Developing a Registry of Healthcare-Associated Infections at Intensive Care Units in West China: Study Rationale and Patient Characteristics

Development and validation of dynamic clinical subphenotypes in acute pancreatitis patients using vital sign trajectories in intensive care units: a multinational cohort study

Acute pancreatitis (AP) is a heterogeneous inflammatory condition. Although emerging therapeutic strategies targeting pathways such as calcium signaling, TNF-α, the NLRP3 inflammasome, and HMGB1 have shown promise, their efficacy may be limited by the underlying biological heterogeneity of the disease. In this multinational retrospective cohort study across three large ICU databases (ICU-HAI, MIMIC-IV, and eICU-CRD), we used group-based trajectory modeling of early vital signs to identify four distinct AP subphenotypes: hyperinflammatory, hypertensive, hypotensive, and hypoinflammatory. These subtypes differed markedly in 30-day mortality, inflammatory burden, and hemodynamic stability. Compared to the hypertensive phenotype, hyperinflammatory and hypotensive patients had significantly higher 30-day mortality risks (HR = 3.38 and HR = 1.87, respectively), while the hypoinflammatory phenotype carried no excess risk. Fluid resuscitation responses were phenotype-specific: hyperinflammatory patients benefited from higher fluid volumes, whereas hypoinflammatory patients were at risk of fluid overload. Notably, distinct subphenotypes displayed unique responses to fluid intake over the first two ICU days. For hyperinflammatory phenotype, the algorithm-estimated lowest-risk fluid range was 4100-4300 mL on day 1 and 3400-3600 mL on day 2; for phenotype hypoinflammatory phenotype, the optimal ranges were 2800-3800 mL and 1400-2600 mL, respectively. Early use of lactated Ringer's solution, which inhibited NLRP3, was associated with reduced mortality in hypotensive phenotype. These findings underscore the clinical relevance of early physiological trajectories and support precision fluid resuscitation based on subtype. This study establishes the largest early-trajectory-based classification of AP to date, offering new insights into immune and vascular mechanisms that drive heterogeneity and therapeutic responsiveness.

Methods for identifying health status from routinely collected health data: An overview

Routinely collected health data (RCD) are currently accelerating publications that evaluate the effectiveness and safety of medicines and medical devices. One of the fundamental steps in using these data is developing algorithms to identify health status that can be used for observational studies. However, the process and methodologies for identifying health status from RCD remain insufficiently understood. While most current methods rely on International Classification of Diseases (ICD) codes, they may not be universally applicable. Although machine learning methods hold promise for more accurately identifying the health status, they remain underutilized in RCD studies. To address these significant methodological gaps, we outline key steps and methodological considerations for identifying health statuses in observational studies using RCD. This review has the potential to boost the credibility of findings from observational studies that use RCD.

Effect of Tanreqing injection on multidrug resistance organisms: A test-negative case-control study and network pharmacology analysis

BACKGROUND

Multidrug resistance organisms (MDROs) pose a major threat in intensive care units (ICUs). Although in vitro studies suggested that Tanreqing (TRQ) was effective against MDROs, evidence about TRQ injection usage and its real-world effectiveness is lacking.

PURPOSE

This study aimed to investigate treatment pattern and real-world effectiveness of TRQ against MDRO infections among ICU patients being treated with antibiotics.

STUDY DESIGN

A real-world data study (i.e., test-negative case control) was conducted, using a large validated multicenter ICU database. Eligible cases were patients infected with any of the six monitored MDROs, including methicillin resistant Staphylococcus aureus (MRSA), carbapenem-resistant Acinetobacter baumannii (CRAB), vancomycin-resistant Enterococcus sp. (VRE), carbapenems-resistant Pseudomonas aeruginosa (CRPA), carbapenem-resistant Enterobacter sp. (CRE), or carbapenem-resistant Klebsiella pneumoniae (CRKP). The controls were individuals infected with antibiotic-sensitive strains.

