Hostname: page-component-586b7cd67f-r5fsc Total loading time: 0 Render date: 2024-11-26T07:06:31.507Z Has data issue: false hasContentIssue false

Is early time to positivity of blood culture associated with clinical prognosis in patients with Klebsiella pneumoniae bloodstream infection?

Published online by Cambridge University Press:  21 February 2023

Weiwei Hou
Affiliation:
Department of Laboratory Medicine, Tongji Hospital, School of Medicine, Tongji University, Shanghai 200065, China
Tiantian Han
Affiliation:
Department of Hospital Infection Control, Tongji Hospital, School of Medicine, Tongji University, Shanghai 200065, China
Guangbo Qu
Affiliation:
Department of Epidemiology and Biostatistics, School of Public Health, Anhui Medical University, Hefei, 230032, Anhui Province, China
Yehuan Sun
Affiliation:
Department of Epidemiology and Biostatistics, School of Public Health, Anhui Medical University, Hefei, 230032, Anhui Province, China
Dianyu Yang
Affiliation:
Department of Laboratory Medicine, Tongji Hospital, School of Medicine, Tongji University, Shanghai 200065, China
Yan Lin*
Affiliation:
Department of Hospital Infection Control, Tongji Hospital, School of Medicine, Tongji University, Shanghai 200065, China
*
Author for correspondence: Yan Lin, E-mail: [email protected]
Rights & Permissions [Opens in a new window]

Abstract

The association between time to positivity (TTP) of blood culture and the clinical prognosis of patients with Klebsiella pneumoniae bloodstream infection (BSI) remains unclear. A retrospective study of 148 inpatients with BSI caused by K. pneumoniae was performed at Shanghai Tongji Hospital, China, from October 2016–2020. The total in-hospital fatality rate was 32%. The median TTP was 11.0 (7.7–16.1) h and the optimal cutoff for prediction of in-hospital mortality was 9.4 h according to the ROC curve. Early TTP (<9.4 h) was a risk factor for in-hospital mortality by univariate analysis (OR = 2.5, 95% CI 1.2–5.0, P = 0.01), but not by multivariate analysis (OR = 2.7, 95% CI 1.0–7.4, P = 0.06). Old age, serum creatinine, white blood cells, and C-reactive protein values were risk factors for in-hospital mortality by multivariate analysis. Early TTP was not a risk factor for septic shock (OR = 1.8, 95% CI 0.6–5.1, P = 0.27) or ICU admission (OR = 1.0, 95% CI 1.0–1.0, P = 0.32). In conclusion, the in-hospital fatality rate of patients with K. pneumoniae BSI was relatively high and associated with an early TTP of blood cultures. However, no increased risk of mortality, septic shock or ICU admission was evident in early TTP patients.

Type
Original Paper
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
Copyright © The Author(s), 2023. Published by Cambridge University Press

Introduction

Klebsiella pneumoniae (K. pneumoniae), a natural resident of the normal human faecal microbiome, can cause community and hospital-acquired infections such as pneumonia, liver abscess, bacteraemia and intraperitoneal and urinary tract infections [Reference Calfee1]. Multidrug resistance is common in the species, ranging from 36.0% to 73.2% and showing an increasing trend [Reference Baral2Reference Brady4]. K. pneumoniae bloodstream infection (BSI) is more likely to occur in patients who require relatively long-term health care, and is associated with high mortality [Reference Brady4], particularly if caused by carbapenem-resistant strains [Reference Xu, Sun and Ma5]. Therefore, appropriate indicators are needed to assess the possible prognosis for patients and inform therapy for such infections.

