A nomogram to predict nosocomial infection in patients on venoarterial extracorporeal membrane oxygenation after cardiac surgery

Abstract

Introduction

After cardiac surgery, patients on venoarterial extracorporeal membrane oxygenation (VA-ECMO) have a higher risk of nosocomial infection in the intensive care unit (ICU). We aimed to establish an intuitive nomogram to predict the probability of nosocomial infection in patients on VA-ECMO after cardiac surgery.

Methods

We included patients on VA-ECMO after cardiac surgery between January 2011 and December 2020 at a single center. We developed a nomogram based on independent predictors identified using univariate and multivariate logistic regression analyses. We selected the optimal model and assessed its performance through internal validation and decision-curve analyses.

Results

Overall, 503 patients were included; 363 and 140 patients were randomly divided into development and validation sets, respectively. Independent predictors derived from the development set to predict nosocomial infection included older age, white blood cell (WBC) count abnormality, ECMO environment in the ICU, and mechanical ventilation (MV) duration, which were entered into the model to create the nomogram. The model showed good discrimination, with areas under the curve (95% confidence interval) of 0.743 (0.692–0.794) in the development set and 0.732 (0.643–0.820) in the validation set. The optimal cutoff probability of the model was 0.457 in the development set (sensitivity, 0.683; specificity, 0.719). The model showed qualified calibration in both the development and validation sets (Hosmer–Lemeshow test, p > .05). The threshold probabilities ranged from 0.20 to 0.70.

Conclusions

For adult patients receiving VA-ECMO treatment after cardiac surgery, a nomogram-monitoring tool could be used in clinical practice to identify patients with high-risk nosocomial infections and provide an early warning.

Keywords

cross-infection extracorporeal membrane oxygenation risk factors predictive model nomogram

Introduction

Since the first successful clinical application of extracorporeal membrane oxygenation (ECMO) in 1972, ¹ ECMO technology has been used clinically for nearly 40 years. Although ECMO is widely used in refractory cardiogenic shock, severe respiratory failure, and cardiac arrest, ² the in-hospital mortality rate associated with venoarterial (VA)-ECMO is as high as 59.6%. ³ It is well known that nosocomial infection is the most common complication of ECMO treatment. ⁴ Based on data issued by the Extracorporeal Life Support Organization (ELSO), patients have a worse prognosis when complicated by nosocomial infection. ^5,6 Prevention and control of nosocomial infections are very important for the survival of patients on ECMO.

However, it is very difficult to identify nosocomial infections (ignoring the source of infection ⁷) in patients with cardiogenic shock after cardiac surgery. ⁸ The most important challenge is to clinically differentiate between sepsis and systemic inflammatory response syndrome. ⁹ Hypotension, fever, and laboratory tests, such as those for C-reactive protein and procalcitonin, have limited value in the diagnosis of nosocomial infection. Moreover, septic shock can be easily neglected because of the presence of cardiogenic shock. ⁸

At present, there is still a lack of effective means to recognize early warning signs of nosocomial infections in patients receiving VA-ECMO treatment after cardiac surgery. Here, we reviewed the occurrence of nosocomial infections in adult patients who received VA-ECMO after cardiac surgery during the entire hospitalization period. We developed a prediction model for nosocomial infections in these populations to provide a useful monitoring tool for use in clinical practice.

Methods

Study design and internal validation

We performed a retrospective observational study on nosocomial infection in patients receiving VA-ECMO therapy after cardiac surgery between 1 January 2011, and 31 December 2020, at the Cardiac Intensive Care Unit, Beijing Anzhen Hospital, Beijing, China. We randomly selected approximately 70% of the cases as the development set and 30% as the validation set. We developed a nomogram from the development set to predict nosocomial infections in patients receiving VA-ECMO after cardiac surgery.

This study was approved by the Institutional Review Board of Beijing Anzhen Hospital, Capital Medical University (No. 2,022,064X) on 21 March 2022. The committee waived the requirement for informed consent from individual patients because of the retrospective study design. ¹⁰

Study population and definition

The inclusion criteria in this study were as follows: (1) age ≥18 years, (2) patients who did not have prior infections and received VA-ECMO support after cardiac surgery, and (3) ECMO support duration ≥48 h. The exclusion criteria were as follows: (1) patients on VA-ECMO treatment with incomplete medical records, (2) death or weaning for any reason within 48 h after ECMO support, (3) venovenous ECMO (VV-ECMO) support, and (4) VA-ECMO for patients who did not undergo cardiac surgery.