METHODS

We used marginal structural models to adjust for time-varying confounding. We also performed network pharmacology analysis to explore the mechanisms by which TRQ exerted effects against MRDOs.

RESULTS

A total of 2890 patients were included. There were significant variations in timing and duration of use of TRQ injection. Over half (54.4 %) of patients received antibiotics plus TRQ injection, and the duration ranged from 1 to 83 days. The addition of TRQ injection was associated with lower probability of that patients become infected with CRE (adjusted odds ratio [ORadj] 0.51; 95 % confidence interval [CI]: 0.35-0.74) and CRKP (ORadj 0.55; 95 % CI: 0.36-0.83). Network pharmacology analysis suggested that TRQ exerts the effect against CRKP by modulating the metabolic pathways of K. pneumoniae and inhibit β-lactamase enzyme. No statistically significant differences were observed between TRQ infection with MDROs (ORadj 1.12; 95 % CI: 0.90-1.38), MRSA (ORadj 1.13; 95 % CI: 0.50-2.54), CRPA (ORadj 0.79; 95 % CI: 0.52-1.20) and CRAB (ORadj 1.36; 95 % CI: 0.67-2.76).

CONCLUSION

TRQ injection was associated with lower CRKP infection risk in ICU patients, potentially via modulation of β-lactam antibiotic resistance and metabolic pathways.

Early deep-to-light sedation versus continuous light sedation for ICU patients with mechanical ventilation: A cohort study

BACKGROUND

Sedation strategies have not been well established for patients being treated with invasive mechanical ventilation (MV). This study aimed to compare the potential effects of alternative sedation strategies - including early deep-to-light sedation (DTLS), continuous deep sedation (CDS) and continuous light sedation (CLS, the currently recommended strategy) - on ventilator, intensive care unit (ICU) or hospital mortality.

METHODS

A cohort study was conducted using two large validated ICU databases, including the Registry of Healthcare-associated Infections in ICUs in China (ICU-HAI) and the Medical Information Mart for Intensive Care (MIMIC). Patients who received MV for more than 3 days with one of three sedation strategies were included. Multivariable survival analyses with inverse probability-weighted competing risk models were conducted separately for ICU-HAI and MIMIC cohorts. Adjusted estimates were pooled using fixed-effects models.

RESULTS

In total, 6700 patients (2627 ICU-HAI, 4073 MIMIC) were included in the cohort study, of whom 2689 received CLS, 2079 CDS and 1932 DTLS. Compared to CLS, DTLS was associated with lower ICU mortality (9.3% vs. 11.0%; pooled adjusted HR 0.78, 95% CI 0.66-0.94) and hospital mortality (16.0% vs. 14.1%; 0.86, CI 0.74-1.00); and CDS was associated with higher ventilator mortality (32.8% vs. 7.0%; 4.65, 3.91-5.53), ICU mortality (40.6% vs. 11.0%; 3.39, 2.95-3.90) and hospital mortality (46.8% vs. 14.1%; 3.27, 2.89-3.71) than CLS. All HRs were qualitatively consistent in both cohorts.

CONCLUSIONS

Compared to the continuous light sedation, early deep-to-light sedation strategy was associated with improved patient outcomes, and continuous deep sedation was confirmed with poorer patient outcomes.

Guidance of development, validation, and evaluation of algorithms for populating health status in observational studies of routinely collected data (DEVELOP-RCD)

BACKGROUND

In recent years, there has been a growing trend in the utilization of observational studies that make use of routinely collected healthcare data (RCD). These studies rely on algorithms to identify specific health conditions (e.g. diabetes or sepsis) for statistical analyses. However, there has been substantial variation in the algorithm development and validation, leading to frequently suboptimal performance and posing a significant threat to the validity of study findings. Unfortunately, these issues are often overlooked.

