Extended-spectrum β-lactamase-producing Escherichia coli and Klebsiella pneumoniae in Guangzhou, China: risk factors and economic burden among Patients with bloodstream infections

doi:10.21203/rs.3.rs-3432080/v2

Download PDF

Research Article

Extended-spectrum β-lactamase-producing Escherichia coli and Klebsiella pneumoniae in Guangzhou, China: risk factors and economic burden among Patients with bloodstream infections

https://doi.org/10.21203/rs.3.rs-3432080/v2

This work is licensed under a CC BY 4.0 License

Version 2

posted

You are reading this latest preprint version

Background.

Extended-spectrum β-lactamase-producing Escherichia coli and Klebsiella pneumoniae(ESBL-EK) are important sources of bloodstream infection (BSI), but their economic impact has not been comprehensively evaluated.

Patients and methods. A retrospective analysis of inpatients diagnosed with Escherichia coli or Klebsiella pneumoniae bacteremia in tertiary hospitals between January 2020 and December 2022 was performed. Clinical data and healthcare spending were collected through a chart review of electronic medical records. The chi-square test was used to analyze risk factors for ESBL, and then propensity score matching (PSM) was used to match patients 1:1 ratio to eliminate confounding factors and economic burden was assessed using disability-adjusted life years (DALYs).

Results. A total of 318 patients meeting study specifications were identified. Before matching, there were statistically significant differences in hospitalization time before infection, organ transplantation rate, ICU admission rate, Age-adjusted Charlson Comorbidity Index (ACCI), and SOFA score between the two groups (all p<0.05). After using PSM-adjusted estimates, the ESBL group still had a lower rate of effective empirical antimicrobial therapy than the non-ESBL group (non-ESBL=79.3% vs ESBL=59.5% p=0.001). The ESBL group had a higher hospitalization cost, but there was no significant difference in total mortality (non-ESBL=19.0% vs ESBL=20.7%, p=0.747) and mean DALYs (non-ESBL=1.84 DALYs vs ESBL=2.12 DALYs, p=0.098) between two groups. Mediation analysis identified that the relationship between ESBL and hospitalization costs is largely or fully mediated by inappropriate empirical antibiotic therapy and hospital stay.

Conclusion. The cost of patients with BSI caused by ESBL-EK is higher than that of patients with BSI caused by non-ESBL-EK. This phenomenon can be attributed to differences in effective empirical antimicrobial therapy rates and length of hospitalization. Reasonable assessment of ESBL risk，fully consider the regional epidemiological situation and precise use of antibiotics are very important to reduce the burden on patients.

Escherichia coli

Klebsiella pneumoniae

bacteremia

ESBL

medical costs

Evidence before this study

We searched PubMed using the keywords “bacteremia”, “disease burden”, “medical costs”, “ESBL”, “antimicrobial resistance”, “EC” and “KP”, from database inception to September 2022 January 11th. Most existing studies show that bacteremia caused by ESBL generally has a higher disease burden compared with bacteremia caused by non-ESBL-producing bacteria. WHO has included extended-spectrum beta-lactamase (ESBL)-producing Enterobacteriaceae in the WHO list of priority pathogens for new antibiotic research and development and classified them as priority category 1: extremely important part. However, in recent bacteremia prevalence data, there are large gaps in estimates of the additional burden caused by ESBL bacteria in different regions and countries. These data gaps mean that large-scale epidemiological surveys globally or across China may not reflect local conditions. Regional epidemiological survey data are crucial for local hospitals’ decision-making on antibiotic selection and nosocomial infection management.

Added value of this study

This report complements the evidence on the disease burden of bacteremia in South China. We use disability-adjusted life years (DALYs) to evaluate the economic burden and uses propensity score matching (PSM) and mediation analysis to eliminate confounding factors and identify sources of burden.

Implications of all the available evidence

Bloodstream infection caused by ESBL-producing EC/KP in our hospital will reduce the effectiveness of empirical medication by 19.8%, extend the total length of hospitalization by an average of 3 days, and increase the patient's financial burden by US$2,047. This study suggests that reasonable assessment of ESBL risks and precise use of antibacterial drugs based on regional epidemiological conditions are important to reduce patient burden. Finally, we believe that continued epidemiological testing and risk assessment of antimicrobial resistance are necessary for clinical and antimicrobial stewardship.

Bloodstream infection (BSI) is one of the most common severe bacterial infections in clinical practice, which has caused a serious disease burden[1][2][3]. A European study estimated that more than 160,000 healthcare-associated (HA) primary BSI in the European Union and the European Economic Area (EU/EEA) are infected each year, with more than 24,000 deaths, and its disease burden is second only to hospital-acquired pneumonia (HAP) among HA infections, becoming a serious public health challenge[4][5]. Globally, 1,500,000 number of deaths were associated with antimicrobial resistance (AMR) among patients with bloodstream infections in 2019, which further hamper treatment options[2].

Among most-considerable BSI causative pathogens, the number of individuals infected by Klebsiella pneumoniae (KP) and Escherichia coli (EC) is amongst the most prevalent and has been increasing globally in latest decade [3][5]. In China, EC and KP are reported isolated in 35% of blood samples, with elevated presence of resistance mechanisms, including extended-spectrum β-lactamase (ESBL) carrying rates (exceeding 40% among isolates) [6]. A previous epidemiological survey in Guangdong Province showed that (ESBL)-producing EC and KP isolates accounted for more than 40% in children, which presumes even greater updated rates [7]. Genes encoding ESBL are usually located on mobile genetic elements (MGE) such as plasmids and are widely transferred among EC and KP [8][9]. Not only they endow the host with resistance to β-lactamases such as cephalosporins, but these plasmids usually carry multiple drug-resistance genes [10][11], which increases the difficulty of clinical anti-infective drug selection and may increase the cost of patient treatment[12].

Literature has suggested differing attempts to quantify the economic and burden impact of ESBL on EC/KP-induced bacteremia displaying inconsistent conclusions. A systematic review pointed out that Study design which controls for severity of illness and same empirical antibiotic therapy in the two groups would be expected to bias the study towards a similar, even negative result[13]. In addition, Recent figures from the World Health Organization (WHO) Global Antimicrobial Resistance and Surveillance System (GLASS) report show that the proportion of Escherichia coli bloodstream infections (BSIs) caused by third-generation cephalosporins resistant E. coli was more than triple in LMICs compared to high-income countries, (58.3 % and 17.53%, respectively)[14] Previous meta-analysis showed that patients with antibiotic-resistant Gram-negative BSI in China showed higher excess mortality and ICU admissions compared with other low- and middle-income countries with reported data rate and length of hospital stay (LOS) [1].