Time to positivity (TTP) is the time interval from incubation of a blood culture sample to a positive signal of bacterial growth, which is associated with bacterial load [Reference Hamilton6]. Previous studies have emphasised the value of TTP in the diagnosis of catheter-related BSIs [Reference Al-Juaid7], potential differentiation of pathogens [Reference Ning8] and for some species a predictor of death in patients with BSI [Reference Martín-Gutiérrez9] caused by various pathogens such as Staphylococcus aureus [Reference Siméon10] and Pseudomonas aeruginosa [Reference Xu11]. A short TTP (<7 h) was found to be significantly associated with mortality at all time points after admission for patients with K. pneumoniae monomicrobial bacteremia [Reference Liao12], and an early TTP (≤13 h) was identified as an independent risk factor for in-hospital mortality for paediatric patients with this organism [Reference Cheng13]. However, no direct evidence has been presented of an association between TTP and 28-day mortality for patients with Klebsiella spp. BSI, in general [Reference Hamilton6], and indicates the need for further studies of such a correlation, as well as the most appropriate cut-off TTP value for these patients.

As a consequence, the present study was undertaken to investigate further the correlation between TTP values of blood cultures and clinical prognosis, verify its capacity as a predictor of in-hospital mortality, and determine the most appropriate cut-off value to inform mortality among patients with K. pneumoniae BSI.

Methods

Study design and patients

A retrospective study was performed at Shanghai Tongji Hospital, China. Patients with K. pneumoniae monomicrobial BSI were selected from clinical and laboratory records from October 2016 to October 2020. The inclusion criteria for subjects were: inpatients with a positive blood culture for K. pneumoniae and with recorded TTP values. Patients were excluded if they were admitted from the ICU or emergency department and if they fulfilled any of the outcomes such as death or septic shock, before the blood culture or had polymicrobial BSIs.

Definitions

K. pneumoniae BSI was defined as a positive blood culture result for the species. Antimicrobial resistance of isolates was categorised as either CRKP (resistant to carbapenems) or multi-drug resistant (MDRKP, resistant to ≥3 different classes of antibiotic agents). TTP was defined as the time interval from incubation of the blood culture sample to a positive signal, and the shortest value was recorded when multiple positive signals were detected. Septic shock was diagnosed according to Third International Consensus Definitions for Sepsis and Septic Shock [Reference Shankar-Hari14].

Microbiological methods

Blood samples (10 ml) from each patient were collected in aerobic and anaerobic culture bottles according to the standard procedure at Shanghai Tongji Hospital. Samples were loaded into the Bactec FX400 blood culture system (Becton Dickinson, Franklin Lakes, NJ, USA), and positive bottles were subcultured and examined by Gram stain. Isolates were identified to species level by matrix-assisted desorption/ionisation time-of-flight mass spectrometry (MALDI-TOF MS), performed using Microflex LT (Bruker Daltonics, Bremen, Germany), and antibiotic susceptibility was determined in the VITEK-2 compact system (bioMerieux, Marcy L'Etoile, France).

Data collection

The following anonymous information was collected for each subject: age, sex, source department, comorbidity, drug resistance, TTP, antibiotic therapy, haemodialysis, transfusion, surgery, length of hospital stay and blood examination indices, including serum creatinine (SCR), total bilirubin (TBIL), mean arterial pressure (MAP), blood platelet count (PLT), procalcitonin (PCT), white blood cell count (WBC) and C-reactive protein (CRP).

Statistical analyses

The median and interquartile range (M, P25–P75) or frequency and proportion (N, %) were used to describe the basic information of the enrolled subjects, and differences between groups were compared using the Mann–Whitney U test and χ 2 or Fisher's exact probability test. A bar chart was drawn to map the distribution of TTP values, and a receiver operating characteristic (ROC) curve along with an area under the curve (AUC) analysis was developed to explore the predictive effect of TTP on in-hospital mortality. Youden's index was calculated to estimate the appropriate TTP cutoff value. Univariate logistic regression analyses were conducted on all variables and those with statistical significance were entered into multivariate logistic regression analyses. TTP values were included into all the latter analyses regardless of their statistical significance by univariate analyses, to explore correlations with in-hospital mortality, septic shock and ICU admission. Odds ratios (OR) and corresponding 95% confidence intervals (95% CI) were calculated, and a two-sided P value >0.05 was statistically significant. Data were analysed using SPSS 24.0 software (IBM Corporation, Armonk, NY, USA).