Based on the definition of nosocomial infection from the US Center for Disease Control and Prevention, lower respiratory tract infection (LRTI) was defined as the presence of clinical manifestations (two or more of the following criteria: fever [more than 38.0°C or less than 36.0°C]; white blood cell [WBC] count >12×10⁹/L or less than 4×10⁹/L; purulent excretion from the lower respiratory tract; new or consistent infiltration from chest photographs), positive sputum culture, and initiation of antimicrobial therapy under the guidance of intensivists or infectious disease specialists. ^11–15 Bloodstream infection (BSI) was defined as the presence of any of the clinical manifestations (including fever >38°C, chills, or hypotension) accompanied by a positive blood culture. ^13,16–19 Urinary tract infection (UTI) was defined as the presence of symptomatic signs or positive urine routine results with a positive urine culture. ¹¹ Surgical site infections (SSI) were defined as infections occurring within 30 days of surgery involving the skin, subcutaneous tissue, or muscle layer, or any item of the following: purulent drainage, positive culture from the drainage fluid, and infection identified by the surgeon. ^13,20 Skin and soft tissue infections (SSTI) are defined as purulent drainage in the skin and soft tissue, any microorganism isolated from the culture of skin or soft tissue, or a surgeon providing debridement confirming of infection. ¹³ Nosocomial infection during ECMO was defined as infection occurring from 24 h after ECMO initiation to 48 h after ECMO removal.^14,15,21 Nosocomial infection after ECMO was defined as infection occurring from 48 h after ECMO removal to discharge from the hospital.

All sources and frequencies of infection were recorded; however, the first infection was regarded as the primary source when multiple infections occurred in the same patient.

ECMO support and infection control measures

The indication for VA-ECMO was determined by cardiac surgery and an extracorporeal life support physician. ¹⁵ VA-ECMO cannulas were surgically implanted by trained team members using the femoral vein-femoral artery approach. The ECMO flow rate titration was adjusted according to lactic acid levels and organ function to ensure adequate tissue perfusion. A systematic heparin anticoagulation strategy was performed so that the activated clotting time was 180–220 s.

All the patients received antibiotics to prevent nosocomial infections when ECMO was initiated. Third-generation cephalosporins were routinely administered. Once nosocomial infection was suspected, empirical combination antibiotics were administered to cover gram-negative bacilli and gram-positive cocci, according to the local microorganisms in the intensive care unit (ICU). When the results of pathogenic culture were clear, antibiotics were administered as the target treatment.

Data collection

We retrieved data from the electronic medical records, including demographic variables (age, sex, nationality, past medical history, and underlying disease), specific data related to VA-ECMO (exact operation duration, duration of ECMO support, use of intra-aortic balloon pump [IABP], continuous renal replacement therapy [CRRT], ventilator usage, ECMO environment, and complications of ECMO therapy), the worst results of laboratory tests within 48 h of ECMO initiation (WBC counts, platelets, procalcitonin [PCT], liver function test, renal function test, blood gas analysis, and microbiological data), maximal vasoactive-inotropic score (VIS), ^22,23 Sequential Organ Failure Assessment (SOFA) score ²⁴), the diagnosis date of nosocomial infection, and outcome after therapy.