These differences in epidemiology and clinical impact between regions mean that large-scale epidemiological studies around the world or across China may not reflect local conditions, and regional epidemiological survey data are crucial for local hospitals to make decisions about antibiotic selection and nosocomial infection control. is crucial [15].

In this study, we explored the epidemiological status of local carriers of ESBL bacteria through a single-center retrospective study, on the basis of previous studies, a mediation analysis was introduced. We use propensity score matching method and comprehensively evaluated the impact of ESBL-caused-infections on the disease burden of patients by counting both direct and indirect economic burden of hospitalization and mortality[16].

Patients and sample selection

Inpatients diagnosed with EC or KP bacteremia at a 2850-bed teaching hospital (Guangzhou, China）from January 2020 to December 2022 were included. The EC and KP specimen information from blood specimen sources were retrieved from the laboratory electronic system. Inclusion criteria for every patient were as follows: (1) had positive blood cultures of EC or KP; (2) showed clinical manifestations of BSI; (3) had complete clinical microbiological and cost data for analysis. For patients with multiple episodes of EC BSI during hospital readmission or within 6 months of discharge, data only from the first episode of bacteremia were included. We excluded patients if they (1) were younger than 18 years (2) their medical records were incomplete (3) were discharged, died or received palliative care on the day of blood collection [17]. Patients infected with ESBL-positive EC or KP were classified as ESBL group, and patients infected with ESBL-negative EC or KP were classified as non-ESBL group.

Microbiological Tests

The clinical isolates were identified using the matrix-assisted laser desorption ionization time-of-flight microbial identification system (MALDI-TOF) (biomacimrieux, France), and the VITEK-2 COMPACT automatic microbial identification system (biomacimrieux, France) was used to perform drug susceptibility testing of the strains. The ESBL production of Escherichia coli and Klebsiella pneumoniae in this study was determined by Gram Negative Susceptibility Card VITEK 2 AST-334 (biomacimrieux, France), This drug susceptibility card implements CLSI M100 file [18].

Definitions

Hospital-acquired bacteremia was defined as those BSIs showing a positive blood cultures 48 hours after admission, whereas community-acquired referred to those positive blood cultures on admission or within 48 hours of admission. BSI’s severity was assessed utilizing the APACHEII score[19]; the SOFA score was used to describe organ dysfunction/failure in patients with sepsis[20]; the overall systemic health status of patients was determined by the aCCI [21]. Initial antibiotic therapy was defined as empiric therapy until susceptibility results were available. Initial antimicrobial therapy was considered effective if the results of in vitro susceptibility testing showed that the isolated pathogen was sensitive to at least one antibiotic in empirical administration in terms of minimum inhibitory concentration[22], and it was defined as inappropriate if no effective in vitro active antimicrobial was delivered within 48 hours of diagnosis. For infections caused by ESBL-EK (extended-spectrum β-lactamase-producing Escherichia coli and Klebsiella pneumoniae), oxyimino-β-lactams (e.g., cefuroxime, cefotaxime, ceftriaxone, ceftazidime, cefepime, and aztreonam) were defined as inappropriate treatment, regardless of in vitro susceptibility [23]. The clinical outcome of this study was in-hospital and 28-day post-discharge mortality in the ESBL and non-ESBL groups calculated as the total number of deaths/total number of cases.

Clinical and sociodemographic characteristics

We collated all patient electronic medical record information. Clinical and laboratory data included patient demographics (sex, age), comorbidities at admission and discharge times, age-adjusted Charlson Comorbidity Index (aCCI), date of BSI onset, organ transplantation during hospitalization, severity of disease (APACHE II score and SOFA score), microbiological data, information on empiric antimicrobial use and its sensitivity, records of transfers in and out of the ICU and discharge records containing a summary of the patient's condition at discharge, and outpatient follow-up records. The study complied with the Declaration of Helsinki; it was approved by the Hospital Ethics Committee (approval number: ES-2023-057-01).

Cost analysis

Treatment costs were retrieved from the hospital information system. Direct and indirect costs (i.e., lost productivity due to hospitalization and death) were analyzed. All costs were converted to USD (on average from 2020-2022, 1 USD = 6.69 RMB). The total direct expenses included medical care (treatment and nursing in the ward), treatment (surgical anesthesia, laboratory tests), rehabilitation, drugs (including antibacterial drugs), traditional Chinese medicine, diagnosis (pathology, microbiology, laboratory, imaging), disposable consumables, blood transfusion, and other expenses (other items specified by the Healthcare Security Administration of Guangdong Province).

The indirect economic burden of EC bacteremia was analyzed through DALYs and human capital approaches. These methods impose an indirect economic burden equal to DALYs times gross-domestic product (GDP) per capita times productivity weights. In 2020, China's GDP per capita was $10737, in 2021 it was $12162, and in 2022 it was $12809 [24]. Different age groups have different productivity weights, children aged 0-14 were assigned a weight of 0.15; 15-44 and 45-59 years old were assigned a weight of 0.75 and 0.80, respectively; over 60 years old were assigned a weight of 0.1 [25].

BSI-associated DALYs were calculated by summing all years of life lost (YLLs) and years living with disabilities (YLDs). Other characteristics, such as life expectancy, age (standardized), time discount rate, and disability weights were included in the calculation of YLLs and YLDs[16]. We utilized the following formula for calculating YLL, previously recommended by Murray and Lopez et al.[26].

The difference with the formula for YLLs is that D (the disability weight) was added in the formula for YLDs [γ,K,β], as follows:

We used a calculator designed by the Brecht Devleesschauwer team based on the R language to calculate economic costs. The weight parameters used the values recommended by the World Health Organization, and the constant value was 0.166 [28][29][30]. According to the global burden of disease (GBD) template provided by the WHO and the Institute for IHME, the value of disability weight (D) ranges from 0 to 1. Since the recent GBD study did not mention the criteria of BSI, our assessment of D was based on acute infection, which was divided into 0.006, 0.051 and 0.133 according to the match severity of mild, moderate, and severe disease to 0.006, 0.051 and 0.133. To calculate L (years lost due to death and years discounted due to disability), we used "Standard Expected Years of Life Lost" (SEYLL) as a good approximation for life expectancy [31].