METHODS

We systematically developed guidance for the development, validation, and evaluation of algorithms designed to identify health status (DEVELOP-RCD). Our initial efforts involved conducting both a narrative review and a systematic review of published studies on the concepts and methodological issues related to algorithm development, validation, and evaluation. Subsequently, we conducted an empirical study on an algorithm for identifying sepsis. Based on these findings, we formulated specific workflow and recommendations for algorithm development, validation, and evaluation within the guidance. Finally, the guidance underwent independent review by a panel of 20 external experts who then convened a consensus meeting to finalize it.

RESULTS

A standardized workflow for algorithm development, validation, and evaluation was established. Guided by specific health status considerations, the workflow comprises four integrated steps: assessing an existing algorithm's suitability for the target health status; developing a new algorithm using recommended methods; validating the algorithm using prescribed performance measures; and evaluating the impact of the algorithm on study results. Additionally, 13 good practice recommendations were formulated with detailed explanations. Furthermore, a practical study on sepsis identification was included to demonstrate the application of this guidance.

CONCLUSIONS

The establishment of guidance is intended to aid researchers and clinicians in the appropriate and accurate development and application of algorithms for identifying health status from RCD. This guidance has the potential to enhance the credibility of findings from observational studies involving RCD.

Blood Urine Nitrogen Trajectories of Acute Pancreatitis Patients in Intensive Care Units

Objective

To identify subclasses of acute pancreatitis (AP) patients in the intensive care unit (ICU) by analyzing blood urea nitrogen (BUN) trajectories.

Methods

AP patients in West China Hospital System (development cohort) and three public databases in the United States (validation cohort) were included. Latent class trajectory modelling was used to identify subclasses based on BUN trajectories within the first 21 days after ICU admission. Clinical characteristics and outcomes were compared, and results were externally validated.

Results

The study comprised 2971 and 930 patients in the development and validation cohorts, respectively, with five subclasses: Class 1 ("Moderate-azotemia, slow decreasing"), Class 2 ("Non-azotemia"), Class 3 ("Severe-azotemia, slow decreasing"), Class 4 ("Moderate-azotemia, rapid increasing"), and Class 5 ('Moderate-azotemia, slow increasing) identified. Azotemia patients showed significantly higher 30-day mortality risk in development and validation cohorts. Specifically, Class 4 patients exhibited notably highest mortality risk in both the development cohort (HR 5.32, 95% CI 2.62-10.82) and validation cohort (HR 6.23, 95% CI 2.93-13.22). Regarding clinical characteristics, AP patients in Class 4 showed lower mean arterial pressure and a higher proportion of renal disease. We also created an online early classification model to further identify Class 4 patients among all patients with moderate azotemia at baseline.

Conclusion

This multinational study uncovers heterogeneity in BUN trajectories among AP patients. Patients with "Moderate-azotemia, rapid increasing" trajectory, had a higher mortality risk than patients with severe azotemia at baseline. This finding complements studies that solely rely on baseline BUN for risk stratification and enhanced our understanding of longitudinal progression of AP.

Characterization and trajectories of hematological parameters prior to severe COVID-19 based on a large-scale prospective health checkup cohort in western China: a longitudinal study of 13-year follow-up

BACKGROUND

The relaxation of the "zero-COVID" policy on Dec. 7, 2022, in China posed a major public health threat recently. Complete blood count test was discovered to have complicated relationships with COVID-19 after the infection, while very few studies could track long-term monitoring of the health status and identify the characterization of hematological parameters prior to COVID-19.

METHODS

Based on a 13-year longitudinal prospective health checkup cohort of ~ 480,000 participants in West China Hospital, the largest medical center in western China, we documented 998 participants with a laboratory-confirmed diagnosis of COVID-19 during the 1 month after the policy. We performed a time-to-event analysis to explore the associations of severe COVID-19 patients diagnosed, with 34 different hematological parameters at the baseline level prior to COVID-19, including the whole and the subtypes of white and red blood cells.