Statistical analysis

We performed the Kolmogorov–Smirnov test to assess the distribution characteristics of the continuous parameters. All numerical variables showed a skewed distribution. Skewed distribution data were expressed as medians and interquartile ranges. Categorical data were reported as frequencies or ratios. Referring to the definition of early onset ventilator-associated pneumonia, we grouped all durations of medical device use into intervals of 5 days. ²⁵ The optimal cutoff values for most of the other continuous variables were divided into two groups. We used the Mann–Whitney U test to analyze differences in continuous variables between groups in the univariate analysis, and the chi-square test or Fisher’s exact test for categorical variables. The collinearity of the covariates was assessed, and model fitting was based on the Hosmer–Lemeshow goodness-of-fit test. Variables from the univariate analysis with p values of less than 0.1 were selected for multivariate analysis. The final selection of the multivariate logistic regression model was based on regression using the Akaike information criterion (AIC). ²⁶ We developed a nomogram based on the results of multivariate analysis in the development set to obtain nosocomial infection probability estimates. The discrimination ability of the model was assessed by calculating the area under the receiver operating characteristic (ROC) curve (AUC). The best cutoff value of the model was regarded as the boundary point of the diagnosis, and patients were stratified into low- or high-risk nosocomial infections. ²⁷ Internal validation was performed on the validation set according to the established model. Calibration ability was assessed using the Hosmer–Lemeshow test. Decision curve analysis (DCA) was used to evaluate the clinical utility of model-assisted decisions. ²⁸ An interactive regression plot based on the prediction model was constructed to calculate the probability of nosocomial infection.

SPSS for Windows, version 26 (SPSS, Chicago, IL, USA), and R statistical software version 4.1.2 (https://www.r-project.org/) were used to analyze the data. Packages “pROC” “calibrate,” “MASS,” “rms,” “rmda,” “regplot” were used to create plots in this study. Statistical significance was set at p < .05.

Results

Patient characteristics

A flowchart of the patient enrolment process is shown in Figure 1. A total of 503 patients were included in this study. Of these, 213 (42.3%) met the diagnostic criteria for nosocomial infection. Using the random sampling method, 363 (approximately 70% of all cases) patients were allocated to the development set, and 140 (approximately 30% of all cases) patients were included in the validation set. Nosocomial infection occurred in 164 (45.2%) patients in the development set and 49 (35.0%) patients in the validation set. The prevalences of LRTI, BSI, UTI, and other infections during ECMO were 97 (19.3%), 60 (11.9%), 1 (0.2%), and 8 (1.6%), respectively. The corresponding prevalences after ECMO were 55 (10.9%), 40 (8.0%), 3 (0.6%), and 32 (6.4%), respectively (Table S1).

Figure 1.

Flowchart for screening patients for the predictive model in the study.

The characteristics of the study population are summarized in Table S2. Among the 40 variables, compared with the validation set, patients in the development set accounted for a higher ratio in the VIS ≤40 and PCT >4 mg/L groups. The remaining variables have the same distribution.

Independent predictors of nosocomial infection

Univariate logistic regression analysis in the development set indicated that age (≥65 years), operation duration (>420 min), mechanical ventilation (MV) duration, ECMO duration, ECMO environment (in the ICU), blood urea nitrogen, and WBC count abnormalities (<4 or >12 × 10⁹/L) were associated with nosocomial infection (Table 1). These seven variables, all with p values less than 0.1, were entered into a multivariate logistic regression analysis in the development set by backward stepwise and forward stepwise, respectively. The same result indicated that the four variables remained in the equation. After comparison and screening of the nosocomial infection models, Model C was regarded as the final model in this study (Table S3). The independent predictors for nosocomial infection included four variables: older age (≥65 years) (odds ratio [OR], 1.699; 95% confidence interval [CI], 1.029–2.806, p = .038), ECMO environment (in ICU; OR, 2.122, 95% CI, 1.328–3.390; p = .002), WBC count abnormality (<4 × 10⁹/L or >12 × 10⁹/L; OR, 1.615; 95% CI, 0.987–2.674; p = .057), and MV duration (0–5 days as the reference group, 6–10 days, or >10 days; OR, 2.112 (1.194–3.737), 7.567 (4.070–14.102), respectively; p = .010 and <0.001) (Table 1).

Table 1.

Logistic regression analysis of predictive factors for nosocomial infection in adult patients on VA ECMO after cardiac surgery.

