Multilevel competing risk models to evaluate the risk of nosocomial infection

An agent-based model on antimicrobial de-escalation in intensive care units: Implications on clinical trial design

Antimicrobial de-escalation refers to reducing the spectrum of antibiotics used in treating bacterial infections. This strategy is widely recommended in many antimicrobial stewardship programs and is believed to reduce patients' exposure to broad-spectrum antibiotics and prevent resistance. However, the ecological benefits of de-escalation have not been universally observed in clinical studies. This paper conducts computer simulations to assess the ecological effects of de-escalation on the resistance prevalence of Pseudomonas aeruginosa-a frequent pathogen causing nosocomial infections. Synthetic data produced by the models are then used to estimate the sample size and study period needed to observe the predicted effects in clinical trials. Our results show that de-escalation can reduce colonization and infections caused by bacterial strains resistant to the empiric antibiotic, limit the use of broad-spectrum antibiotics, and avoid inappropriate empiric therapies. Further, we show that de-escalation could reduce the overall super-infection incidence, and this benefit becomes more evident under good compliance with hand hygiene protocols among health care workers. Finally, we find that any clinical study aiming to observe the essential effects of de-escalation should involve at least ten arms and last for four years-a size never attained in prior studies. This study explains the controversial findings of de-escalation in previous clinical studies and illustrates how mathematical models can inform outcome expectations and guide the design of clinical studies.

A stochastic dynamical model for nosocomial infections with co-circulation of sensitive and resistant bacterial strains

Nosocomial infections (hospital-acquired) has been an important public health problem, which may make those patients with infections or involved visitors and hospital personnel at higher risk of worse clinical outcomes or infection, and then consume more healthcare resources. Taking into account the stochasticity of the death and discharge rate of patients staying in hospitals, in this paper, we propose a stochastic dynamical model describing the transmission of nosocomial pathogens among patients admitted for hospital stays. The stochastic terms of the model are incorporated to capture the randomness arising from death and discharge processes of patients. Firstly, a sufficient condition is established for the stochastic extinction of disease. It shows that introducing randomness in the model will result in lower potential of nosocomial outbreaks. Further, we establish a threshold criterion on the existence of stationary distribution and ergodicity for any positive solution of the model. Particularly, the spectral radius form of stochastic threshold value is calculated in the special case. Moreover, the numerical simulations are conducted to both validate the theoretical results and investigate the effect of prevention and control strategies on the prevalence of nosocomial infection. We show that enhancing hygiene, targeting colonized and infected patients, improving antibiotic treatment accuracy, shortening treatment periods are all crucial factors to contain nosocomial infections.

SAPS2, APACHE2, SOFA, and Core-10-TISS upon admission as risk indicators for ICU-acquired infections: a retrospective cohort study

PURPOSE

Early identification of high-risk patients is an important component in improving infection prevention. The SAPS2, APACHE2, Core-10-TISS, and SOFA scores are already widely used to estimate mortality, morbidity and nursing workload, but this study evaluated their usefulness in assessing a patient's risk of ICU-acquired infection.

METHODS

We conducted a retrospective cohort study by analyzing all patient admissions to seven ICUs at Charité Berlin, Germany in 2017 and 2018. The four scores were documented by physicians on the day of admission. The infection control staff monitored daily whether the patients experienced lower respiratory tract infections (LRTIs), urinary tract infections (UTIs), or primary blood stream infections (PBSIs). For each combination of scoring system and infection type, an adjusted Fine and Gray model was fitted.

RESULTS

We analyzed 5053 ICU admissions and observed at least one ICU-acquired infection in N = 253 patients (incidence density: 4.73 per 1000 days). 59.0% (N = 2983) of the patients were male, median age was 66 years (IQR 55-77) and median length of stay was 6 days (IQR 4-12). All models showed that patients with a higher score value were at higher risk for ICU-acquired first PBSI, LRTI, or UTI, except for the model of APACHE2 and PBSI. Patients with a SAPS2 score of > 50 points showed an increased risk of infection of sHR = 2.34 for PBSIs (CI 1.06-5.17, p < 0.05), sHR = 2.33 for LRTIs (1.53-2.55, p < 0.001) and sHR = 2.25 for UTIs (1.23-4.13, p < 0.01) when compared to the reference group with 0-30 points.