RESULTS

A total of 998 participants with a positive SARS-CoV-2 test were documented in the cohort, 42 of which were severe cases. For white blood cell-related parameters, a higher level of basophil percentage (HR = 6.164, 95% CI = 2.066-18.393, P = 0.001) and monocyte percentage (HR = 1.283, 95% CI = 1.046-1.573, P = 0.017) were found associated with the severe COVID-19. For lymphocyte-related parameters, a lower level of lymphocyte count (HR = 0.571, 95% CI = 0.341-0.955, P = 0.033), and a higher CD4/CD8 ratio (HR = 2.473, 95% CI = 1.009-6.059, P = 0.048) were found related to the risk of severe COVID-19. We also observed that abnormality of red cell distribution width (RDW), mean corpuscular hemoglobin concentration (MCHC), and hemoglobin might also be involved in the development of severe COVID-19. The different trajectory patterns of RDW-SD and white blood cell count, including lymphocyte and neutrophil, prior to the infection were also discovered to have significant associations with the risk of severe COVID-19 (all P < 0.05).

CONCLUSIONS

Our findings might help decision-makers and clinicians to classify different risk groups of population due to outbreaks including COVID-19. They could not only optimize the allocation of medical resources, but also help them be more proactive instead of reactive to long COVID-19 or even other outbreaks in the future.

Effects of Tanreqing Injection on ICU Mortality among ICU Patients Receiving Mechanical Ventilation: Time-Dependent Cox Regression Analysis of A Large Registry

OBJECTIVE

To assess whether the use of Tanreqing (TRQ) Injection could show improvements in time to extubation, intensive care unit (ICU) mortality, ventilator-associated events (VAEs) and infection-related ventilator associated complication (IVAC) among patients receiving mechanical ventilation (MV).

METHODS

A time-dependent cox-regression analysis was conducted using data from a well-established registry of healthcare-associated infections at ICUs in China. Patients receiving continuous MV for 3 days or more were included. A time-varying exposure definition was used for TRQ Injection, which were recorded on daily basis. The outcomes included time to extubation, ICU mortality, VAEs and IVAC. Time-dependent Cox models were used to compare the clinical outcomes between TRQ Injection and non-use, after controlling for the influence of comorbidities/conditions and other medications with both fixed and time-varying covariates. For the analyses of time to extubation and ICU mortality, Fine-Gray competing risk models were also used to measure competing risks and outcomes of interest.

RESULTS

Overall, 7,685 patients were included for the analyses of MV duration, and 7,273 patients for the analysis of ICU mortality. Compared to non-use, patients with TRQ Injection had a lower risk of ICU mortality (Hazards ratios (HR) 0.761, 95% CI, 0.581-0.997), and was associated with a higher hazard for time to extubation (HR 1.105, 95% CI, 1.005-1.216), suggesting a beneficial effect on shortened time to extubation. No significant differences were observed between TRQ Injection and non-use regarding VAEs (HR 1.057, 95% CI, 0.912-1.225) and IVAC (HR 1.177, 95% CI, 0.929-1.491). The effect estimates were robust when using alternative statistic models, applying alternative inclusion and exclusion criteria, and handling missing data by alternative approaches.

CONCLUSION

Our findings suggested that the use of TRQ Injection might lower mortality and improve time to extubation among patients receiving MV, even after controlling for the factor that the use of TRQ changed over time.

