A Data-Driven Framework for Identifying Intensive Care Unit Admissions Colonized With Multidrug-Resistant Organisms.

Background: The rising prevalence of multi-drug resistant organisms (MDROs), such as Methicillin-resistant Staphylococcus aureus (MRSA), Vancomycin-resistant Enterococci (VRE), and Carbapenem-resistant Enterobacteriaceae (CRE), is an increasing concern in healthcare settings. Materials and Methods: Leveraging data from electronic healthcare records and a unique MDRO universal screening program, we developed a data-driven modeling framework to predict MRSA, VRE, and CRE colonization upon intensive care unit (ICU) admission, and identified the associated socio-demographic and clinical factors using logistic regression (LR), random forest (RF), and XGBoost algorithms. We performed threshold optimization for converting predicted probabilities into binary predictions and identified the cut-off maximizing the sum of sensitivity and specificity. Results: Four thousand six hundred seventy ICU admissions (3,958 patients) were examined. MDRO colonization rate was 17.59% (13.03% VRE, 1.45% CRE, and 7.47% MRSA). Our study achieved the following sensitivity and specificity values with the best performing models, respectively: 80% and 66% for VRE with LR, 73% and 77% for CRE with XGBoost, 76% and 59% for MRSA with RF, and 82% and 83% for MDRO (i.e., VRE or CRE or MRSA) with RF. Further, we identified several predictors of MDRO colonization, including long-term care facility stay, current diagnosis of skin/subcutaneous tissue or infectious/parasitic disease, and recent isolation precaution procedures before ICU admission. Conclusion: Our data-driven modeling framework can be used as a clinical decision support tool for timely predictions, characterization and identification of high-risk patients, and selective and timely use of infection control measures in ICUs.

Carbapenem MICs in Escherichia coli and Klebsiella Species Producing Extended-Spectrum ß-Lactamases in Critical Care Patients from 2001 to 2009.

Extended-spectrum ß-lactamase (ESBL)-producing Enterobacteriaceae strains are increasing in prevalence worldwide. Carbapenem antibiotics are used as a first line of therapy against ESBL-producing Enterobacteriaceae We examined a cohort of critical care patients for gastrointestinal colonization with carbapenem-resistant ESBL-producing strains (CR-ESBL strains). We cultured samples from this cohort of patients for ESBL-producing Klebsiella spp. and Escherichia coli and then tested the first isolate from each patient for susceptibility to imipenem, doripenem, meropenem, and ertapenem. Multilocus sequence typing was performed on isolates that produced an ESBL and that were carbapenem resistant. Among all patients admitted to an intensive care unit (ICU), 4% were positive for an ESBL-producing isolate and 0.64% were positive for a CR-ESBL strain on surveillance culture. Among the first ESBL-producing E. coli and Klebsiella isolates from the patients' surveillance cultures, 11.2% were carbapenem resistant. Sequence type 14 (ST14), ST15, ST42, and ST258 were the dominant sequence types detected in this cohort of patients, with ST15 and ST258 steadily increasing in prevalence from 2006 to 2009. Patients colonized by a CR-ESBL strain were significantly more likely to receive antipseudomonal and anti-methicillin-resistant Staphylococcus aureus (anti-MRSA) therapy prior to ICU admission than patients colonized by carbapenem-susceptible ESBL-producing strains. They were also significantly more likely to have received a cephalosporin or a carbapenem antibiotic than patients colonized by carbapenem-susceptible ESBL-producing strains. In conclusion, in a cohort of patients residing in intensive care units within the United States, we found that 10% of the isolates were resistant to at least one carbapenem antibiotic. The continued emergence of carbapenem-resistant ESBL-producing strains is of significant concern, as infections due to these organisms are notoriously difficult to treat.

Assessment of the Overall and Multidrug-Resistant Organism Bioburden on Environmental Surfaces in Healthcare Facilities.

OBJECTIVE To determine the typical microbial bioburden (overall bacterial and multidrug-resistant organisms [MDROs]) on high-touch healthcare environmental surfaces after routine or terminal cleaning. DESIGN Prospective 2.5-year microbiological survey of large surface areas (>1,000 cm2). SETTING MDRO contact-precaution rooms from 9 acute-care hospitals and 2 long-term care facilities in 4 states. PARTICIPANTS Samples from 166 rooms (113 routine cleaned and 53 terminal cleaned rooms). METHODS Using a standard sponge-wipe sampling protocol, 2 composite samples were collected from each room; a third sample was collected from each Clostridium difficile room. Composite 1 included the TV remote, telephone, call button, and bed rails. Composite 2 included the room door handle, IV pole, and overbed table. Composite 3 included toileting surfaces. Total bacteria and MDROs (ie, methicillin-resistant Staphylococcus aureus, vancomycin-resistant enterococci [VRE], Acinetobacter baumannii, Klebsiella pneumoniae, and C. difficile) were quantified, confirmed, and tested for drug resistance. RESULTS The mean microbial bioburden and range from routine cleaned room composites were higher (2,700 colony-forming units [CFU]/100 cm2; ≤1-130,000 CFU/100 cm2) than from terminal cleaned room composites (353 CFU/100 cm2; ≤1-4,300 CFU/100 cm2). MDROs were recovered from 34% of routine cleaned room composites (range ≤1-13,000 CFU/100 cm2) and 17% of terminal cleaned room composites (≤1-524 CFU/100 cm2). MDROs were recovered from 40% of rooms; VRE was the most common (19%). CONCLUSIONS This multicenter bioburden summary provides a first step to determining microbial bioburden on healthcare surfaces, which may help provide a basis for developing standards to evaluate cleaning and disinfection as well as a framework for studies using an evidentiary hierarchy for environmental infection control. Infect Control Hosp Epidemiol 2016;1426-1432.

Accuracy of a radiofrequency identification (RFID) badge system to monitor hand hygiene behavior during routine clinical activities.

BACKGROUND: Hand hygiene (HH) is a critical part of infection prevention in health care settings. Hospitals around the world continuously struggle to improve health care personnel (HCP) HH compliance. The current gold standard for monitoring compliance is direct observation; however, this method is time-consuming and costly. One emerging area of interest involves automated systems for monitoring HH behavior such as radiofrequency identification (RFID) tracking systems. METHODS: To assess the accuracy of a commercially available RFID system in detecting HCP HH behavior, we compared direct observation with data collected by the RFID system in a simulated validation setting and to a real-life clinical setting over 2 hospitals. RESULTS: A total of 1,554 HH events was observed. Accuracy for identifying HH events was high in the simulated validation setting (88.5%) but relatively low in the real-life clinical setting (52.4%). This difference was significant (P < .01). Accuracy for detecting HCP movement into and out of patient rooms was also high in the simulated setting but not in the real-life clinical setting (100% on entry and exit in simulated setting vs 54.3% entry and 49.5% exit in real-life clinical setting, P < .01). CONCLUSION: In this validation study of an RFID system, almost half of the HH events were missed. More research is necessary to further develop these systems and improve accuracy prior to widespread adoption.

Using the theory of planned behavior to predict self-medication with over-the-counter analgesics.

Millions of people worldwide use over-the-counter analgesics on a regular basis; yet little is known about how decisions to self-medicate are made. This study used the theory of planned behavior to explore the influence of beliefs about medicines (Beliefs about Medicines Questionnaire) and individual pain experience as predictors of intent to self-medicate. Both emerged as significant predictors of intent to self-medicate. Furthermore, intent to self-medicate significantly predicted reported use of analgesics. These findings indicate that use of over-the-counter pain medication is more likely when the value of the pain relief is greater than concerns about harm.