CONCLUSIONS

The result of this study showed that admission scores of SAPS2, Core-10-TISS, APACHE2, and SOFA might be adequate indicators for assessing a patient's risk of ICU-acquired infection.

Clinical characteristics and risk factors associated with ICU-acquired infections in sepsis: A retrospective cohort study

Intensive care unit (ICU)-acquired infection is a common cause of poor prognosis of sepsis in the ICU. However, sepsis-associated ICU-acquired infections have not been fully characterized. The study aims to assess the risk factors and develop a model that predicts the risk of ICU-acquired infections in patients with sepsis.

Do Contact Precautions Reduce the Incidence of Intensive Care Unit-Acquired Pseudomonas aeruginosa Infections? The DPCPYO (Detection and Contact Precautions for Patients With P. aeruginosa) Cluster-Randomized Crossover Trial

BACKGROUND

The issue of contact precautions as contributory factors for reducing Pseudomonas aeruginosa (Pa) infections in intensive care units (ICUs) remains questioned. We evaluated the impact of the addition of contact precautions to standard precautions for Pa-positive patients on incidence of ICU-acquired Pa infections.

METHODS

In this multicenter, cluster-randomized crossover trial, 10 French ICUs were randomly assigned (1:1) to sequence 0-1 (6-month control period [CP]/3-month washout period/6-month intervention period [IP]) or sequence 1-0 (6-month IP/3-month washout period/6-month CP). A surveillance screening program for Pa was implemented. Competing-risks regression models were built with death and discharge without the occurrence of ICU-acquired Pa infection (the primary outcome) as competing events. Models were adjusted for within-ICU correlation and patient- and ICU-level covariates. The Simpson diversity index (SDI) and transmission index (TI) of Pa isolates were derived from pulsed-field gel electrophoresis typing.

RESULTS

Within recruited ICUs, the cumulative incidence and incidence rate of ICU-acquired Pa infections were 3.38% (55/1625) versus 3.44% (57/1658) and 3.31 versus 3.52 per 1000 patient-days at risk during the CP and IP, respectively. Multivariable models indicated that the intervention did not significantly change the cumulative incidence (subdistribution hazard ratio, .91; 95% confidence interval [CI], .49-1.67; P = .76) or rate (cause-specific hazard ratio, 1.36; 95% CI, .71-2.63; P = .36) of the primary outcome. SDI and TI did not significantly differ between CP and IP.

CONCLUSIONS

The addition of contact precautions to standard precautions for Pa-positive patients with a surveillance screening program does not significantly reduce ICU-acquired Pa infections in non-outbreak situations. Clinical Trials Registration. ISRCTN92710225.

Mortality review as a tool to assess the contribution of healthcare-associated infections to death: results of a multicentre validity and reproducibility study, 11 European Union countries, 2017 to 2018