Evaluation of healthcare-associated infections in general intensive care unit in Meram State Hospital

Aim: Healthcare-associated infections are a major source of concern in all areas of hospitals, particularly in intensive care units. The goal of our study was to look at the current situation and evaluate the measures that can be taken based on the data obtained by examining the rates and factors of healthcare-associated infections in the general intensive care units of our hospital over a one-year period. Materials and Methods: Between January 2020 and December 2020, 665 patients who were followed up and treated in the general intensive care unit of Meram State Hospital were followed up in terms of healthcare-associated infections, and their outcomes were evaluated. Results: 5354 hospitalization days of 665 patients who were followed up in the general intensive care units for a year were evaluated, and it was determined that 53 of the patients developed healthcare-associated infections. Twenty-two (41.5%) of patients with healthcare-associated infections were female, while 31 (58.5%) were male. It was discovered that the patients' mean age was 71,7±14 (19-94). The infection rate was calculated to be 5.86 and the density to be 7.28. Furthermore, the rates of invasive device-associated nosocomial infection are as follows: 1.02 for central line-associated bloodstream infections, 0.56 for catheter-associated urinary tract infections, and 0 for ventilator-associated pneumonia. Conclusion: Healthcare-associated infections are a significant cause of mortality and morbidity in intensive care units. Due to the improvement in medical care and the increase in life expectancy in parallel with this, effective surveillance practices are of critical importance.

Associations Between Antithrombosis and Ventilator-Associated Events, ICU Stays, and Mortality Among Mechanically Ventilated Patients: A Registry-Based Cohort Study

Background: The effect of thromboembolism prophylaxis on clinical outcomes, such as ventilator-associated events (VAEs), ICU stays, and mortality, remains controversial. This study was conducted to evaluate the effect of pharmacological thromboprophylaxis on VAEs, ICU stays, and ICU mortality among patients receiving mechanical ventilation (MV).

Materials and Methods: A retrospective cohort study was conducted based on a well-established registry of healthcare-associated infections at ICUs in the West China Hospital system. Patients who consistently received MV for at least 4 days from 1 April 2015 to 31 December 2018 were included. Hazard ratios (HRs) were compared for three tiers of VAEs, ICU stays, and ICU mortality among patients receiving pharmacological thromboprophylaxis versus those without using the time-dependent Cox model. For the analyses of ICU stays and ICU mortality, we also used Fine-Gray models to disentangle the competing risks and outcomes of interest.

Results: Overall, 6,140 patients were included. Of these, 3,805 received at least one prescription of antithrombosis agents. Treatments with antithrombosis agents were associated with lower risk of VAEs (HR: 0.87, 95% CI: 0.77, 0.98) and ICU mortality (HR: 0.72, 95% CI: 0.61, 0.86) than those without. Anticoagulants but not antiplatelet agents were associated with decreased risk of VAEs (HR: 0.86, 95% CI: 0.75, 0.98), ICU mortality (HR: 0.62, 95% CI: 0.51, 0.76), and less time to ICU discharge (HR: 1.15, 95% CI: 1.04, 1.28). Antithrombosis may be associated with decreased risk of VAEs in patients with D-dimer >5 mg/LFEU (HR: 0.84, 95%CI: 0.72, 0.98).

Conclusions: Pharmacological thromboprophylaxis was associated with lower risk of VAEs and ICU mortality. Similar effects were observed between unfractionated heparins versus low-molecular-weight heparins.

Associations of Fentanyl, Sufentanil, and Remifentanil With Length of Stay and Mortality Among Mechanically Ventilated Patients: A Registry-Based Cohort Study

Background: As the first-line treatment for mechanically ventilated patients with critical illness, fentanyl and its analogs (e.g., sufentanil and remifentanil) are commonly used in the intensive care unit (ICU). However, the pharmacokinetics, metabolism, and potency of these agents differed. Their effects on clinical outcomes have not been well-understood.

Materials and Methods: Using a well-established registry, we conducted a cohort study. Patients who consistently underwent mechanical ventilation (MV) for more than 24 h were identified. We used a time-varying exposure definition, in which we coded each type of opioids as prescribed or not prescribed on each day from initiation of MV to extubation and ICU discharge. We used Fine-Gray competing risk models to compare the effects of fentanyl, sufentanil, and remifentanil on hazards for extubation, ventilator mortality, ICU discharge, and ICU mortality. All models were adjusted using a combination of fixed-time and time-varying covariates. Missing data were imputed using multiple imputation by chained equations.