Variables	Univariate logistic regression analysis		Multivariate logistic regression analysis
Variables	OR (95% CI)	p value	OR (95% CI)	p value
Age ≥65	1.588 (1.013–2.489)	.043	1.699 (1.029–2.806)	0.038
IABP duration	1.491 (1.127–1.971)	.005
MV group (1-5)d	1	—	1	—
MV group (6-10)d	1.859 (1.074–3.217)	.027	2.112 (1.194–3.737)	0.010
MV group (>10 days)	6.827 (3.755–12.413)	< .001	7.567 (4.070–14.102)	< .001
CRRT duration	1.840 (1.304–2.597)	.001
ECMO duration	1.801 (1.294–2.508)	< .001
ECMO site	2.065 (1.35–3.159)	.001	2.122 (1.328–3.390)	0.002
Bun	1.317 (0.971–1.788)	.077
WBC group	1.552 (0.994–2.422)	.053	1.615 (0.987–2.674)	0.057

Abbreviation: IABP: Intra-Aortic Balloon Pump; CRRT: Continuous Renal Replacement Therapy;MV: mechanical ventilation; ECMO: extracorporeal membrane oxygenation; WBC: white blood cell

Model assessment and internal validation

We performed the discrimination assessment in both the development and validation sets. The area under the curve (AUC) in the development set was 0.743 (95% CI, 0.692–0.794) (Figure 2(a)), and that in the validation set was 0.732 (95% CI, 0.643–0.820) (Figure 2(b)). The AUC values in the development and validation sets were greater than 0.70, indicating that the predictive model had good discrimination.

Figure 2.

The area under the curve (AUC), calibration curve and decision curve analysis in development set and validation set, respectively. (A) Receiver operating characteristic curve (ROC) in the development set. The area under the curve (AUC) for the model in the development set was 0.743 (95% confidence interval [CI] 0.692–0.794). (B) ROC in the validation set. The AUC for the model in the validation set was 0.732 (95% CI 0.643–0.820). (C) Calibration curve in the derivation set. (Hosmer–Lemeshow test, χ²=5.761, p = .764). (D) Calibration curve in the validation set. (Hosmer–Lemeshow test, χ²=12.419, p = .191). (E) Decision curve analysis in the development set. (F) Decision curve analysis in the validation set.

The calibration curve of the predictive model was shown by the Hosmer–Lemeshow test, which revealed no significant difference in either the development set (χ² = 5.761; p = .764) or the validation set (χ²=12.419; p = .191). (Figure 2(c) and Figure 2(d)). The calibration curve in the validation set lied below the reference line, suggesting that the model might slightly overestimate the risk of nosocomial infections.

DCA was performed on the development and validation sets (Figures 2(e) and (f)). When threshold probability was in the range of 0.20–0.70, the net benefit would gain from the model compared to “treat all” or “treat none” strategies.

Nomogram for the predicting model

We created a nomogram to predict the probability of nosocomial infections in adult patients receiving VA-ECMO after cardiac surgery. The nomogram is shown in Figure 3. The optimal probability for predicting nosocomial infection was 0.457. At this cutoff value, the Youden index was 0.402, sensitivity was 71.9%, specificity was 68.3%, positive predictive value was 64.1%, negative predictive value was 75.5%. The corresponding total score for the model is approximately 85 points.

Figure 3.

Nomogram for predicting nosocomial infection in patients on venoarterial extracorporeal membrane oxygenation (VA-ECMO) after cardiac surgery. (A) Nomogram for the model. (B) Interactive nomogram for the patient as an example. Figure legend. For example, a 68-year-old patient who was admitted with coronary artery disease for coronary artery bypass grafting. After surgery, he was transferred to the intensive care unit (ICU). He received MV for 3 days after surgery, after which he received VA-ECMO treatment because of circulatory instability. The white blood cell (WBC) count was 15×109/L. Hence, his total score from the model of nosocomial infection was as follows: for age >65 years, the score was 26; for WBC count in the abnormal range, the score was 24; for ECMO environment being the ICU, the score was 37; and for an MV duration of 3 days, the score was 0. Hence, the total score was 26+24+37=87, and the probability of nosocomial infection was approximately 45.8%, according to the interactive nomogram (more than the cutoff point 0.457), which would suggest that the patient was in the infectious state. The total score calculated by R software was shown automatically with a red dot, and the corresponding probability was shown with a red arrow in the interactive nomogram (B).