IntroductionThe contribution of healthcare-associated infections (HAI) to mortality can be estimated using statistical methods, but mortality review (MR) is better suited for routine use in clinical settings. The European Centre for Disease Prevention and Control recently introduced MR into its HAI surveillance.AimWe evaluate validity and reproducibility of three MR measures.MethodsThe on-site investigator, usually an infection prevention and control doctor, and the clinician in charge of the patient independently reviewed records of deceased patients with bloodstream infection (BSI), pneumonia, Clostridioides difficile infection (CDI) or surgical site infection (SSI), and assessed the contribution to death using 3CAT: definitely/possibly/no contribution to death; WHOCAT: sole cause/part of causal sequence but not sufficient on its own/contributory cause but unrelated to condition causing death/no contribution, based on the World Health Organization's death certificate; QUANT: Likert scale: 0 (no contribution) to 10 (definitely cause of death). Inter-rater reliability was assessed with weighted kappa (wk) and intra-cluster correlation coefficient (ICC). Reviewers rated the fit of the measures.ResultsFrom 2017 to 2018, 24 hospitals (11 countries) recorded 291 cases: 87 BSI, 113 pneumonia , 71 CDI and 20 SSI. The inter-rater reliability was: 3CAT wk 0.68 (95% confidence interval (CI): 0.61-0.75); WHOCAT wk 0.65 (95% CI: 0.58-0.73); QUANT ICC 0.76 (95% CI: 0.71-0.81). Inter-rater reliability ranged from 0.72 for pneumonia to 0.52 for CDI. All three measures fitted 'reasonably' or 'well' in > 88%.ConclusionFeasibility, validity and reproducibility of these MR measures was acceptable for use in HAI surveillance.

The influence of individual and neighborhood-level characteristics on rural-urban disparities in cardiovascular disease mortality among U.S. women diagnosed with breast and gynecologic cancers

OBJECTIVE

Rural-urban disparities exist in cancer and cardiovascular disease (CVD) mortality. Investigations of CVD mortality among breast and gynecologic cancer (BGC) survivors from rural/urban communities are limited. We evaluated the influence of individual and neighborhood-level factors on rural-urban disparities in CVD mortality among BGC survivors.

METHODS

Data were from 1,139,767 women aged ≥20 years from the Surveillance, Epidemiology, and End Results program who were diagnosed with BGC from 2000 to 2016 that was merged with Area Health Resource Files for neighborhood-level factors (smoking, cancer screening, primary care provider density and socioeconomic index). Standardized mortality ratios (SMRs) for CVD mortality were calculated and multilevel Cox models, accounting for competing events, were used to estimate hazards ratios (HR) and 95% confidence intervals (CI).

RESULTS

The average age of BGC survivors was 60 years, with 10.3% of them living in rural counties. During a median follow-up of 5.1 years, 47,995 CVD deaths occured. Women with BGC had excess CVD mortality compared to general population women (SMR 6.05; CI: 6.00-6.11). This risk was highest among women aged <50 years (SMR = 27.16; CI: 25.74-28.62). In models adjusted for demographics, cancer stage and cancer therapy, women with BGC in rural communities had higher CVD deaths than those in urban communities (HR = 1.10, CI:1. 05-1.15). Additional adjustment for neighborhood-level characteristics attenuated the relation of rurality with CVD mortality (HR = 1.02, CI: 0.98-1.07).

CONCLUSIONS

BGC survivors living in rural communities have elevated risk of CVD mortality. Neighborhood-level characteristics explained the rural-urban disparities in CVD mortality observed among BGC survivors.

Protocol for a prospective cohort study: Prevention of Transmissions by Effective Colonisation Tracking in Neonates (PROTECT-Neo)

INTRODUCTION

Transmissions of opportunistic bacterial pathogens between neonates increase the risk of infections with negative repercussions, including higher mortality, morbidity and permanent disabilities. The probability of transmissions between patients is contingent on a set of intrinsic (patient-related) and extrinsic (ward-related) risk factors that are not clearly quantified. It is the dual objective of the Prevention of Transmissions by Effective Colonisation Tracking-Neo study to determine the density of transmission events in a level III neonatal intensive care unit (NICU) and to identify risk factors that may be causally associated with transmission events.

METHODS AND ANALYSIS

A full cohort of patients treated in a 17-bed level III NICU will be prospectively followed and transmission events between two or more patients will be documented. A transmission event occurs when isogenic isolates from two different patients can be identified. Isolates will be obtained by routine weekly screening. Isogenicity will be determined by whole-genome sequencing. During the study, relevant intrinsic and extrinsic risk factors will be recorded. Specimen and data will be collected for 1 year. We postulate that transmission density increases during episodes when demand for intensive care cannot be met by existing staff, and that threshold dynamics have a bearing on cohorting and hand hygiene performance. Poisson logistic regression, proportional hazard and multilevel competing risk models will be used to estimate the effect of explanatory variables.