Results

A total of 220 patients with K. pneumoniae BSIs were identified, of which 72 were excluded (63 from the ICU, three had a polymicrobial BSI, one was an outpatient and five with incomplete data); the remaining 148 patients were enrolled in the study.

Table 1 shows that the median age of the study cohort was 68 years; ranging from 18–96 years; 66.2% were male, 99.3% had at least one comorbidity and 66.2% had undergone antibiotic therapy before blood sample collection. The median length of hospital stay was 22.0 (12.0–42.8) days. Drug resistance was detected in 41.2% K. pneumoniae isolates, 30.4% CRKP and 10.8% MDRKP. Drug-resistant isolates were more commonly recovered in death cases (53.2% vs. 35.6%). Two patients (1.4%) had undergone haemodialysis, 65.5% had received a blood transfusion and 60.1% had recent surgical treatment. In total, 47 patients (31.8%) had died, 16 (10.8%) had septic shock and 28.6% required admission to the ICU.

Table 1. Clinical characteristics of survival and death groups among patients with K. pneumoniae BSI

Annotation: a, antibiotic therapy before blood collection.

The median TTP for all patients was 11.0 (7.7–16.1 h and values ranged from 1.2–105.5 h), 11.5 (8.6–21.3 h) for the survival group and 9.3 (6.6–13.1 h) in the death group (Supplementary Fig. S1). Figure 1 shows that the ROC curve of TTP discriminated in-hospital mortality among patients with K. pneumoniae BSI, with an AUC of 0.61 (95% CI 0.52–0.71). A maximum Youden's index was reached when the TTP was 9.4 h, with a sensitivity of 70.3% and specificity of 51.1%. An early TTP (<9.4 h) was observed in 36.5% of the cohort patients who had higher in-hospital mortality rates than those with late TTP (44.4% vs. 24.5%). Among patients without antibiotic treatment before blood culture, early TTP was associated with in-hospital mortality according to χ 2 tests after correlation for continuity (P = 0.02). The ROC showed a predictive, but statistically nonsignificant value of TTP (AUC = 0.67, P = 0.49). By contrast, statistically significant differences in TBIL, PLT, WBC, transfusion, surgery, and in-hospital mortality rates were observed between early (<9.4 h) and late TTP (≥9.4 h) patients with K. pneumoniae BSI (Supplementary Table S1).

Fig. 1. ROC curve of TTP on in-hospital mortality among patients with K. pneumoniae BIS.

In the univariate logistic regression analysis of factors related to in-hospital mortality of the study patients, several factors were identified (Table 2) including: age > 60 years (OR = 3.6, 95% CI 1.4–9.4, P < 0.01), SCR > 106 μmol/l (OR = 4.6, 95% CI 2.1–10.0, P < 0.01), PCT > 2 ng/ml (OR = 2.5, 95% CI 1.2–5.2, P = 0.02), WBC < 4 × 109/l or >20 × 109/l (OR = 3.7, 95% CI 1.7–7.7, P < 0.01), CRP > 100 mg/l (OR = 3.5, 95% CI 1.6–7.8, P < 0.01), early TTP (OR = 2.5, 95% CI 1.2–5.0, P = 0.01), antibiotic therapy before blood collection (OR = 2.4, 95% CI 1.1–5.4, P = 0.03) and transfusion (OR = 3.6, 95% CI 1.5–8.5, P < 0.01). However, in the multivariate analysis, only age > 60 years (OR = 3.6, 95% CI 1.0–12.2, P = 0.04), SCR>106 μmol/l (OR = 6.1, 95% CI 2.0–18.4, P < 0.01), WBC<4 × 109/l or >20 × 109/l (OR = 4.4, 95% CI 1.5–12.4, P < 0.01), and CRP > 100 mg/l (OR = 4.0, 95% CI 1.3–12.0, P = 0.01) were confirmed as independent risk factors. Likewise, no significant influence of early TTP on mortality (OR = 2.7, 95% CI 1.0–7.4, P = 0.06) was observed.