Model clinical application

Four independent variables were converted to the corresponding values depending on the degree of risk. If age was ≥65 years, WBC count in the abnormal range (<4 × 10⁹/L or >12 × 10⁹/L), ECMO environment in the ICU, then the scores were 26, 24, and 37, respectively. If the MV duration was in the range of 6–10 days or >10 days, the corresponding scores were 37 and 100, respectively. Subsequently, the total score was calculated, and the probability of nosocomial infection was easily found in the nomogram (Table 2, Figure 3). When the probability was ≥0.457, the patient was considered at a high risk of nosocomial infection. Otherwise, the patient was in a low-risk state of nosocomial infection.

Table 2.

The model score for nosocomial infection in adult patients on VA ECMO after cardiac surgery.

Risk factor	Categories	Reference	Coefficient	Scores
Age_65	< 65	0	0.539	0
Age_65	≥65	1		26
WBC group	4–12*G/L	0	0.479	0
WBC group	< 4 or > 12*G/L	1		24
ECMO site	Non-ICU	0	0.752	0
ECMO site	ICU	1		37
MV group	(1-5) d	1	—	0
	(6-10) d	2	0.748	37
	(>10) d	3	2.025	100

Abbreviation: MV: mechanical ventilation; ECMO: extracorporeal membrane oxygenation; WBC: white blood cell

Discussion

To the best of our knowledge, this is the first model to predict the probability of nosocomial infection based on independent factors after multivariate logistic regression. In this study, for adult patients receiving VA-ECMO treatment after cardiac surgery, four independent risk predictors were included: older age (≥65 years), WBC abnormality, ECMO environment in the ICU, and MV duration. The predictive model conducted using these four variables had good discrimination and calibration, which can play a role in identifying high-risk clinical warnings of nosocomial infections.

Many studies have investigated the risk factors for nosocomial infections in patients receiving ECMO therapy. Most of these studies have suggested that ECMO support duration is an independent risk factor. ^{6,14,15,19,29–34} Additionally, other studies reported independent risk factors including SOFA score, ^33,35,36 older age, ^37–39 ICU length of stay, ^40,41 preoperative creatinine level, ³² VV-ECMO mode, ⁴² ventilator duration before ECMO removal, ⁴² VA-ECMO mode, ⁴³ and immunosuppression status. ⁴³ These discrepancies were due to the different sources of the included patients, smaller sample sizes, and different definitions of nosocomial infection among these studies.

MV duration was an independent risk factor in our model, which was consistent with the findings of a study by Gao et al. ⁴⁴ in which the risk of hospital-acquired infection increased by 18% per 24-hour increment in ventilator usage. In our study, the risk of nosocomial infection reached 2.1 fold and 7.6 fold, with 6–10 days and more >10 days of MV, respectively. Consistent with the findings of most studies, ^37–39 we also found that the risk of nosocomial infection increased significantly in older age.

Unexpectedly, ECMO environment in the ICU was an independent risk factor in our model, contrasting that in the non-ICU. Compared with the operating room environment, the ICU environment was vulnerable to microorganism contamination. However, a more reasonable explanation might be that some patients received ECMO treatment in the ICU because deterioration was caused by nosocomial infection after surgery at that time.

Nomograms are more intuitive and convenient than other visualization methods. If the probability reaches the cutoff point of 0.457, the patient is most likely at a high-risk infectious state. As variables such as age, ECMO environment, and WBC count were unmodifiable clinical factors, only MV duration was a potentially adjustable factor. From the nomogram, the probability of nosocomial infection in patients who received MV for more than 10 days was over 50%, indicating that those patients were at high risk of infection. Therefore, it is better to extubate as soon as possible within 10 days after surgery to reduce nosocomial infection opportunities.

When a patient was thought to be at a high risk of nosocomial infection, some measures were recommended. Increasing the frequency of testing of various pathogens (blood, urine, drainage, sputum), screening via ultrasound and radiological bedside check-up, strengthening the anti-infection drugs, and assessing and removing unnecessary catheters or devices as soon as possible would help reduce nosocomial infections. “Awake ECMO” might reduce the incidence of ventilator-associated pneumonia, as reported by Montero S et al. ⁴⁵

The prevalence of nosocomial infection in adult patients receiving ECMO treatment in our study was 42.3%, which was higher than that (20.9%) reported by ELSO. ^5,6 A possible explanation is that these studies focused only on nosocomial infections during ECMO, whereas our study included nosocomial infections not only during ECMO support but also after ECMO removal.