ETHICS AND DISSEMINATION

This study has been approved by the local ethics committee (study ID 287/18). The results will be published in peer-reviewed medical journals, communicated to participants, the general public and all relevant stakeholders.

TRIAL REGISTRATION NUMBER

The German Clinical Trials Registry (DRKS00017733); Pre-results.

Individual participant data meta-analysis of intervention studies with time-to-event outcomes: A review of the methodology and an applied example

Many randomized trials evaluate an intervention effect on time-to-event outcomes. Individual participant data (IPD) from such trials can be obtained and combined in a so-called IPD meta-analysis (IPD-MA), to summarize the overall intervention effect. We performed a narrative literature review to provide an overview of methods for conducting an IPD-MA of randomized intervention studies with a time-to-event outcome. We focused on identifying good methodological practice for modeling frailty of trial participants across trials, modeling heterogeneity of intervention effects, choosing appropriate association measures, dealing with (trial differences in) censoring and follow-up times, and addressing time-varying intervention effects and effect modification (interactions).We discuss how to achieve this using parametric and semi-parametric methods, and describe how to implement these in a one-stage or two-stage IPD-MA framework. We recommend exploring heterogeneity of the effect(s) through interaction and non-linear effects. Random effects should be applied to account for residual heterogeneity of the intervention effect. We provide further recommendations, many of which specific to IPD-MA of time-to-event data from randomized trials examining an intervention effect.We illustrate several key methods in a real IPD-MA, where IPD of 1225 participants from 5 randomized clinical trials were combined to compare the effects of Carbamazepine and Valproate on the incidence of epileptic seizures.

Clinical characteristics, risk factors, immune status and prognosis of secondary infection of sepsis: a retrospective observational study

Background

Secondary infection has a higher incidence in septic patients and affects clinical outcomes. This study aims to investigate the clinical characteristics, risk factors, immune status and prognosis of secondary infection of sepsis.

Methods

A four-year retrospective study was carried out in Zhongshan Hospital, Fudan University, enrolling septic patients admitted between January, 2014 and January, 2018. Clinical data were acquired from medical records. CD14⁺ monocyte human leukocyte antigen-D related (HLA-DR) expression and serum cytokines levels were measured by flow cytometry and enzyme-linked immunosorbent assay (ELISA) respectively.

Results

A total of 297 septic patients were enrolled, 92 of whom developed 150 cases of secondary infections. Respiratory tract was the most common site of secondary infection (n = 84, 56%) and Acinetobacter baumanii the most commonly isolated pathogen (n = 40, 31%). Urinary and deep venous catheterization increased the risk of secondary infection. Lower HLA-DR expression and elevated IL-10 level were found in secondary infection group. The expected prolonged in-hospital stay owing to secondary infection was 4.63 ± 1.87 days. Secondary infection was also associated with higher in-hospital, 30-day and 90-day mortality. Kaplan-Meier survival analysis and Log-rank test revealed that secondary infection group had worse survival between day 15 and day 90.

Conclusions

Urinary and deep venous catheterization increased the risk of secondary infection, in which underlying immunosuppression might also play a role. Secondary infection affected the prognosis of septic patients and prolonged in-hospital length of stay.

Organisational and financial consequences of the early discharge of patients treated for acute bacterial skin and skin structure infection and osteomyelitis in infectious disease departments in Greece, Italy and Spain: a scenario analysis

OBJECTIVE

The aim of the analysis is to assess the organisational and economic consequences of adopting an early discharge strategy for the treatment of acute bacterial skin and skin structure infection (ABSSSI) and osteomyelitis within infectious disease departments.

SETTING

Infectious disease departments in Greece, Italy and Spain.

PARTICIPANTS

No patients were involved in the analysis performed.

