeruginosa bacteraemia, specifically by examining the effects of various interventions, including ID consultation, administration of appropriate antimicrobial treatment, and the implementation of source control procedures when deemed necessary.

Materials and methods

This retrospective study was conducted at Lausanne University Hospital, Switzerland from 2015 to 2021. Inclusion criteria were adult patients (≥ 18 years old) and presence of at least one blood culture for Pseudomonas aeruginosa. The sole exclusion criterion was the patient’s refusal of use of clinical data.

BACTEC™ FX BacT/ALERT System (Becton, Dickinson and Company, Franklin Lakes, USAbioMerieux, Marcy l’Etoile, France) was used for incubation of blood cultures. For rapid identification, bacterial pellet preparations were performed on the positive blood cultures (2015 to 2018 based on Croxatto et al. method, and from 2018 with the rapid BACpro® II Nittobo Medical Co., Tokyo, Japan) [15]. Bacterial pellets were identified with a Matrix-assisted laser desorption-ionization time of flight mass spectrometry (MALDI-TOF MS; Bruker Daltonics, Bremen, Germany) was used for species identification from 07:00 to 19:00. Susceptibility results (Vitek® 2, bioMerieux Marc l’Etoile, France and Kirby-Bauer methods) were obtained from the microbiology laboratory database and assessed in accordance with the EUCAST criteria [16].

Fourteen-day all-cause mortality was the primary outcome. Data on demographics (age, sex), comorbidities, Charlson Comorbidity Index, ID consultation, antimicrobial treatment, source control, the presence of sepsis, and the site of infection were retrieved from patients’ electronic health records. All data were collected, stored and managed using REDCap by an ID specialist. REDCap electronic data capture tools are hosted at Lausanne University Hospital. REDCap (Research Electronic Data Capture) is a secure, web-based software platform designed to support data capture for research studies [17, 18].

In our institution, ID consultants receive notification regarding patients with positive blood cultures following species identification. For P. aeruginosa bacteraemia, ID consultation is not mandatory. However, patients undergoing induction or consolidation/maintenance chemotherapy, those with lymphoma or multiple myeloma undergoing autologous hematopoietic stem cell transplantation are hospitalized in the ID service while those after allogeneic stem cell or solid organ transplantation are closely monitored by ID consultants. The decision to perform follow-up blood cultures is at the discretion of the treating physician.

The date of collection of the first positive blood culture was defined as bacteraemia onset. A new episode was included if more than 30 days had elapsed since the cessation of antibiotic treatment for the initial bacteraemia. The Sepsis-3 International Consensus criteria were used for sepsis definition [19]. The determination of the infection site was based on the assessment by the ID consultant, or treating physician, taking into account clinical, radiological, microbiological, and operative findings. Appropriate antimicrobial treatment was defined as the initiation of at least one antimicrobial agent with in vitro activity against the infecting isolate. Source control was deemed necessary in the following situations: removal of venous catheter in patients with catheter-related bacteraemia or bacteraemia of unknown origin in the presence of a venous catheter; imaging-guided or surgical drainage of infected collections, such as abscesses, empyema, etc.; joint fluid drainage; cardiac surgery in endocarditis patients when indicated for heart failure; and correction of urinary-tract obstruction. We used the cutoff of 48 h to define early interventions (antimicrobial treatment initiation, source control, ID consultation) from bacteraemia onset Multidrug-resistance was defined as resistance to at least one agent in three or more antimicrobial categories [20]. Time to positivity of blood cultures was evaluated in episodes where at least one bottle grew only P. aeruginosa.

SPSS version 26.0 (SPSS, Chicago, IL, USA) were used for data analyses. Categorical variables were analyzed using the chi-square or Fisher exact test and continuous variables with Mann–Whitney U test. Two univariate logistic regression models were assessed with 14-day and 30-day mortality as dependent variables. Clinically relevant non collinear covariates, assessed through variance inflation factor, were used in multivariable analysis. After checking Cox assumptions, two multivariable Cox proportional hazards regression models were performed with 14-day and 30-day mortality as the time-to-event. Adjusted Hazzard ratios (aHRs) and 95% confidence intervals (CIs) were calculated to evaluate the strength of any association. All statistic tests were 2-tailed and P < 0.05 was considered statistically significant. We finally performed Kaplan-Meier curves of the survival probability of patients with P. aeruginosa bacteraemia according to appropriate source control within 48 h from bacteraemia onset.