Some limitations of this study should be considered when interpreting the results. First, as a retrospective study, the time span was up to 10 years; this may have been much more advanced, and the prevalence of infection is unstable over time. Thus, the prediction efficiency decreased with time. Second, the model is only suitable for patients without prior infection who received VA-ECMO after cardiac surgery. Our results cannot be generalized to other centers because we did not have external validation. Third, similar to other models for predicting nosocomial infection, ^27,44,46 our model predicts the probability of nosocomial infection, but does not describe the exact infection site.

Conclusions

Four independent predictors of nosocomial infection in patients during and after VA-ECMO after cardiac surgery were older age (≥65 years), WBC count abnormality (<4 or >12×10⁹/L), ECMO environment (ICU), and MV duration (>5 days or >10 days). The nomogram created using these risk factors showed reliable performance in predicting nosocomial infection in patients during and after VA-ECMO after cardiac surgery.

Supplemental Material

Supplemental Material - A nomogram to predict nosocomial infection in patients on venoarterial extracorporeal membrane oxygenation after cardiac surgery

Supplemental Material for A nomogram to predict nosocomial infection in patients on venoarterial extracorporeal membrane oxygenation after cardiac surgery by Xiyuan Li, Liangshan Wang, Chenglong Li, Xiaomeng Wang, Xing Hao, Zhongtao Du, Haixiu Xie, Feng Yang, Hong Wang, and Xiaotong Hou in Perfusion

Footnotes

Acknowledgements

We would like to thank senior engineer Wu Yuhao, Electronic Medical Record, and the Information Center of Anzhen Hospital.

Author contributions

XYL, LSW, CLL, XMW, XH, HW, ZTD, HXX, and FY contributed to the data acquisition. XYL and LSW wrote the manuscript’s initial draft. XYL, LSW, and XMW performed statistical analyses. XYL, LSW, HW, and XTH contributed to the study conception and design. HW, and XTH contributed to the revision of this paper and final approval of the version to be published. All authors have read and approved the final manuscript.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the 1. Beijing Hospitals Authority “Ascent Plan” (FDL20190601, to X Hou). 2. The “Sailing” Program of key medical specialty of Beijing Hospitals Authority in 2021: for critical care medicine (fund number for extracorporeal life support: ZYLX202111, to X Li). The funding body had no role in the design of the study; collection, analysis, and interpretation of data; or in writing the manuscript.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Ethics approval

Ethical approval was granted according to national requirements. This study was approved by the Medical Ethics Committee of the Beijing Anzhen Hospital, Capital Medical University. The requirement for informed consent was waived because the study was performed retrospectively and no interventions were applied.

Supplemental Material

Supplemental material for this article is available online.

Appendix

References

Hill

O'Brien

Murray

, et al. Prolonged extracorporeal oxygenation for acute post-traumatic respiratory failure (shock-lung syndrome). Use of the Bramson membrane lung. N Engl J Med 1972; 286(12): 629–634.

Wrisinger

Thompson

. Basics of extracorporeal membrane oxygenation. Surg Clin North Am 2022; 102(1): 23–35.

Smith

Vukomanovic

Brodie

, et al. Duration of veno-arterial extracorporeal life support (VA ECMO) and outcome: an analysis of the extracorporeal life support organization (ELSO) registry. Crit Care 2017; 21(1): 45.

MacLaren

Schlapbach

Aiken

. Nosocomial infections during extracorporeal membrane oxygenation in neonatal, pediatric, and adult patients: a comprehensive narrative review. Pediatr Crit Care Med 2020; 21(3): 283–290.

Bizzarro

Conrad

Kaufman

, et al. Infections acquired during extracorporeal membrane oxygenation in neonates, children, and adults. Pediatr Crit Care Med 2011; 12(3): 277–281.

Vogel

Lew

Kao

, et al. Defining risk for infectious complications on extracorporeal life support. J Pediatr Surg 2011; 46(12): 2260–2264.