INTERVENTIONS

An analytic framework was developed to consider two alternative scenarios: standard hospitalisation care or an early discharge strategy for patients hospitalised due to ABSSSI and osteomyelitis, from the perspective of the National Health Services of Greece, Italy and Spain. The variables considered were: the number of annual hospitalisations eligible for early discharge, the antibiotic treatments considered (ie, oral antibiotics and intravenous long-acting antibiotics), diagnosis-related group (DRG) reimbursements, number of days of hospitalisation, incidence and costs of hospital-acquired infections, additional follow-up visits and intravenous administrations. Data were based on published literature and expert opinions.

PRIMARY AND SECONDARY OUTCOME MEASURES

Number of days of hospitalisation avoided and direct medical costs avoided.

RESULTS

The total number of days of hospitalisation avoided on a yearly basis would be between 2216 and 5595 in Greece (-8/-21 hospital beds), between 15 848 and 38 444 in Italy (-57/-135 hospital beds) and between 7529 and 23 520 in Spain (-27/-85 hospital beds). From an economic perspective, the impact of the early discharge scenario is a reduction between €45 036 and €149 552 in Greece, a reduction between €182 132 and €437 990 in Italy and a reduction between €292 284 and €884 035 in Spain.

CONCLUSIONS

The early discharge strategy presented would have a positive organisational impact on National Health Services, leading to potential savings in beds, and to a reduction of hospital-acquired infections and costs.

Economic evaluation of the treatment of Acute Bacterial Skin and Skin Structure Infections (ABSSSIs) from the national payer perspective: introduction of a new treatment to the patient journey. A simulation of three European countries

Background: The aim of this study was to develop a spending predictor model to evaluate the direct costs associated with the management of ABSSSIs from the National health-care provider's perspective of Italy, Romania, and Spain. Methodology: A decision-analytic model was developed to evaluate the diagnostic and clinical pathways of hospitalized ABSSSI patients based on scientific guidelines and real-world data. A Standard of Care (SoC) scenario was compared with a dalbavancin scenario in which the patients could be discharged early. The epidemiological and cost parameters were extrapolated from national administrative databases (i.e., hospital information system). A probabilistic sensitivity analysis (PSA) and one-way sensitivity analysis (OWA) were performed. Results: Overall, the model estimated an average annual number of patients with ABSSSIs of approximately 50,000 in Italy, Spain, and Romania. On average, the introduction of dalbavancin reduced the length of stay by 3.3 days per ABSSSI patient. From an economic perspective, dalbavancin did not incur any additional cost from the National Healthcare perspective, and the results were consistent among the countries. The PSA and OWA demonstrated the robustness of these results. Conclusion: This model represents a useful tool for policymakers by providing information regarding the economic and organizational consequences of an early discharge approach in ABSSSI management.

Impact of a multifaceted prevention program on ventilator-associated pneumonia including selective oropharyngeal decontamination

PURPOSE

We describe the impact of a multifaceted program for decreasing ventilator-associated pneumonia (VAP) after implementing nine preventive measures, including selective oropharyngeal decontamination (SOD).

METHODS

We compared VAP rates during an 8-month pre-intervention period, a 12-month intervention period, and an 11-month post-intervention period in a cohort of patients who received mechanical ventilation (MV) for > 48 h. The primary objective was to assess the effect on first VAP occurrence, using a Cox cause-specific proportional hazards model. Secondary objectives included the impact on emergence of antimicrobial resistance, antibiotic consumption, duration of MV, and ICU mortality.

RESULTS

Pre-intervention, intervention and post-intervention VAP rates were 24.0, 11.0 and 3.9 VAP episodes per 1000 ventilation-days, respectively. VAP rates decreased by 56% [hazard ratio (HR) 0.44, 95% CI 0.29-0.65; P < 0.001] in the intervention and by 85% (HR 0.15, 95% CI 0.08-0.27; P < 0.001) in the post-intervention periods. During the intervention period, VAP rates decreased by 42% (HR 0.58, 95% CI 0.38-0.87; P < 0.001) after implementation of eight preventive measures without SOD, and by 70% after adding SOD (HR 0.30, 95% CI 0.13-0.72; P < 0.001) compared to the pre-intervention period. The incidence density of intrinsically resistant bacteria (to colistin or tobramycin) did not increase. We documented a significant reduction of days of therapy per 1000 patient-days of broad-spectrum antibiotic used to treat lower respiratory tract infection (P < 0.028), median duration of MV (from 7.1 to 6.4 days; P < 0.003) and ICU mortality (from 16.2 to 13.5%; P < 0.049) for patients ventilated > 48 h between the pre- and post-intervention periods.