Statistical analyses

Our study determined the causal relationship between different clinical factors and whether covariates were confounding factors by checking chronological orders. We did not consider the inoculum effect of antimicrobial agents on ESBL bacteria and assumed that the phenotype of bacteria in a single patient was fixed during treatment. The variables were divided into three categories according to "whether the variable was preceded by bacteremia in time sequence" and "whether the variable was affected by the ESBL gene": ①Variables preceding the onset of bacteremia (such as hospitalization time before bacteremia or the aCCI) are performed by chi-square (χ²) test to confirm the risk factors of ESBL infection, and at the same time, the difference between groups was excluded when calculating the economic cost. ② Variables following the onset of bacteremia, but its impact on clinical outcome of patient was independent of ESBL (such as SOFA score). These variables not included in the risk factor analysis of ESBL in the study, but they were used as confounding factors in the study of the impact of ESBL on economic burden to exclude differences between groups through PSM.③ Factors influenced by ESBL, like initial antimicrobial susceptibility, are not adjusted for economic assessments using PSM, and their mediating impact has been verified [32].

For descriptive purposes, we computed mean, standard deviation or median (interquartile range [IQR]) for continuous variables and percentage (%) for categorical variables. For statistical analysis, Student’s t-test (for variables with normal distribution) or Mann–Whitney U test (for variables with non-normal distribution) was used, and chi-square (χ2) test was used for categorical variables. We computed a PSM-corrected logistic regression using a backward stepwise method. We matched individuals from the ESBL and non-ESBL groups utilizing a caliper with a match tolerance set to 0.02. Mediating effects were confirmed by Baron and Kenny's steps for mediation analysis between ESBL and hospitalization costs via the mediating impact of LOS and ineffective empirical antimicrobial therapy[33][34]. We employed a holistic and comprehensive approach considering LOS and IEAT as simultaneous mediators of the association between ESBL infections and hospitalization costs. We utilized structural equation modeling (SEM) through the “lavaan” package (in the R software) to account for the correlations and interactions between mediators and ESBL and economic costs using linear regressions. Bootstrapped standard errors with 1,000 replications. P-value of less than 0.05 was considered statistically significant. Statistical analyses were performed in R version 4.3.1 and SPSS version 26.0 (SPSS, Chicago, IL, USA), including the R-based DALYs calculator [28].

Analytical sample

From January 2020 to December 2022, we identified a total of 372 isolates of Escherichia coli and Klebsiella pneumoniae from the blood of 367 patients, and excluded 5 duplicate isolates from the same patient, and then excluded 49 electronic medical records according to the exclusion criteria described in the methodology section. Finally, the analysis included 318 nonrepeating E. coli clinical and K. pneumoniae isolates isolated from the blood of 318 patients. These 318 patients included 152 non-ESBL-EC/KP BSI patients (47.8%) and 166 (52.2%) ESBL-EC/KP BSI patients.

Descriptive statistics before matching and by ESBL group

Of the 318 patients, 179 were male, and the mean ages for ESBL-BSI and non-ESBL-BSI were 56.5 and 63 years, respectively (p=0.064). There was no statistically significant difference in strain distribution between the two groups in this study (p=0.407). The ESBL detection rates of EC and KP were 54.0% and 49.1%, respectively. The results of EC were consistent with the results of China Antimicrobial Surveillance Network (CHINET) [35], but the ESBL positive rate of KP in adults was higher than our previous study of the provincial epidemiologic survey. (49.1% vs 28.7%) [7]. The organ transplant rate (non-ESBL=1.3% vs ESBL=6%, p=0.028) and ICU admission rate (non-ESBL=7.2% vs ESBL=19.9%, p=0.001) were higher in the ESBL group before onset than in the non-ESBL group, but nosocomial infection rates were not significantly different between the two groups (non-ESBL=40.1% vs ESBL=35.5%, p=0.399).

The most common comorbidities in ESBL group and non-ESBL group were malignant tumor (non-ESBL=36.2% vs ESBL=26.5%), hepatobiliary disease (non-ESBL=14.5% vs ESBL=22.9%) and diabetes (non-ESBL=14.5% vs ESBL=16.9%). Among comorbidities, only moderate to severe nephropathy had a statistically significant difference between groups (6.6% for non-ESBL vs 15.7% for ESBL, p=0.011). Notably, while there were partial differences in comorbidity distribution between the two groups of patients, there was no statistically significant differences in quantified aCCI (non-ESBL=4(2,6) vs ESBL=4(3,6), p=0.232). Correspondingly, the APACHEII (non-ESBL=10(7,13) vs ESBL=12(7,19), p=0.001) score and SOFA score (non-ESBL=1(0,4) vs ESBL=3(0,7), p=0.004) reflecting the patient's basic condition and the severity of sepsis in the ESBL group were higher than those in the non-ESBL group.

Although more carbapenems were used in the ESBL group (non-ESBL=41.4% vs ESBL=53%, p=0.039), the effective rate of empiric antibacterial therapy was lower than that of the non-ESBL group (non-ESBL=80.3% vs ESBL=58.4%, p<0.001), and the total mortality rate (non-ESBL=17.8% vs ESBL=27.1%, p=0.047) was higher than that of the non-ESBL group.

Estimation of the propensity score

To further confirm the impact of ESBL on the clinical outcome and economic burden of patients, we performed PSM matching on the two groups of patients. After PSM matching processing, in the newly generated data set, there were 121 patients in the ESBL-producing group and 121 patients in the non-ESBL-producing group, and the sample size was 76.1% of the original sample size. After matching, there was no significant difference in clinical characteristics considered as confounding factors between the two groups (all P>0.05), including gender, age, bacterial species, hospital infection rate, hospitalization time before infection, organ transplantation during hospitalization, empirical drug category, APACHE II score, SOFA score, and aCCI. The raw and PSM-adjusted ESBL+ and non-ESBL proportions are depicted in Supplementary Figure A1. We found similar distributions across ESBL proportions after correcting the estimates utilizing the PSM.