Allyn

Ferdynus

Lo Pinto

, et al. Complication patterns in patients undergoing venoarterial extracorporeal membrane oxygenation in intensive care unit: multiple correspondence analysis and hierarchical ascendant classification. PLoS One 2018; 13(9): e0203643.

Gopalakrishnan

Vashisht

. Sepsis and ECMO. Indian J Thorac Cardiovasc Surg 2021; 37(suppl 2): 267–274.

Carlet

. Rapid diagnostic methods in the detection of sepsis. Infect Dis Clin North Am 1999; 13(2): 483–494. xi.

10.

Wang

Huang

Zhou

, et al. Predicting the occurrence of multidrug-resistant organism colonization or infection in ICU patients: development and validation of a novel multivariate prediction model. Antimicrob Resist Infect Control 2020; 9(1): 66.

11.

Quintana

Mazzeffi

Galvagno

, et al. A retrospective study of infection in patients requiring extracorporeal membrane oxygenation support. Ann Thorac Surg 2021; 112(4): 1168–1175.

12.

Bouglé

Bombled

Margetis

, et al. Ventilator-associated pneumonia in patients assisted by veno-arterial extracorporeal membrane oxygenation support: epidemiology and risk factors of treatment failure. PLoS One 2018; 13(4): e0194976.

13.

Garner

Jarvis

Emori

, et al. CDC definitions for nosocomial infections, 1988. Am J Infect Control 1988; 16(3): 128–140.

14.

Hsu

Chiu

Huang

, et al. Risk factors for nosocomial infection during extracorporeal membrane oxygenation. J Hosp Infect 2009; 73(3): 210–216.

15.

Wang

Jia

, et al. Extracorporeal membrane oxygenation-related nosocomial infection after cardiac surgery in adult patients. Braz J Cardiovasc Surg 2021; 36(6): 743–751.

16.

Silvetti

Ranucci

Pistuddi

, et al. Bloodstream infections during post-cardiotomy extracorporeal membrane oxygenation: incidence, risk factors, and outcomes. Int J Artif Organs 2019; 42(6): 299–306.

17.

Yun

Hong

Jung

, et al. Epidemiology and clinical characteristics of bloodstream infection in patients under extracorporeal membranous oxygenation. J Intensive Care Med 2020; 36(6): 1053–1060.

18.

Menaker

Galvagno

Rabinowitz

, et al. Epidemiology of blood stream infection in adult extracorporeal membrane oxygenation patients: a cohort study. Heart and Lung 2019; 48(3): 236–239.

19.

Zhang

. Analysis of nosocomial infection and risk factors in patients with ECMO treatment. Infect Drug Resist 2021; 14: 2403–2410.

20.

Horan

Andrus

Dudeck

. CDC/NHSN surveillance definition of health care-associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control 2008; 36(5): 309–332.

21.

Schmidt

Bréchot

Hariri

, et al. Nosocomial infections in adult cardiogenic shock patients supported by venoarterial extracorporeal membrane oxygenation. Clin Infect Dis 2012; 55(12): 1633–1641.

22.

Gaies

Gurney

Yen

, et al. Vasoactive-inotropic score as a predictor of morbidity and mortality in infants after cardiopulmonary bypass. Pediatr Crit Care Med 2010; 11(2): 234–238.

23.

Hyun

Kim

Lee

, et al. Vasoactive-inotropic score as a determinant of timely initiation of venoarterial extracorporeal membrane oxygenation in patients with cardiogenic shock. Circ J 2021; 86: 687–694.

24.

Vincent

Moreno

Takala

, et al. The SOFA (sepsis-related organ failure assessment) score to describe organ dysfunction/failure. On behalf of the working group on sepsis-related problems of the European society of intensive care medicine. Intensive Care Med 1996; 22(7): 707–710.

25.

Hugonnet

Uçkay

Pittet

. Staffing level: a determinant of late-onset ventilator-associated pneumonia. Crit Care 2007; 11(4): R80.

26.

Wang

Han

, et al. Prognostic value of preoperative inflammatory markers in patients with hepatocellular carcinoma who underwent curative resection. Cancer Cel Int 2021; 21(1): 500.

27.

Zhang

Yang

Mou

, et al. An interactive nomogram to predict healthcare-associated infections in ICU patients: a multicenter study in GuiZhou Province, China. PLoS One 2019; 14(7): e0219456.