CONCLUSIONS

Our preventive program produced a sustained decrease in VAP incidence. SOD provides an additive value.

Risk factors for health care-associated infection in hospitalized adults: Systematic review and meta-analysis

BACKGROUND

Health care-associated infections (HAIs) are a public health problem that increase health care costs. This article aimed to systematically review the literature and meta-analyze studies investigating risk factors (RFs) independently associated with HAIs in hospitalized adults.

METHODS

Electronic databases (MEDLINE, Embase, and LILACS) were searched to identify studies from 2009-2016. Pooled risk ratios (RRs) or odds ratios (ORs) or mean differences (MDs) and 95% confidence intervals (CIs) were calculated and compared across the groups. This review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement.

RESULTS

Of 867 studies, 65 met the criteria for review, and the data of 18 were summarized in the meta-analysis. The major RFs independently associated with HAIs were diabetes mellitus (RR, 1.76; 95% CI, 1.27-2.44), immunosuppression (RR, 1.24; 95% CI, 1.04-1.47), body temperature (MD, 0.62; 95% CI, 0.41-0.83), surgery time in minutes (MD, 34.53; 95% CI, 22.17-46.89), reoperation (RR, 7.94; 95% CI, 5.49-11.48), cephalosporin exposure (RR, 1.77; 95% CI, 1.30-2.42), days of exposure to central venous catheter (MD, 5.20; 95% CI, 4.91-5.48), intensive care unit (ICU) admission (RR, 3.76; 95% CI, 1.79-7.92), ICU stay in days (MD, 21.30; 95% CI, 19.81-22.79), and mechanical ventilation (OR, 12.95; 95% CI, 6.28-26.73).

CONCLUSIONS

Identifying RFs that contribute to develop HAIs may support the implementation of strategies for their prevention, therefore maximizing patient safety.

Multistate Modeling to Analyze Nosocomial Infection Data: An Introduction and Demonstration

OBJECTIVE Multistate and competing risks models have become an established and adequate tool with which to quantify determinants and consequences of nosocomial infections. In this tutorial article, we explain and demonstrate the basics of these models to a broader audience of professionals in health care, infection control, and hospital epidemiology. METHODS Using a publicly available data set from a cohort study of intensive care unit patients, we show how hospital infection data can be displayed and explored graphically and how simple formulas are derived under some simplified assumptions for illustrating the basic ideas behind multistate models. Only a few simply accessible values (event counts and patient days) and a pocket calculator are needed to reveal basic insights into cumulative risk and clinical outcomes of nosocomial infection in terms of mortality and length of stay. RESULTS We show how to use these values to perform basic multistate analyses in own data or to correct biased estimates in published data, as these values are often reported. We also show relationships between multistate-based hazard ratios and odds ratios, which are derived from the popular logistic regression model. CONCLUSIONS No sophisticated statistical software is required to apply a basic multistate model and to avoid typical pitfalls such as time-dependent or competing-risks bias. Infect Control Hosp Epidemiol 2017;38:953-959.