Association between ESBL+ and outcome variables related to clinical burden

We noticed that after removing the interference of confounding factors, the ESBL group still had a lower rate of effective empirical antimicrobial therapy than the non-ESBL group (non-ESBL=79.3% vs ESBL=59.5%, p=0.001), longer total hospitalization time (ESBL=18(11, 28.5) vs non-ESBL=14(10, 22.5), p=0.02) and longer post-infection hospitalization time (non-ESBL=9(6, 13) vs ESBL=12(7.5, 19.5), p<0.001), but there was no significant difference in the overall mortality rate between non-ESBL group and ESBL group (non-ESBL=19.0% vs ESBL=20.7%, p=0.747).

Table 1：Characteristics of Patients with Bloodstream Infection (BSI) Stratified by Extended-Spectrum-Lactamase (ESBL) Production

Category	Comparison before PSM			Comparison after PSM
Parameter/Category	Non-ESBL（n=152）	ESBL（n=166）	p	Non-ESBL（n=121）	ESBL（n=121）	p
Male sex，n（%）	75(49.3%)	104(62.7%)	0.017	58,(47.9%)	66(54.5)	0.304
EC,n（%）	93(61.2%)	109(65.7%)	0.407	90(74.4%)	88(72.7%)	0.771
KP,n（%）	59(38.8%)	57(34.3%)	0.407	31(25.6%)	33(27.3%)	0.771
Age in years, [M(Q1, Q3)]	56.5(46,68.8)	63（50，71）	0.064	57(48,69)	60(49,71)	0.523
Nosocomial bacteremia, n（%）	61(40.1%)	59(35.5%)	0.399	51(42.1%)	48(39.7%)	0.695
organ transplantation，n （%）	2(1.3%)	10(6%)	0.028	1(0.8%)	5(4.1%)	0.098
LOS hospital length of stay (LOS)
Total LOS（M，IQR）	15(10,26.75)	21(11,35)	0.003	14(10,22.5)	18(11,28.5)	0.02
LOS before the bacteremia, [d，M(Q1, Q3)]	4(1,12)	5(1,15)	<0.001	4(1,12)	4(1,10)	0.348
LOS after the bacteremia, [d，M(Q1, Q3)]	9(6,14)	13(7,21)	0.007	9(6,13)	12(7.5,19.5)	<0.001
ICU admission before the bacteremia，n （%）	11(7.2%)	33(19.9%)	0.001	11(9.1%)	7(5.8%)	0.327
APACHE score，[M(Q1, Q3)]	10(7,13)	12(7,19.25）	0.001	10(7,13)	10(7,15)	0.223
sofa score，[M(Q1, Q3)]	1(0,4)	3(0,7)	0.004	2(0,4)	2(0,4.5)	0.399
Comorbid illnesses
Malignant tumor，n（%）	55（36.2%）	44（26.5%）	0.063	39	25	0.041
Hepatobiliary disease，n （%）	22（14.5%）	38（22.9%）	0.055	20	32	0.06
Leukemia n （%）	18（11.8%）	13（7.8%）	0.228	15	11	0.406
Lymphoma，n （%）	10（6.6%）	8（4.8%）	0.498	6	6	1
Kidney disease，n（%）	10（6.6%）	26（15.7%）	0.011	10	21	0.034
Diabetes，n（%）	22（14.5%）	28（16.9%）	0.558	18	22	0.489
COPD,n （%）	11（7.2%）	16（9.6%）	0.331	7	11	0.327
Cardio-Cerebrovascular Disease，n （%）	21（13.8%）	27（16.2%）	0.542	16	15	0.94
Peptic ulcer，n （%）	9（5.9%）	7（4.2%）	0.487	6	6	1
rheumatism disease ,n （%）	13（8.6%）	15(9%)	0.879	11	10	0.819
Other ,n （%）	1（0.7%）	5(3%)	0.123	1	3	0.313
aCCI [M(Q1, Q3)]	4(2,6)	4(3,6)	0.232	4(2,6)	4(2,6)	0.419
Empirical antimicrobial
Cephalosporins，n （%）	22(14.5%)	25(15.1%)	0.883	17	23	0.299
BLBLI， n （%）	70(46.1%)	76(45.8%)	0.962	54	50	0.603
Carbapenems，n （%）	63(41.4%)	88(53%)	0.039	46	57	0.153
Aminoglycosides，n（%）	8(5.3%)	10(6%)	0.769	6	5	0.758
Fluoroquinolone， n （%）	22(14.5%)	18(10.8%)	0.329	14	14	1
Glycopeptides &Tigecycline，n（%)	25+1(17.1%)	27+2(17.5%)	0.932	16+1	13+1	0.564
Antifungal drug，n（%)	26(17.1)	30(18.1%)	0.821	17	14	0.564
Other,n （%）	18(11.8%)	19(11.4%)	0.924	14	7	0.11
Effective empirical antimicrobial therapy，n （%）	122(80.3%)	97(58.4%)	<0.001	96(79.3%)	72(59.5%)	0.001
Mortality
Total mortality，n （%）	27(17.8%)	45(27.1%)	0.047	23(19.0%）	25(20.7%)	0.747
28-day mortality，n （%）	15(9.9%)	21(12.6%)	0.434	12(9.9%)	12(9.9%)	1
In-hospital mortality，n （%）	12(7.9%)	24(14.5%)	0.065	11(9.1%)	13(10.7%)	0.667

Notes: ICU= Intensive care unit. aCCI= Age-adjusted Charlson comorbidity test. COPD= Chronic obstructive pulmonary disease. Q= quartile. LOS= Length of hospital stay. BSI= Bloodstream infection. BLBLI= β-lactam/β-lactamase inhibitor. IQR= Interquartile range. P= p-value for the t-test or χ² tests. PSM= Propensity score matching.