Table 2. Factors related to in-hospital mortality among patients with K. pneumoniae BSI: univariate and multivariate logistic regression

Annotation: N = 135; Hosmer–Lemeshow test: P = 0.51; Overall prediction accuracy: 78.5%; a, antibiotic therapy before blood collection.

In the univariate logistic regression analysis of factors related to septic shock in 16 patients (Table 3), PCT > 2 ng/ml (OR = 4.6, 95% CI 1.2–17.0, P = 0.02), WBC < 4 × 109/l or >20 × 109/l (OR = 3.7, 95% CI 1.3–10.8, P = 0.02), MDRKP (OR = 5.3, 95% CI 1.2–22.4, P = 0.03), antibiotic therapy before blood collection (OR = 8.6, 95% CI 1.1–67.2, P = 0.04), and transfusion (OR = 8.6, 95% CI 1.1–67.2, P = 0.04) proved to be significant risk factors.

Table 3. Factors related to septic shock among patients with K. pneumoniae BSI: univariate logistic regression

Annotation: N = 137; a, antibiotic therapy before blood collection.

Several factors related to ICU admission were identified in the univariate analysis (Table 4) including: gender, elevations of MAP and hospital stays, CRKP, MDRKP, antibiotic therapy before blood collection, transfusion and surgery; by multivariate analysis, only surgery (OR = 3.6, 95% CI 1.2–11.1, P = 0.03) was linked to ICU admission. Female sex (OR = 0.2, 95% CI 0.1–0.6, P < 0.01) was an independent protective factor. No significant influence of increasing TTP on ICU admission was found in the univariate (OR = 1.0, 95% CI 1.0–1.0, P = 0.07) or multivariate analyses (OR = 1.0, 95% CI 1.0–1.0, P = 0.32).

Table 4. Factors related to ICU admission among patients with K. pneumoniae BSI: univariate logistic regression and multivariate logistic regression

Annotation: N = 147; Hosmer–Lemeshow test: P = 0.40; Overall prediction accuracy: 71.4%; a, antibiotic therapy before blood collection.

Discussion

In the present study, the median TTP was 11.0 h for patients with K. pneumoniae BSI, and there was a significant discriminative effect of its value on in-hospital mortality, with an optimal cut-off of 9.4 h according to the ROC curve. This result was similar to the median TTP of 10 h observed in a recent study on newborns [Reference Huggard15] and was lower than the median of 14.2 h and cut-off value of 13 h among paediatric patients [Reference Cheng13], but higher than median and cut-off values (9.6 and 7 h respectively) for K. pneumoniae BSI in adults in another study [Reference Liao12]. These outcomes can perhaps be partly explained by differences in physiological properties and bacterial load between children and adults. Unlike the latter two studies, antibiotic therapy was administered to over half of our patients in both the survival and mortality groups representing a much higher comparative number [Reference Liao12, Reference Cheng13], which may account for such differences [Reference Lambregts16]. Furthermore, it is inaccurate to divide subjects into short and long TTP (break point = 7 h) groups based only a higher OR without taking account of the sensitivity and specificity of the blood culture system [Reference Liao12].

In our series approximately one-third (36.5%) of patients exhibited an early TTP (<9.4 h); these patients had a 2.5 times higher risk of in-hospital mortality than those with late TTP values, according to univariate analysis. This observation was consistent with previous K. pneumoniae BSI studies [Reference Liao12, Reference Cheng13]. Among patients who had not received antibiotics prior to blood culture, early TTP was associated with in-hospital mortality despite a lack of significant discrimination in ROC (Supplementary Fig. S2). After controlling for several patient characteristics (Tables 1 and 2) patients with an early TTP had a 2.7 times higher risk of in-hospital death, but this did not prove statistically significant by multivariate analysis. Similar results were reported for BSI due to Enterococcus spp. [Reference Michelson, Löffler and Höring17]. Here, we observed a number of independent risk factors for in-hospital mortality which included age > 60 years, SCR > 106 μmol/l, abnormal WBC count, and CRP > 100 mg/l. SCR is a biomarker for renal function and high values are indicative of acute kidney injury and associated with mortality of hospitalised patients [Reference Laszczyńska18]. Likewise, the WBC count is an initial diagnostic marker of sepsis [Reference Fan19] and correlates to severe bacteraemia [Reference Paquette20]. Moreover, as found in this study, an elevated CRP value was reported to be linked to mortality in S. aureus bacteraemia [Reference Siméon10].