Results: A total of 8,165 patients were included. There were, respectively, 4,778, 4,008, and 2,233 patients receiving at least 1 day of fentanyl, sufentanil, and remifentanil dose. Compared to fentanyl, sufentanil was associated with shorter duration to extubation (hazard ratio 1.31, 95% CI, 1.20–1.41) and ICU discharge (hazard ratio 1.63, 95% CI, 1.38–1.92), and remifentanil was associated with shorter duration to extubation (hazard ratio 1.60, 95% CI, 1.40–1.84) and ICU discharge (hazard ratio 2.02, 95% CI, 1.43–2.84). No significant differences in time to extubation (Hazard ratio 1.14, 95% CI, 0.92–1.41) and ICU discharge (Hazard ratio 1.31, 95% CI, 0.81–2.14) were found between sufentanil and remifentanil. No differences were observed between any two of the agents regarding ventilator mortality or ICU mortality. The effects were similar in patients with versus without surgery.

Conclusion: Sufentanil and remifentanil may be superior to fentanyl in shortening the time to extubation and ICU discharge. The effects on ventilator mortality and ICU mortality appeared similar across these agents, while further research is warranted.

Utilization of traditional Chinese medicine in the intensive care unit

Previous studies showed that traditional Chinese medicine (TCM) may be effective for patients with critical conditions. However, the extent to which TCM is used in intensive care units (ICUs) remains less investigated. This study aimed to investigate the utilization of TCM among ICU patients. Using a cross-sectional study design, we examined the use of TCMs among ICU patients. The data were from a well-established ICU registry from a large teaching hospital in west China. Our study found that TCM was widely used among ICU, in particular TCM injections and oral liquids. The use of TCM often differed by ICUs, and TCM injections and oral liquids were mostly used in neurological ICU, while Chinese herbal medicines (CHMs) were often used in general ICU. The use of TCM was also highly associated with patient characteristics. Patients with pancreatitis were administered with most TCM, patients with cerebrovascular disease with most TCM injections and those with chronic renal failure with most oral liquids. Future efforts should include generation of high-quality evidence guidelines about clinical effects of TCM interventions among ICU patients.

Fluid Balance and Ventilator-Associated Events Among Patients Admitted to ICUs in China: A Nested Case-Control Study

Supplemental Digital Content is available in the text.

Fluid therapy is an important component of intensive care management, however, optimal fluid management is unknown. The relationship between fluid balance and ventilator-associated events has not been well established. This study investigated the dose-response relationship between fluid balance and ventilator-associated events.

Developing a research database of primary aldosteronism: rationale and baseline characteristics

BACKGROUND

Management of primary aldosteronism (PA) has become a research hotspot in the field of endocrinology. To obtain reliable research evidence, it is necessary to establish a high-quality PA research database.

METHODS

The establishment of PA research database involved two steps. Firstly, patients with confirmation of PA diagnosis between 1 Jan 2009 to 31 Aug 2019 at West China Hospital were identified and data were extracted. Secondly, patients with confirmatory testing for PA will be enrolled into a prospective cohort. Data will be prospectively collected based on the case report forms since 1 Sep 2019. We evaluated the quality of research database through assessment of quality of key variables.

RESULTS

Totally, 862 patients diagnosed as PA were identified, of which 507 patients who had positive confirmatory testing for PA were included into the retrospective database. Among 862 patients diagnosed as PA, the mean systolic blood pressure (SBP) was 156.1 (21.7) mmHg, mean diastolic blood pressure (DBP) was 97.2 (14.5) mmHg. Among included patients, the mean serum potassium level was 2.85 (IQR, (2.47-3.36) mmol/L, and the mean plasma aldosterone concentration (PAC) was 28.1 (IQR, 20.0-40.4) ng/dL. The characteristics of patients with positive confirmatory testing for PA were similar. Validation of data extracting and linking showed the accuracy were 100%. Evaluation of missing data showed that the completeness of BMI (95.9%), SBP (99.4%) and DBP (99.4%) were high.