Table 2：Costs of Patients Bloodstream Infection (BSI) Stratified by Extended-Spectrum-Lactamase (ESBL) Production

Costs (Median, IQR, $)	Comparison before PSM			Comparison after PSM
	Non-ESBL（n=152）	ESBL（n=166）	p	Non-ESBL （n=121）	ESBL （n=121）	p
Direct economic
Total direct economic burden	6176(3061,14137)	8954(35927,23169)	<0.001	5638(3007,13346)	7685(4817,13960)	0.014
General medical services（Nursing care)	652(315,1442)	1066(503,2562)	<0.001	530(304,1226)	772(423,1646)	0.013
Diagnosis & Laboratory tests	1450(887,2557)	1956(1166,4262)	<0.001	1326(896,2516)	1628(1044,2884)	0.051
Treatment& Surgery	309(63,1386)	1189(171,3120)	<0.001	342(57,1629)	1075(107,2397)	0.022
Rehabilitation	0(0,0)	0(0,1)	0.031	0(0,0)	0(0,0)	0.134
Traditional Chinese medicines	0(0,0)	0(0,0)	0.035	0(0,0)	0(0,0)	0.154
Medicine	1573(821,4324)	2860(7808,1167)	0.002	1344(769,4100)	2472(986,4914)	0.021
Antimicrobial	471(156,1148)	1042(336,3387)	<0.001	400(148,1001)	834(316,2506)	<0.001
Proportion of antimicrobials [M(Q1, Q3)]	28.0%(16.6%,45.3%)	39.3%(27.8%,56.5%)	<0.001	27.3%(15.4%,48.1%)	39.0%(27.6%,55.8%)	0.002
Blood transfusion	0(0,448)	123(0,752)	0.036	0(0,371)	0(0,401)	0.599
Medical consumables	654(259,1260)	1112(446,2602)	<0.001	583(257,1338)	974(394,1779)	0.022
Others	183(61,456)	257(97,830)	0.003	167(56,352)	197(68,491)	0.341
Indirect economic burden
DALYs	1.78	2.46	0.003	1.84	2.12	0.098
Indirect economic burden	11791.2	17454.7	0.444	13143.2	15820.7	0.702

Notes: Q= quartile. IQR= Interquartile range. DALYs= Disability-adjusted life years. P= p-value for the t-test or χ² tests. PSM= Propensity score matching.

Economic costs

After PSM, the median total hospitalization cost was $5638 for ESBL patients and $7685 for non-ESBL patients (p=0.014) (Figure 1). In the two groups, the cost of antibiotics accounted for 27.3% and 39.0% of the drug, respectively (p=0.002). The median cost of antibiotics during hospitalization was 400 for ESBL-EC BSI patients and 834 for non-ESBL-EC BSI patients, p<0.001).

In the non-ESBL-EC group after PSM matching, patients lost an average of 1.84 DALYs. In the ESBL-EC group, patients lost an average of 2.12 DALYs (p=0.098). There was no significant difference in mean indirect loss in the ESBL-EC group compared to the non-ESBL-EC group. After giving different productivity weights according to patient age groups, there was no significant difference in the indirect economic burden, regardless of whether confounding factors were excluded.

Mediation analyses

To identify mediating variables, Baron and Kenny's steps for mediation analysis were performed for the mediating impact of LOS and inappropriate empirical antibiotic therapy (IEAT) on hospitalization costs. ESBL was positively associated with IEAT (Estimate = 0.215, SE=, p<0.05) and LOS (Estimate = 0.151, SE=0.05, p<0.05) (Table 3). However, ESBL did not show a direct impact on economic costs (-0.03, SE=, p>0.05), whereby IEAT and LOS were significantly and directly associated with economic costs (Estimate= 0.13, SE=0.06, p-value=0.037, Estimate= 0.62, SE=0.14, p-value<0.001, respectively). All estimated variances for the residuals (errors) of the three models (a, b, and c) are significant, indicating variability around the predicted values of these variables (Table 3, variance). The total indirect impact of ESBL on cost was significant (Estimate= 0.12, SE=0.06, p-value<0.027), with marginal indirect impacts from IEAT and more significant impacts via LOS (Table 3, impacts of indirect effects). The total indirect effect impact was significant suggesting that the relationship between ESBL and COST is largely or fully mediated by IEAT and LOS.

Table 3. Mediation effect analysis results

	Regressions model	Association type		Estimate		SE	p-value
	(a)	IEAT ← ESBL		0.22		0.06	0.001
	(b)	LOS ← ESBL		0.15		0.06	0.019
	(c)	Economic costs ←
		ESBL		-0.03		0.05	0.591
		IEAT		0.13		0.06	0.037
		LOS		0.62		0.14	0.000
	Estimated variance	Model outcome		Estimate		SE	p-value
	(a)	IEAT		0.95		0.05	0.000
	(b)	LOS		0.97		0.24	0.000
	(c)	Economic costs		0.58		0.22	0.007
	Impact of indirect and direct effects on costs			Estimate		SE	p-value
Indirect effect of ESBL via IEAT			0.03		0.02			0.077
Indirect effect of ESBL via LOS			0.09		0.05			0.048
Total indirect effect of ESBL (via IEAT and LOS)			0.12		0.06			0.027
Total effect (direct and indirect)			0.09		0.06			0.151

Notes: LOS= Length of hospital stay. IEAT= Inappropriate empirical antibiotic therapy. ESBL= Extended Spectrum Beta-Lactamase. SE= Standard error. Figure A2 presents the direct and indirect impacts of ESBL on economic costs, using IEAT and LOS as mediators.

Our results showed that comparing to BSI patients in the non-ESBL group, the ESBL group had greater risks of empirical medication failure and mortality, prolonged hospital stay, and incurred more medical costs. After excluding confounding factors, there were still statistically significant differences in effective empirical antimicrobial therapy rate, length of hospitalization and hospitalization costs between the two groups.

We believe that the exclusion of confounding factors and the inclusion of more comprehensive economic indicators in retrospective studies through chronological order and causality can more accurately reflect the impact of ESBL-producing pathogens on the economic burden of patients. When calculating the economic burden, we used hospital expenses, which are more comprehensive than patient bills, to reflect the impact of ESBL-producing pathogens on the entire medical system, and included DALYs to calculate the overall burden on patients’ families and society brought about by the lack of labor due to diseases [36].

Similar to most of the studies in the article counted by Zhen X et al. in a recent systematic review [13], patients in the ESBL group in our study had longer hospital stays and higher financial burden. We found a 33.3% increase in median post-bacteremia LOS (from 9 days to 12 days), a 28.6% increase in total LOS (from 14 days to 18 days), and a 36.3% increase in hospital costs (an additional $2047 per patient) in the post-PSM ESBL group.