Results

Among the 308 episodes of P. aeruginosa bacteraemia, 278 episodes were included, involving 261 patients (Fig. 1). Twelve and five patients had two and three episodes, respectively. These 22 subsequent episodes of bacteremia, occurred at a median of 2 months from the previous episode (interquartile range: 2–3 months). Resistance was observed in 36 (13%) isolates for piperacillin/tazobactam, 37 (13%) for ceftazidime, 28 (10%) for cefepime, 39 (14%) for imipenem, 23 (8%) for meropenem, 9 (3%) for amikacin, and 31 (11%) for ciprofloxacin. Twenty (7%) isolates were classified as multidrug-resistant. Resistance to colistin was observed in 5% (one isolate among 22 tested), to ceftazidime/avibactam in 40% (6 out of 15), and to ceftolozane/tazobactam in 29% (5 out of 17). The susceptibility profiles and the respective administered antimicrobial treatment of the 20 episodes due to multidrug-resistant P. aeruginosa isolates are shown in Supplementary Table 1. Follow-up blood cultures were conducted until sterilization in 221 (80%) episodes, among which 28 (13%) exhibited persistent bacteraemia for at least 48 h.

Fig. 1
figure 1

Flowchart of patients’ inclusion

Full size image

The most common type of infection was low respiratory tract infection (58 episodes; 21%), followed by catheter-related (54; 19%) and abdominal infection (54; 19%). Sepsis was present in the majority of episodes (152; 55%).

ID consultation was performed in 224 (81%) episodes, with 203 (73%) consultations provided within 48 h of bacteraemia onset. Appropriate antimicrobial treatment was administered within 48 h of bacteraemia onset in 257 (92%) episodes. The impact of ID consultation on antimicrobial treatment is shown in Supplementary Table 2. Among the 80 episodes where appropriate antimicrobial treatment was not administered within 24 h of bacteraemia onset, 49 (61%) had an early ID consultation. More episodes in the early ID consultation group had appropriate antimicrobial treatment initiated within 48 h compared to those without ID consultation in the same timeframe (90% versus 52%; P < 0.001). Among the 60 episodes where appropriate treatment by carbapenems was administered within the first 24 h, 48 (80%) had early ID consultation. More episodes in the early ID consultation group had a de-escalation to ceftazidime, cefepime, or piperacillin/tazobactam within 48 h compared to those without ID consultation in the same timeframe (48% versus 8%; P < 0.001). For the majority of episodes (145; 52%), source control was considered necessary, with 93 (64%) of them undergoing such interventions within 48 h.

The overall 14-day mortality was 15% (42 episodes), while 30-day mortality was 22% (60 episodes). Table 1 shows the comparison of survivors and not survivors at day 14. Sepsis was more prevalent among deceased patients (93% versus 48%; P < 0.001). Lower respiratory tract infections were more common among deceased patients (60% versus 14%; P < 0.001). Early ID consultation (50% versus 77%; P 0.001) and source control (25% versus 68%; P < 0.001) were less frequently observed in deceased patients. No association was found between multidrug-resistance, time to blood culture positivity, appropriate antimicrobial treatment within 48 h, and 14-day mortality.

Table 1 Comparison of survivors and not survivors at day 14
Full size table

The Cox multivariable regression model of 14-day mortality is presented in Table 2. Fourteen-day mortality was associated with sepsis (P 0.002; aHR 6.58, CI 1.95–22.16), and lower respiratory tract infection (P < 0.001; aHR 4.63, CI 1.78–12.06). Conversely, interventions performed within 48 h of bacteraemia onset, such as ID consultation (P 0.036; HR 0.51, CI 0.27–0.96), and source control (P 0.009; aHR 0.17, CI 0.47–0.64) were associated with improved outcome.

Table 2 Univariable and multivariable Cox proportional hazard regression of 14-day mortality among patients with bacteraemia due to P. aeruginosa
Full size table

Figure 2 illustrates Kaplan–Meier survival probability curves for episodes with P. aeruginosa bacteraemia categorized by the requirement for and execution of source control within 48 h. Source control within 48 h was linked to a more favorable outcome when compared to episodes where it was warranted but not performed (Log-rank test, P 0.004).

Fig. 2
figure 2

Kaplan–Meier curves of the survival probability curves for episodes with P. aeruginosa bacteraemia categorized by the requirement for and execution of source control within 48 h. Comparison of episodes where source control was performed versus those where it was indicated but not performed: Log-rank test, P 0.004. Comparison of episodes where source control was performed versus those where source control was not warranted: Log-rank test, P < 0.001. Comparison of episodes where source control was not performed versus those where source control was not warranted: Log-rank test, P 0.267

Full size image

The comparison of survivors and not survivors at day 30 and the Cox multivariable regression model of 30-day mortality are presented in Supplementary Tables 3 and 4, respectively. Thirty-day mortality was associated with a Charlson comorbidity index superior to 4 (P 0.001; aHR 2.83, CI 1.58–5.09), sepsis (P 0.002; aHR 7.48, CI 1.52–6.84), and lower respiratory tract infection (P < 0.001; aHR 4.60, CI 2.15–9.81). Conversely, interventions performed within 48 h of bacteraemia onset, such as ID consultation (P 0.049; HR 0.59, CI 0.35–0.99), and source control (P 0.009; aHR 0.29, CI 0.11–0.74) were associated with improved outcome.