Neither early TTP nor increasing TTP proved to be significant factors in septic shock or ICU admission for patients with K. pneumoniae BSI. However, a meta-analysis of the literature noted that a short TTP was a general predictor of septic shock in patients with general bacterial BSI [Reference Hsieh21], but did not specifically include K. pneumoniae. Paediatric patients with early TTP were reported to be more likely to present with septic shock [Reference Cheng13], was not found here which might suggest that the prognosis of young patients is less affected by underlying conditions more commonly associated with the elderly.

This study has a number of limitations mainly due to its retrospective and observational nature. Consequently, diagnostic clinical score data, sources and sites of infection along with underlying conditions and comorbidities were not recorded. Likewise, the relatively low sample size of the cohort may have impacted on the validity of the multivariate analysis.

In conclusion, our findings indicate the relatively high in-hospital fatality rate of patients with K. pneumoniae BSI was affected by age and blood markers, notably SCR, CRP and WBC. Early TTP was statistically associated with in-hospital mortality of the patient cohort but a higher risk of fatality was not evident for septic shock or ICU admission among the early TTP group.

Supplementary material

The supplementary material for this article can be found at https://doi.org/10.1017/S0950268823000262.

Acknowledgements

The informed consent of enrolled patients was waived owing to the retrospective design of this study. We thank all enrolled patients and the support from Shanghai Tongji Hospital. All authors approved the final manuscript.

Financial support

This research received no specific grant from any funding agency, commercial or not-for-profit sectors.

Data availability statements

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to their containing information that could compromise the privacy of research participants.

Footnotes

*

These authors contributed equally to this work.