28.

Vickers

Cronin

Elkin

, et al. Extensions to decision curve analysis, a novel method for evaluating diagnostic tests, prediction models and molecular markers. BMC Med Inform Decis Mak 2008; 8: 53.

29.

Allou

Lo Pinto

Persichini

, et al. Cannula-related infection in patients supported by peripheral ECMO: clinical and microbiological characteristics. ASAIO J 2019; 65(2): 180–186.

30.

Juthani

Macfarlan

, et al. Incidence of nosocomial infections in adult patients undergoing extracorporeal membrane oxygenation. Heart and Lung 2018; 47(6): 626–630.

31.

Kutleša

Santini

Krajinović

, et al. Nosocomial blood stream infections in patients treated with venovenous extracorporeal membrane oxygenation for acute respiratory distress syndrome. Minerva Anestesiologica 2017; 83(5): 493–501.

32.

Kim

Lee

Park

, et al. Nosocomial infection in adult patients undergoing veno-arterial extracorporeal membrane oxygenation. J Korean Medical Science 2017; 32(4): 593–598.

33.

Pieri

Agracheva

Fumagalli

, et al. Infections occurring in adult patients receiving mechanical circulatory support: the two-year experience of an italian national referral tertiary care center. Medicina Intensiva 2013; 37(7): 468–475.

34.

Sun

Tsai

, et al. Infections occurring during extracorporeal membrane oxygenation use in adult patients. J Thorac Cardiovasc Surg 2010; 140(5): 1125–1132.

35.

Wang

Huang

, et al. Bloodstream infections in patients undergoing extracorporeal membrane oxygenation. Pakistan J Med Sci 2020; 36(6): 1171–1176.

36.

Huh

Kim

, et al. Nosocomial infections in in-hospital cardiac arrest patients who undergo extracorporeal cardiopulmonary resuscitation. PLoS ONE 2020; 15(12): e0243838.

37.

Gastmeier

Sohr

Geffers

, et al. Risk factors for death due to nosocomial infection in intensive care unit patients: findings from the krankenhaus infektions surveillance system. Infect Control Hosp Epidemiol 2007; 28(4): 466–472.

38.

Blot

Cankurtaran

Petrovic

, et al. Epidemiology and outcome of nosocomial bloodstream infection in elderly critically ill patients: a comparison between middle-aged, old, and very old patients. Crit Care Med 2009; 37(5): 1634–1641.

39.

Fang

Yang

, et al. Application and comparison of scoring indices to predict outcomes in patients with healthcare-associated pneumonia. Crit Care 2011; 15(1): R32.

40.

Hessels

Kuo

Ahmed

. Epidemiology and impact of healthcare-associated infections in trauma patients: a national data analysis. Surg Infect (Larchmt) 2020; 21(10): 871–876.

41.

Hermann

Hauw-Berlemont

Augy

, et al. Epidemiology and predictors of long-stays in medical ICU: a retrospective cohort study. J Intensive Care Med 2021; 36(9): 1066–1074.

42.

Wang

Huang

, et al. Risk factors and prognosis of nosocomial pneumonia in patients undergoing extracorporeal membrane oxygenation: a retrospective study. J Int Med Res 2020; 48(10): 300060520964701.

43.

Austin

Kerr

Al-Soufi

, et al. Nosocomial infections acquired by patients treated with extracorporeal membrane oxygenation. Crit Care Resusc 2017; 19(suppl 1): 68–75.

44.

Gao

Wang

, et al. Establishment and validation of a nomogram to predict hospital-acquired infection in elderly patients after cardiac surgery. Clin Interv Aging 2022; 17: 141–150.

45.

Montero

Huang

Rivas-Lasarte

, et al. Awake venoarterial extracorporeal membrane oxygenation for refractory cardiogenic shock. Eur Heart J Acute Cardiovasc Care 2021; 10(6): 585–594.

46.

Jiang

Shi

Zheng

, et al. Establishment and validation of a nomogram to predict the in-hospital death risk of nosocomial infections in cancer patients. Antimicrob Resist Infect Control 2022; 11(1): 29.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.33 MB