Discussion

In this study, we investigated predictors of mortality among patients with P. aeruginosa bacteraemia and underscored the importance of ID consultation and timely source control in enhancing outcomes.

The mortality rate in our cohort (22%) was similar to previous research [3,4,5, 8, 11], but lower than studies conducted in regions with higher resistance rates [6, 7, 9, 10]. The relatively low incidence of multidrug resistance among P. aeruginosa isolates, coupled with close monitoring by ID specialists for high-risk patients (such as those undergoing chemotherapy or transplantation), likely contributed to the higher rate of appropriate antimicrobial treatment within the first 48 h of bacteraemia onset in the present study, compared to prior research [2, 3, 6, 7, 10, 11].

The most important finding was the positive impact of ID consultation on managing P. aeruginosa infections, particularly in guiding antimicrobial therapy and source control measures. Consistent with two previous studies on P. aeruginosa bacteraemia, our findings emphasize the importance of ID consultation in improving patient outcomes [8, 11]. The aforementioned studies suffered from survival bias, with only patients surviving long enough receiving ID consultation. One of these studies did not include a timeframe for the realization of ID consultation [8], while the other included ID consultation within two weeks from the onset of bacteraemia [11]. This issue was addressed in our study by implementing a timeframe of 48 h from the onset of bacteraemia. In the present study, the early involvement of ID consultants led to an earlier initiation of appropriate treatment and earlier de-escalation from carbapenems to other beta-lactams with a narrower spectrum. Although the dosages of the administered antimicrobials were not collected in the present study, a previous study from our institution showed that ID consultation among all P. aeruginosa infections led to an increase in the administered dosages of non-carbapenem antibiotics [21]. However, the influence of ID consultation might have been underestimated in our study, as patients at higher risk of mortality, such as those with neutropenia following chemotherapy for hematologic malignancies or undergoing transplantation, were either closely monitored by ID consultants or hospitalized in the ID service, thus de facto receiving consultation within 48 h [3].

In a prior study involving patients with infections due to multidrug-resistant pathogens, ID consultation proved beneficial for methicillin-resistant S. aureus and multidrug-resistant Enterobacteriaceae, but not for P. aeruginosa, likely due to the limited number of patients included with infection caused by this pathogen [22]. Furthermore, although previous research extensively evaluated the effect of appropriate empirical antimicrobial treatment on survival in P. aeruginosa bacteraemia, the role of early source control measures among P. aeruginosa bacteraemic patients was rarely investigated [8]. The impact of timely source control on better outcomes (survival, clearance of bacteraemia or candidaemia) has been demonstrated in various types of infections, such as intra-abdominal infections, necrotizing fasciitis, sepsis, and bloodstream infections caused by different pathogens, including S. aureus, streptococci, and Candida spp [13, 14, 23,24,25,26]. Especially in sepsis which was associated with poorer outcomes in our study, prompt source control is recommended by the Surviving Sepsis Campaign Guidelines to enhance management and improve outcomes [27]. This recommendation is supported by a study on critically ill patients, regardless of the causative pathogen [28].

As previously demonstrated, an elevated Charlson comorbidity index, which encompasses age and various comorbid conditions, was correlated with increased 30-day mortality [1,2,3, 7, 8, 10, 11]. Previous research has highlighted the significance of the focus of infection in determining outcomes, with bacteraemias secondary to lower respiratory tract or pulmonary infections associated with worse prognoses [1, 2, 29, 30].

In prior studies, the time to positivity of blood cultures, serving as an indirect measure of the microbial load of the infecting organism, was found to be indicative of poorer outcomes among patients with P. aeruginosa bacteraemia [29, 30]. No such correlation was identified in the current study, underscoring the limitations in evaluating the time to positivity of blood cultures. Variability in protocols for blood culture drawing, processing, and incubation across different centers can lead to discrepancies in the time to positivity results.

The current study is subject to several limitations. Firstly, it was a retrospective analysis conducted at a single center in a setting with low resistance rates, thus caution must be exercised when generalizing our findings, particularly to settings with higher resistance rates. Moreover, factors such as decisions to restrict treatment and the readiness of surgeons or interventional radiologists to undertake source control procedures could influence the relationship between outcome and prompt interventions. In our study, among episodes warranting source control intervention, only three patients succumbed or had care withdrawn within the 48-hour timeframe, minimizing their impact on our results.

In conclusion, we have demonstrated the beneficial role of timely ID consultation in patients with P. aeruginosa bacteraemia, aligning it with other pathogens such as S. aureus, Candida spp., enterococci, and streptococci, thereby advocating for ID consultation as an integral component of management to enhance patient outcomes. Additionally, we have underscored the importance of implementing source control interventions when appropriate. Future studies are needed to evaluate the impact of comprehensive approach with early interventions such as ID consultation, tailored antimicrobial treatment, and source control in enhancing patient outcomes.