CONCLUSION

Through integrating retrospective and prospective cohort of PA, a research database of PA with high quality and comprehensive data can be established. We anticipate that the research database will provide a high level of feasibility for management of PA in China.

Association between blood transfusion and ventilator-associated events: a nested case-control study

OBJECTIVES

The association between blood transfusion and ventilator-associated events (VAEs) has not been fully understood. We sought to determine whether blood transfusion increases the risk of a VAE.

DESIGN

Nested case-control study.

SETTING

This study was based on a registry of healthcare-associated infections in intensive care units at West China Hospital system.

PATIENTS

1,657 VAE cases and 3,293 matched controls were identified.

METHODS

For each case, 2 controls were randomly selected using incidence density sampling. We defined blood transfusion as a time-dependent variable, and we used weighted Cox models to calculate hazard ratios (HRs) for all 3 tiers of VAEs.

RESULTS

Blood transfusion was associated with increased risk of ventilator-associated complication-plus (VAC-plus; HR, 1.47; 95% CI, 1.22-1.77; P <.001), VAC-only (HR, 1.29; 95% CI, 1.01-1.65; P = .038), infection-related VAC-plus (IVAC-plus; HR, 1.78; 95% CI, 1.33-2.39; P < .001), and possible ventilator-associated pneumonia (PVAP; HR, 2.10; 95% CI, 1.10-3.99; P = .024). Red blood cell (RBC) transfusion was also associated with increased risk of VAC-plus (HR, 1.34; 95% CI, 1.08-1.65; P = .007), IVAC-plus (HR, 1.70; 95% CI, 1.22-2.36; P = .002), and PVAP (HR, 2.49; 95% CI, 1.17-5.28; P = .018). Compared to patients without transfusion, the risk of VAE was significantly higher in patients with RBC transfusions of >3 units (HR, 1.73; 95% CI, 1.25-2.40; P = .001) but not in those with RBC transfusions of 0-3 units.

CONCLUSION

Blood transfusions were associated with increased risk of all tiers of VAE. The risk was significantly higher among patients who were transfused with >3 units of RBCs.

Clinical outcomes and risk factors for mortality from ventilator-associated events: A registry-based cohort study among 30,830 intensive care unit patients

OBJECTIVE

To investigate the clinical impact of ventilator-associated events (VAEs) on adverse prognoses and risk factors for mortality among intensive care unit (ICU) patients receiving invasive mechanical ventilation (IMV) based on an ICU healthcare-associated infection (ICU-HAI) registry.

DESIGN

A cohort study was conducted based on an ICU-HAI registry including 30,830 patients between 2015 and 2018.

SETTING

The study was conducted using data from 5 adult ICUs of a referral hospital.

PATIENTS

Adult patients in the ICU-HAI registry who received ≥4 consecutive IMV days.

METHODS

Clinical outcomes and mortality risk factors for VAEs were analyzed using propensity score matching (PSM), multivariate regression models, and sensitivity analyses.

RESULTS

Of 6,426 included patients, 1,803 developed 1,899 VAEs. After PSM, patients with VAEs did have prolonged length of stay in the ICU and in the hospital, increased hospitalization costs, longer days on mechanical ventilation, higher proportion of ≥9 days on mechanical ventilation, higher rate of failure in extubating mechanical ventilation, and excess all-cause mortality in the ICU. Older age (adjusted OR [aOR], 1.02), higher APACHE II score on ICU admission (aOR, 1.06), pneumonia (aOR, 1.49), blood transfusion (aOR 1.43), immunosuppressive drugs (aOR, 1.69), central-line catheter (aOR, 2.06), and ≥2 VAEs in the ICU (aOR, 1.99) were associated with higher risks for all-cause mortality in an ICU.