At present, how ESBL affects the burden of disease is still unclear. Most scholars agree that ESBL-induced prolonged hospitalization is an important reason for the increase in hospitalization costs. However, the contribution of the initial treatment effectiveness and the direct cost of antibiotics to the overall direct burden is still controversial. We believe that this can be related to the comparison of medical services and drug prices in different regions[13]. The results of the mediation effect test in this study we found that ESBL had statistically significant effects on IEAT and LOS, and these variables, in turn, have significant effects on hospitalization costs, suggesting that interventions aimed at modifying ESBL might indirectly influence costs through changes in IEAT and LOS. We suggest that addressing and studying the mediators can give more insight into the mechanism through which ESBL influences hospitalization costs. In our study, the median difference in antimicrobial spending accounted for 21.2% (434/2037) of the median difference in total hospitalization spending, and the higher cost of antimicrobials in the ESBL group and its share in Western medicine (27.3% vs 39%, p = 0.002) suggested that the higher cost of drugs required for ESBL-BSI itself also increased hospitalization costs.

In this study, we observed that in the composition of the disease burden, in the average disability years (1.84-2.12 years), the average hospitalization days of 14-18 days produced less than 1% of the YLDs of the DALYs after the weight of the disease was assigned. We think that this is normal due to the short Disease-process of bloodstream infection and the high mortality rate. We suggest that for investigators with no access to complete data, the use of YLDs alone in the absence of bacteremia-associated sequelae is sufficient to approximately reflect the burden of disease in adults with acute bloodstream infections.

In this study, we observed that clinicians' decisions to use carbapenems were not random（non-ESBL=45.4% vs ESBL=53%，p=0.039）. However, after excluding the interference of SOFA score and APACHE II score, there was no significant difference in the drug usage rate between the two groups（p=0.153）. We speculate that this may be due to the fact that doctors used the drug according to the severity of clinical symptoms rather than risk factors caused by ESBL before obtaining the drug sensitivity report. When inquiring medical records, we found that most of the electronic admission records did not record the antibacterial drug use information in the six months before admission. This is concerning, since some reports suggest that the use of antibacterial drugs other than β-lactams is also a risk factor for ESBL[17]. Therefore, we believe that it is necessary to trace and record all antimicrobial drugs in detail during admission consultation for more refined antimicrobial stewardship (AMS).

In previous studies, nosocomial infection was usually a high-risk factor for ESBL-producing bacterial infection [37]. However, in this study, although the LOS of the ESBL group before PSM was longer, there was no significant difference in the nosocomial infection rate between the two groups (non-ESBL=40.1% vs ESBL= 35.5%, p=0.399), suggesting that blaESBL in this area has been widely spread in the community. Combining the relatively high ESBL detection rate in the local area and, clinicians should carefully choose to use third-generation cephalosporins before obtaining definite drug susceptibility results [7][35]. There are also studies in China showing that choosing antimicrobial agents covering ESBL-producing bacteria as empirical antibiotic therapy can improve the cure rate, reduce the direct and indirect costs for individual and society, clinics need to balance individual and social needs [31]. We believe that the ESBL detection tricycle plan and AWaRe tool recommended by WHO should be better implemented, and a multi-dimensional molecular epidemiological early warning and antimicrobial stewardship system should be established to face the comprehensive and multidisciplinary health challenge of antimicrobial resistance [38][39].

Although we ruled out the confounding interference of different indicators, there may still be potential selection biases. Since the Charlson score assigned different weights to the disease, in order to maximize the enrollment rate of patients in the PSM, we did not include a single comorbidity in the pairing criteria in the PSM. Previous literature pointed out that ESBL plasmids may be related to some virulence factors[40], but this study only considered the interference of drug resistance genes on patients, and potential interference factors such as virulence factors were eliminated by incorporating indicators such as SOFA score into PSM matching.

In summary, we support the notion that bloodstream infection from ESBL-producing EC/KP led to reduced efficacy of empiric medication, prolonged hospital stays, and increased patient financial burden. This study suggests that a reasonable assessment of ESBL risk and precise use of antibiotics are very important in reducing the burden on patients. Finally, we believe that continuous epidemiological detection and risk assessment of antimicrobial resistance is necessary for clinical and antimicrobial stewardship.

Funding: This research was funded by Special Foundation for National Science and Technology Basic Research Program of China 2019FY101200。.

Role of the funding sources: The funders had no role in the design, conduct, analysis or interpretation of the study.

Acknowledgments: All authors attest they meet the ICMJE criteria for authorship and have reviewed and approved the final article. We acknowledge Prof Lixiong Yang, School of Labor and Human Resources, Renmin University of China and Xiaoyi Fang, Institute of Population research Peking University for their invaluable help and expertise.

Conflicts of Interest: The authors declare no conflict of interests.

Author Contributions: Conceptualization (Chao Zhuo); methodology (Wenwei Luo；Chuyue Zhuo; Nanhao He); formal analysis (Yingyi Guo; Jiong Wang；Likang Yao; Jiahui Li; Yexin Lin); writing—original draft preparation (Jiakang Chen); writing—review, editing and commenting (Kasim Allel; Xu Yang; Ruiyang Tu); supervision (Laith Yakob; Chao Zhuo). All authors have read and approved the final version of the manuscript.