Benefits and unintended consequences of antimicrobial de-escalation: Implications for stewardship programs

Sequential antimicrobial de-escalation aims to minimize resistance to high-value broad-spectrum empiric antimicrobials by switching to alternative drugs when testing confirms susceptibility. Though widely practiced, the effects de-escalation are not well understood. Definitions of interventions and outcomes differ among studies. We use mathematical models of the transmission and evolution of Pseudomonas aeruginosa in an intensive care unit to assess the effect of de-escalation on a broad range of outcomes, and clarify expectations. In these models, de-escalation reduces the use of high-value drugs and preserves the effectiveness of empiric therapy, while also selecting for multidrug-resistant strains and leaving patients vulnerable to colonization and superinfection. The net effect of de-escalation in our models is to increase infection prevalence while also increasing the probability of effective treatment. Changes in mortality are small, and can be either positive or negative. The clinical significance of small changes in outcomes such as infection prevalence and death may exceed more easily detectable changes in drug use and resistance. Integrating harms and benefits into ranked outcomes for each patient may provide a way forward in the analysis of these tradeoffs. Our models provide a conceptual framework for the collection and interpretation of evidence needed to inform antimicrobial stewardship.

A case-cohort approach for multi-state models in hospital epidemiology

Analysing the determinants and consequences of hospital-acquired infections involves the evaluation of large cohorts. Infected patients in the cohort are often rare for specific pathogens, because most of the patients admitted to the hospital are discharged or die without such an infection. Death and discharge are competing events to acquiring an infection, because these individuals are no longer at risk of getting a hospital-acquired infection. Therefore, the data is best analysed with an extended survival model - the extended illness-death model. A common problem in cohort studies is the costly collection of covariate values. In order to provide efficient use of data from infected as well as uninfected patients, we propose a tailored case-cohort approach for the extended illness-death model. The basic idea of the case-cohort design is to only use a random sample of the full cohort, referred to as subcohort, and all cases, namely the infected patients. Thus, covariate values are only obtained for a small part of the full cohort. The method is based on existing and established methods and is used to perform regression analysis in adapted Cox proportional hazards models. We propose estimation of all cause-specific cumulative hazards and transition probabilities in an extended illness-death model based on case-cohort sampling. As an example, we apply the methodology to infection with a specific pathogen using a large cohort from Spanish hospital data. The obtained results of the case-cohort design are compared with the results in the full cohort to investigate the performance of the proposed method. Copyright © 2016 John Wiley & Sons, Ltd.

Multiple time scales in modeling the incidence of infections acquired in intensive care units

Background

When patients are admitted to an intensive care unit (ICU) their risk of getting an infection will be highly depend on the length of stay at-risk in the ICU. In addition, risk of infection is likely to vary over calendar time as a result of fluctuations in the prevalence of the pathogen on the ward. Hence risk of infection is expected to depend on two time scales (time in ICU and calendar time) as well as competing events (discharge or death) and their spatial location. The purpose of this paper is to develop and apply appropriate statistical models for the risk of ICU-acquired infection accounting for multiple time scales, competing risks and the spatial clustering of the data.

Methods

A multi-center data base from a Spanish surveillance network was used to study the occurrence of an infection due to Methicillin-resistant Staphylococcus aureus (MRSA). The analysis included 84,843 patient admissions between January 2006 and December 2011 from 81 ICUs. Stratified Cox models were used to study multiple time scales while accounting for spatial clustering of the data (patients within ICUs) and for death or discharge as competing events for MRSA infection.

Results

Both time scales, time in ICU and calendar time, are highly associated with the MRSA hazard rate and cumulative risk. When using only one basic time scale, the interpretation and magnitude of several patient-individual risk factors differed. Risk factors concerning the severity of illness were more pronounced when using only calendar time. These differences disappeared when using both time scales simultaneously.

Conclusions

The time-dependent dynamics of infections is complex and should be studied with models allowing for multiple time scales. For patient individual risk-factors we recommend stratified Cox regression models for competing events with ICU time as the basic time scale and calendar time as a covariate. The inclusion of calendar time and stratification by ICU allow to indirectly account for ICU-level effects such as local outbreaks or prevention interventions.

Electronic supplementary material

The online version of this article (doi:10.1186/s12874-016-0199-y) contains supplementary material, which is available to authorized users.