CONCLUSIONS

Patients with VAEs indeed had poorer clinical outcomes. Older age, higher APACHE II score on ICU admission, pneumonia, blood transfusion, immunosuppressive drugs, central-line catheter, and ≥2 VAEs in the ICU were risk factors for all-cause mortality of VAE patients in the ICU.

Early Prediction of Seven-Day Mortality in Intensive Care Unit Using a Machine Learning Model: Results from the SPIN-UTI Project

Patients in intensive care units (ICUs) were at higher risk of worsen prognosis and mortality. Here, we aimed to evaluate the ability of the Simplified Acute Physiology Score (SAPS II) to predict the risk of 7-day mortality, and to test a machine learning algorithm which combines the SAPS II with additional patients' characteristics at ICU admission. We used data from the "Italian Nosocomial Infections Surveillance in Intensive Care Units" network. Support Vector Machines (SVM) algorithm was used to classify 3782 patients according to sex, patient's origin, type of ICU admission, non-surgical treatment for acute coronary disease, surgical intervention, SAPS II, presence of invasive devices, trauma, impaired immunity, antibiotic therapy and onset of HAI. The accuracy of SAPS II for predicting patients who died from those who did not was 69.3%, with an Area Under the Curve (AUC) of 0.678. Using the SVM algorithm, instead, we achieved an accuracy of 83.5% and AUC of 0.896. Notably, SAPS II was the variable that weighted more on the model and its removal resulted in an AUC of 0.653 and an accuracy of 68.4%. Overall, these findings suggest the present SVM model as a useful tool to early predict patients at higher risk of death at ICU admission.

The epidemiology and clinical outcomes of ventilator-associated events among 20,769 mechanically ventilated patients at intensive care units: an observational study

Background

Ventilator-associated pneumonia (VAP) is the most common hospital-acquired infection (HAI) in intensive care units (ICUs). Ventilator-associated event (VAE), a more objective definition, has replaced traditional VAP surveillance and is now widely used in the USA. However, the adoption outside the USA is limited. This study aims to describe the epidemiology and clinical outcomes of VAEs in China, based on a prospectively maintained registry.

Methods

An observational study was conducted using an ICU-HAI registry in west China. Patients that were admitted to ICUs and underwent mechanical ventilation (MV) between April 1, 2015, and December 31, 2018, were included. The characteristics and outcomes were compared between patients with and without VAEs. The rates of all VAEs dependent on different ICUs were calculated, and the pathogen distribution of patients with possible VAP (PVAP) was described.

Results

A total of 20,769 ICU patients received MV, accounting for 21,723 episodes of mechanical ventilators and 112,697 ventilator-days. In all, we identified 1882 episodes of ventilator-associated condition (VAC) events (16.7 per 1000 ventilator-days), 721 episodes of infection-related ventilator-associated complications (IVAC) events (6.4 per 1000 ventilator-days), and 185 episodes of PVAP events (1.64 per 1000 ventilator-days). The rates of VAC varied across ICUs with the highest incidence in surgical ICUs (23.72 per 1000 ventilator-days). The median time from the start of ventilation to the onset of the first VAC, IVAC, and PVAP was 5 (3–8), 5 (3–9), and 6 (4–13) days, respectively. The median length of hospital stays was 28.00 (17.00–43.00), 30.00 (19.00–44.00), and 30.00 (21.00–46.00) days for the three VAE tiers, which were all longer than that of patients without VAEs (16.00 [12.00–23.00]). The hospital mortality among patients with VAEs was more than three times of those with non-VAEs.

Conclusions

VAE was common in ICU patients with ≥ 4 ventilator days. All tiers of VAEs were highly correlated with poor clinical outcomes, including longer ICU and hospital stays and increased risk of mortality. These findings highlight the importance of VAE surveillance and the development of new strategies to prevent VAEs.