Allel K, Stone J, Undurraga EA, Day L, Moore CE, Lin L, Furuya-Kanamori L, Yakob L. The impact of inpatient bloodstream infections caused by antibiotic-resistant bacteria in low- and middle-income countries: A systematic review and meta-analysis. PLoS Med. 2023 Jun 22;20(6):e1004199. doi: 10.1371/journal.pmed.1004199. PMID: 37347726; PMCID: PMC10287017. Jin L, Zhao C, Li H, Wang R, Wang Q, Wang H. Clinical Profile, Prognostic Factors, and Outcome Prediction in Hospitalized Patients With Bloodstream Infection: Results From a 10-Year Prospective Multicenter Study. Front Med (Lausanne). 2021 May 20;8:629671. doi: 10.3389/fmed.2021.629671. PMID: 34095163; PMCID: PMC8172964.
Antimicrobial Resistance Collaborators. Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. Lancet. 2022 Feb 12;399(10325):629-655. doi: 10.1016/S0140-6736(21)02724-0. Epub 2022 Jan 19. Erratum in: Lancet. 2022 Oct 1;400(10358):1102. PMID: 35065702; PMCID: PMC8841637.
GBD 2019 Antimicrobial Resistance Collaborators (2022). Global mortality associated with 33 bacterial pathogens in 2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet (London, England), 400(10369), 2221–2248. https://doi.org/10.1016/S0140-6736(22)02185-7. doi: 10.1007/s00134-012-2695-9. Epub 2012 Sep 26. PMID: 23011531.
Cassini A, Plachouras D, Eckmanns T, Abu Sin M, Blank HP, Ducomble T, Haller S, Harder T, Klingeberg A, Sixtensson M, Velasco E, Weiß B, Kramarz P, Monnet DL, Kretzschmar ME, Suetens C. Burden of Six Healthcare-Associated Infections on European Population Health: Estimating Incidence-Based Disability-Adjusted Life Years through a Population Prevalence-Based Modelling Study. PLoS Med. 2016 Oct 18;13(10):e1002150. doi: 10.1371/journal.pmed.1002150. PMID: 27755545; PMCID: PMC5068791.
Tabah A, Buetti N, Staiquly Q, Ruckly S, Akova M, Aslan AT, Leone M, Conway Morris A, Bassetti M, Arvaniti K, Lipman J, Ferrer R, Qiu H, Paiva JA, Povoa P, De Bus L, De Waele J, Zand F, Gurjar M, Alsisi A, Abidi K, Bracht H, Hayashi Y, Jeon K, Elhadi M, Barbier F, Timsit JF; EUROBACT-2 Study Group, ESICM, ESCMID ESGCIP and the OUTCOMEREA Network. Epidemiology and outcomes of hospital-acquired bloodstream infections in intensive care unit patients: the EUROBACT-2 international cohort study. Intensive Care Med. 2023 Feb;49(2):178-190. doi: 10.1007/s00134-022-06944-2. Epub 2023 Feb 10. PMID: 36764959; PMCID: PMC9916499.
Chen Y, Ji J, Ying C, Liu Z, Yang Q, Kong H, Xiao Y; Blood Bacterial Resistant Investigation Collaborative System (BRICS) Study Group. Blood bacterial resistant investigation collaborative system (BRICS) report: a national surveillance in China from 2014 to 2019. Antimicrob Resist Infect Control. 2022 Jan 24;11(1):17. doi: 10.1186/s13756-022-01055-5. PMID: 35074014; PMCID: PMC8785473.
Zhuo CY, Guo YY, Liu NJ, Liu BM, Xiao SN, Zhang Y, Guo XC, Li DN, Tan SY, He NH, Mai Y, Guan J, Zhuo C. [Epidemiological analysis of pathogens causing bloodstream infections in department of hematology in Guangdong Province]. Zhonghua Xue Ye Xue Za Zhi. 2020 Dec 14;41(12):996-1001. Chinese. doi: 10.3760/cma.j.issn.0253-2727.2020.12.005. PMID: 33445846; PMCID: PMC7840548.
Partridge SR, Kwong SM, Firth N, Jensen SO. Mobile Genetic Elements Associated with Antimicrobial Resistance. Clin Microbiol Rev. 2018 Aug 1;31(4):e00088-17. doi: 10.1128/CMR.00088-17. PMID: 30068738; PMCID: PMC6148190.
Lerminiaux NA, Cameron ADS. Horizontal transfer of antibiotic resistance genes in clinical environments. Can J Microbiol. 2019 Jan;65(1):34-44. doi: 10.1139/cjm-2018-0275. Epub 2018 Sep 24. PMID: 30248271.
Schwaber MJ, Navon-Venezia S, Schwartz D, Carmeli Y. High levels of antimicrobial coresistance among extended-spectrum-beta-lactamase-producing Enterobacteriaceae. Antimicrobial Agents and Chemotherapy. 2005 May;49(5):2137-2139. DOI: 10.1128/aac.49.5.2137-2139.2005. PMID: 15855548; PMCID: PMC1087677.
Mammeri H, Van De Loo M, Poirel L, Martinez-Martinez L, Nordmann P. Emergence of plasmid-mediated quinolone resistance in Escherichia coli in Europe. Antimicrobial Agents and Chemotherapy. 2005 Jan;49(1):71-76. DOI: 10.1128/aac.49.1.71-76.2005. PMID: 15616277; PMCID: PMC538905.
Gandra S, Barter DM, Laxminarayan R. Economic burden of antibiotic resistance: how much do we really know? Clin Microbiol Infect. 2014 Oct;20(10):973-80. doi: 10.1111/1469-0691.12798. Epub 2014 Nov 7. PMID: 25273968.
Zhen X, Lundborg CS, Sun X, Hu X, Dong H. Economic burden of antibiotic resistance in ESKAPE organisms: a systematic review. Antimicrobial Resistance and Infection Control. 2019 ;8:137. DOI: 10.1186/s13756-019-0590-7. PMID: 31417673; PMCID: PMC6692939.
World Health Organization. Global antimicrobial resistance and use surveillance system (GLASS) report: 2021. 2021.
Hu B, Ye H, Xu Y, et al. Clinical and economic outcomes associated with community-acquired intra-abdominal infections caused by extended spectrum beta-lactamase (ESBL) producing bacteria in China. Current Medical Research and Opinion. 2010 Jun;26(6):1443-1449. DOI: 10.1185/03007991003769068. PMID: 20394469.
Murray CJ. Quantifying the burden of disease: the technical basis for disability-adjusted life years. Bull World Health Organ. 1994;72(3):429-445. PMID: 8062401; PMCID: PMC2486718.
MacVane SH, Tuttle LO, Nicolau DP. Impact of extended-spectrum β-lactamase-producing organisms on clinical and economic outcomes in patients with urinary tract infection. Journal of Hospital Medicine. 2014 Apr;9(4):232-238. DOI: 10.1002/jhm.2157. PMID: 24464783.
Clinical and Laboratory Standards Institute (CLSI). Performance Standards for Antimicrobial Susceptibility Testing. 33rd ed. CLSI supplement M100 (ISBN 978-1-68440-170-3 [Print]; ISBN 978-1-68440-171-0 [Electronic]). Clinical and Laboratory Standards Institute, USA, 2023.