Impact of a targeted isolation strategy at intensive-care-unit-admission on intensive-care-unit-acquired infection related to multidrug-resistant bacteria: a prospective uncontrolled before-after study

Isolation of patients with multidrug resistant (MDR) bacteria is recommended to reduce cross-transmission of these bacteria. However, isolation of critically ill patients has several negative side effects. Therefore, we hypothesized that a targeted isolation strategy, based on the presence of at least one risk factor for MDR bacteria, would be not inferior to a systematic isolation strategy at intensive-care unit (ICU) admission. This prospective before-after study was conducted in a mixed ICU, during two 12-month periods, separated by a 1-month 'wash-out' period. During the before period, isolation was systematically performed in all patients at admission. During the after period, isolation was only performed in patients with at least one risk factor for MDR bacteria at admission. During the two periods, routine screening for MDR bacteria was performed at ICU admission, and isolation prescription was modified after receipt of screening result. Primary outcome was the percentage of patients with ICU-acquired infection (ICUAI) related to MDR bacteria, measured from ICU admission until ICU discharge or day 28, whatever happens first. A total of 1221 patients were included. No significant difference was found in ICUAI related to MDR bacteria (85 of 585 (14.5%) vs. 84 of 636 (13.2%) patients, risk difference, -1.3%, 95% confidence interval [-5.2 to 2.6%]) between the two periods, confirming the non-inferiority hypothesis. Our results suggest that targeted isolation of patients at ICU admission is not inferior to systematic isolation, regarding the percentage of patients with ICUAI related to MDR bacteria. Further randomized controlled multicentre studies are needed to confirm our results.

A full competing risk analysis of hospital-acquired infections can easily be performed by a case-cohort approach

OBJECTIVES

We provide a case-cohort approach and show that a full competing risk analysis is feasible even in a reduced data set. Competing events for hospital-acquired infections are death or discharge from the hospital because they preclude the observation of such infections.

STUDY DESIGN AND SETTING

Using surveillance data of 6,568 patient admissions (full cohort) from two Spanish intensive care units, we propose a case-cohort approach which uses only data from a random sample of the full cohort and all infected patients (the cases). We combine established methodology to study following measures: event-specific as well as subdistribution hazard ratios for all three events (infection, death, and discharge), cumulative hazards as well as incidence functions by risk factor, and also for all three events.

RESULTS

Compared with the values from the full cohort, all measures are well approximated with the case-cohort design. For the event of interest (infection), event-specific and subdistribution hazards can be estimated with the full efficiency of the case-cohort design. So, standard errors are only slightly increased, whereas the precision of estimated hazards of the competing events is inflated according to the size of the subcohort.

CONCLUSION

The case-cohort design provides an appropriate sampling design for studying hospital-acquired infections in a reduced data set. Potential effects of risk factors on the competing events (death and discharge) can be evaluated.

Multilevel competing risks in the evaluation of nosocomial infections: time to move on from proportional hazards and even from hazards altogether

A competing risk is an event (for example, death in the ICU) that hinders the occurrence of an event of interest (for example, nosocomial infection in the ICU) and it is a common issue in many critical care studies. Not accounting for a competing event may affect how results related to a primary event of interest are interpreted. In the previous issue of Critical Care, Wolkewitz and colleagues extended traditional models for competing risks to include random effects as a means to quantify heterogeneity among ICUs. Reported results from their analyses based on cause-specific hazards and on sub-hazards of the cumulative incidence function were indicative of lack of proportionality of these hazards over time. Here, we argue that proportionality of hazards can be problematic in competing-risk problems and analyses must consider time by covariate interactions as a default. Moreover, since hazards in competing risks make it difficult to disentangle the effects of frequency and timing of the competing events, their interpretation can be murky. Use of mixtures of flexible and succinct parametric time-to-event models for competing risks permits disentanglement of the frequency and timing at the price of requiring stronger data and a higher number of parameters. We used data from a clinical trial on fluid management strategies for patients with acute respiratory distress syndrome to support our recommendations.