References

Calfee, DP (2017) Recent advances in the understanding and management of Klebsiella pneumoniae. F1000 Research 6, 1760.CrossRefGoogle ScholarPubMed
Baral, R et al. (2021) Low yield but high levels of multidrug resistance in urinary tract infections in a tertiary hospital, Nepal. Public Health in Action 11, 7076.CrossRefGoogle Scholar
Albu, S et al. (2018) Bacteriuria and asymptomatic infection in chronic patients with indwelling urinary catheter: the incidence of ESBL bacteria. Medicine (Baltimore) 97, e11796.CrossRefGoogle ScholarPubMed
Brady, M et al. (2016) Klebsiella pneumoniae bloodstream infection, antimicrobial resistance and consumption trends in Ireland: 2008 to 2013. European Journal of Clinical Microbiology and Infectious Diseases 35, 17771785.CrossRefGoogle ScholarPubMed
Xu, L, Sun, X and Ma, X (2017) Systematic review and meta-analysis of mortality of patients infected with carbapenem-resistant Klebsiella pneumoniae. Annals of Clinical Microbiology and Antimicrobials 16, 18.CrossRefGoogle ScholarPubMed
Hamilton, F et al. (2022) Time to positivity in bloodstream infection is not a prognostic marker for mortality: analysis of a prospective multicentre randomized control trial. Clinical Microbiology and Infection 28, 136.e7136.e13.CrossRefGoogle Scholar
Al-Juaid, A et al. (2012) Differential time to positivity: vascular catheter drawn cultures for the determination of catheter-related bloodstream infection. Scandinavian Journal of Infectious Diseases 44, 721725.CrossRefGoogle ScholarPubMed
Ning, Y et al. (2016) Time to positivity of blood culture and its prognostic value in bloodstream infection. European Journal of Clinical Microbiology and Infectious Diseases 35, 619624.CrossRefGoogle ScholarPubMed
Martín-Gutiérrez, G et al. (2017) Time to positivity of blood cultures in patients with bloodstream infections: a useful prognostic tool. Enfermedades Infecciosas Microbiologia Clinica 35, 638644.Google ScholarPubMed
Siméon, S et al. (2019) Time to blood culture positivity: an independent predictor of infective endocarditis and mortality in patients with Staphylococcus aureus bacteraemia. Clinical Microbiology and Infection 25, 481488.CrossRefGoogle ScholarPubMed
Xu, H et al. (2020) Prognostic role of time to positivity of blood culture in children with Pseudomonas aeruginosa bacteremia. BMC Infectious Diseases 20, 665.CrossRefGoogle ScholarPubMed
Liao, CH et al. (2009) Correlation between time to positivity of blood cultures with clinical presentation and outcomes in patients with Klebsiella pneumoniae bacteraemia: prospective cohort study. Clinical Microbiology and Infection 15, 11191125.CrossRefGoogle ScholarPubMed
Cheng, J et al. (2020) Time to positivity of Klebsiella pneumoniae in blood culture as prognostic indicator for pediatric bloodstream infections. European Journal of Pediatrics 179, 16891698.CrossRefGoogle ScholarPubMed
Shankar-Hari, M et al. (2016) Developing a new definition and assessing new clinical criteria for septic shock: Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). Journal of the American Medical Association 315, 775787.CrossRefGoogle ScholarPubMed
Huggard, D et al. (2021) Time to positivity (TTP) of neonatal blood cultures: a trend analysis over a decade from Ireland. Journal of Maternal and Fetal Neonatal Medicine 34, 780786.CrossRefGoogle Scholar
Lambregts, MMC et al. (2019) Time to positivity of blood cultures supports early re-evaluation of empiric broad-spectrum antimicrobial therapy. PLoS One 14, e0208819.CrossRefGoogle ScholarPubMed
Michelson, K, Löffler, B and Höring, S (2021) Time to positivity as a prognostic factor in bloodstream infections with Enterococcus spp. Diagnostic Microbiology and Infectious Diseases 101, 115396.CrossRefGoogle ScholarPubMed
Laszczyńska, O et al. (2020) Serum creatinine trajectories in real-world hospitalized patients: clinical context and short-term mortality. Journal of Investigative Medicine 68, 870881.CrossRefGoogle ScholarPubMed
Fan, SL et al. (2016) Diagnosing sepsis – the role of laboratory medicine. Clinica Chimica Acta 460, 203210.CrossRefGoogle ScholarPubMed
Paquette, K et al. (2021) Neither blood culture positivity nor time to positivity Is associated with mortality among patients presenting with severe manifestations of sepsis: the FABLED Cohort Study. Open Forum Infectious Diseases 8, ofab321.CrossRefGoogle ScholarPubMed
Hsieh, YC et al. (2022) Short time to positivity of blood culture predicts mortality and septic shock in bacteremic patients: a systematic review and meta-analysis. BMC Infectious Diseases 22, 142.CrossRefGoogle ScholarPubMed
Figure 0

Table 1. Clinical characteristics of survival and death groups among patients with K. pneumoniae BSI

Figure 1

Fig. 1. ROC curve of TTP on in-hospital mortality among patients with K. pneumoniae BIS.

Figure 2

Table 2. Factors related to in-hospital mortality among patients with K. pneumoniae BSI: univariate and multivariate logistic regression

Figure 3

Table 3. Factors related to septic shock among patients with K. pneumoniae BSI: univariate logistic regression

Figure 4

Table 4. Factors related to ICU admission among patients with K. pneumoniae BSI: univariate logistic regression and multivariate logistic regression

Supplementary material: File

Hou et al. supplementary material

Figures S1-S2 and Table S1

Download Hou et al. supplementary material(File)
File 274.2 KB