Knaus、WA,、Draper、EA,、Wagner、DP,、Zimmerman、JE.、APACHE、II:、a、severity、of、disease、classification、system.、Critical、Care、Medicine.、1985、Oct;13(10):818-829.、PMID:、3928249.
Singer、M,、Deutschman、CS,、Seymour、C,、Shankar-Hari、M,、Annane、D,、Bauer、M,、et、al.、The、third、international、consensus、definitions、for、sepsis、and、septic、shock、(sepsis-3)、JAMA.、2016;315(8):801–810.、doi:、10.1001/jama.2016.0287.
、Charlson、M,、Szatrowski、TP,、Peterson、J,、Gold、J.、Validation、of、a、combined、comorbidity、index.、Journal、of、Clinical、Epidemiology.、1994、Nov;47(11):1245-1251.、DOI:、10.1016/0895-4356(94)90129-5.、PMID:、7722560.
Hill、PC,、Birch、M,、Chambers、S,、et、al.、Prospective、study、of、424、cases、of、Staphylococcus、aureus、bacteraemia:、determination、of、factors、affecting、incidence、and、mortality.、Internal、Medicine、Journal.、2001、Mar;31(2):97-103.、DOI:、10.1111/j.1444-0903.2001.00029.x.、PMID:、11480485.
Apisarnthanarak A, Kiratisin P, Mundy LM. Predictors of mortality from community-onset bloodstream infections due to extended-spectrum beta-lactamase-producing Escherichia coli and Klebsiella pneumoniae. Infect Control Hosp Epidemiol. 2008 Jul;29(7):671-4. doi: 10.1086/588082. PMID: 18624669.
China’s Gross Domestic Product (GDP). Available from: http://data.stats.gov.cn/search.htm?s=GDP.
Barnum H. Evaluating healthy days of life gained from health projects. Soc Sci Med. 1987;24(10):833–841. doi:10.1016/0277-9536(87)90184-5
Murray CJL, Lopez AD, eds. The Global Burden of Disease: Acomprehensive Assessment of Mortality and Disability from Dis-Eases, Injuries, and Risk Factors in 1990 and Projected to 2020. TheGlobal Burden of Disease and Injury Series. Vol. 1 Boston, MA: Harvard University Press, for Harvard School of PublicHealth on behalf of the World Health Organisation and TheWorld Bank; 1996b.
Lopez AD, Mathers CD, Ezzati M, et al., editors. Global Burden of Disease and Risk Factors. Washington (DC): The International Bank for Reconstruction and Development / The World Bank; 2006. Available from: https://www.ncbi.nlm.nih.gov/books/NBK11812/ Co-published by Oxford University Press, New York.
Center for the Evaluation of Value and Risk in Health. The Cost-Effectiveness Analysis Registry [Internet]. (Boston), Institute for Clinical Research and Health Policy Studies, Tufts Medical Center. Available from: www.cearegistry.org Accessed on [Date].
Devleesschauwer B, Havelaar AH, Maertens de Noordhout C, Haagsma JA, Praet N, Dorny P, Duchateau L, Torgerson PR, Van Oyen H, Speybroeck N. Calculating disability-adjusted life years to quantify burden of disease. Int J Public Health. 2014 Jun;59(3):565-9. doi: 10.1007/s00038-014-0552-z. Epub 2014 Apr 22. PMID: 24752429.
JA Fox Rushby , K Hanson, Calculating and presenting disability adjusted life years (DALYs) in cost-effectiveness analysis, Health Policy and Planning, Volume 16, Issue 3, September 2001, Pages 326–331, https://doi.org/10.1093/heapol/16.3.326
Wang Y, Xiao T, Zhu Y, et al. Economic Burden of Patients with Bloodstream Infections Caused by Extended-Spectrum β-Lactamase-Producing <i>Escherichia coli</i>. Infection and Drug Resistance. 2020 ;13:3583-3592. DOI: 10.2147/idr.s271230. PMID: 33116678; PMCID: PMC7568608.
Caliendo, M. and Kopeinig, S. (2008), SOME PRACTICAL GUIDANCE FOR THE IMPLEMENTATION OF PROPENSITY SCORE MATCHING. Journal of Economic Surveys, 22: 31-72. https://doi.org/10.1111/j.1467-6419.2007.00527.x
Bommae Kim. 2016. “Introduction to Mediation Analysis.” UVA Library StatLab. https://library.virginia.edu/data/articles/introduction-to-mediation-analysis/ (accessed September 14, 2023).
SobelM E . D irect and indirect effects in linear structural equation mode ls . In:J S Long (Ed). Common problem s /proper solutions . Beverly H ills, CA:Sage. 1988. 46～ 64
China Antimicrobial Surveillance Network http://www.chinets.com/Chinet.
Tumbarello M, Spanu T, Di Bidino R, et al. Costs of bloodstream infections caused by Escherichia coli and influence of extended-spectrum-beta-lactamase production and inadequate initial antibiotic therapy. Antimicrobial Agents and Chemotherapy. 2010 Oct;54(10):4085-4091. DOI: 10.1128/aac.00143-10. PMID: 20660675; PMCID: PMC2944559.
Quan J, Zhao D, Liu L, Chen Y, Zhou J, Jiang Y, Du X, Zhou Z, Akova M, Yu Y. High prevalence of ESBL-producing Escherichia coli and Klebsiella pneumoniae in community-onset bloodstream infections in China. J Antimicrob Chemother. 2017 Jan;72(1):273-280. doi: 10.1093/jac/dkw372. Epub 2016 Sep 13. PMID: 27624571.
World Health Organization. (‎2021)‎. WHO integrated global surveillance on ESBL-producing E. coli using a “One Health” approach: implementation and opportunities. World Health Organization. https://apps.who.int/iris/handle/10665/340079. License: CC BY-NC-SA 3.0 IGO
World Health Organization. (‎2019)‎. AWaRe – a new WHO tool to help countries improve antibiotic treatment, increase access and reduce resistance. https://adoptaware.org
Karisik E, Ellington MJ, Pike R, et al.Differentia expression of CTX-M-15-β-lactamase between two major Escherichia coli strains responsible for outbreaks in the UK[J].Clin Microbiol Infect Dis,2006, 16(4):1-4.

No competing interests reported.

Supplementarymaterial.docx

Download PDF

Version 2

posted

You are reading this latest preprint version

Extended-spectrum β-lactamase-producing Escherichia coli and Klebsiella pneumoniae in Guangzhou, China: risk factors and economic burden among Patients with bloodstream infections

Status:

Version 2

Abstract

Figures

Highlights

Introduction

Patients and Methods

Results

Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 2