A Machine Learning-Based Risk Score for Prediction of Infective Endocarditis Among Patients With Staphylococcus aureus Bacteremia-The SABIER Score.

BACKGROUND: Early risk assessment is needed to stratify Staphylococcus aureus infective endocarditis (SA-IE) risk among patients with S. aureus bacteremia (SAB) to guide clinical management. The objective of the current study was to develop a novel risk score that is independent of subjective clinical judgment and can be used early, at the time of blood culture positivity. METHODS: We conducted a retrospective big data analysis from territory-wide electronic data and included hospitalized patients with SAB between 2009 and 2019. We applied a random forest risk scoring model to select variables from an array of parameters, according to the statistical importance in predicting SA-IE outcome. The data were divided into derivation and validation cohorts. The areas under the curve of the receiver operating characteristic (AUCROCs) were determined. RESULTS: We identified 15 741 SAB patients, among them 658 (4.18%) had SA-IE. The AUCROC was 0.74 (95%CI 0.70-0.76), with a negative predictive value of 0.980 (95%CI 0.977-0.983). The four most discriminatory features were age, history of infective endocarditis, valvular heart disease, and community onset. CONCLUSIONS: We developed a novel risk score with performance comparable with existing scores, which can be used at the time of SAB and prior to subjective clinical judgment.

Complementary use of MALDI-TOF MS and real-time PCR-melt curve analysis for rapid identification of methicillin-resistant staphylococci and VRE.

OBJECTIVES: To develop a rapid method for routine screening of methicillin-resistant staphylococci and VRE for clinical isolates and positive blood cultures. METHODS: Our method consisted of two parts: MALDI-TOF MS was used for identification of staphylococci and enterococci, followed by antibiotic resistance detection by real-time PCR-melt curve analysis without DNA extraction. The latter part included a triplex reaction for staphylococcal culture isolates (mecA, mecALGA251 and Panton-Valentine leucocidin genes), dual PCR of mecA/mecALGA251 and nuc genes for staphylococcal blood cultures, and a duplex reaction for enterococci (vanA and vanB genes). A total of 124 clinical isolates and 56 positive blood cultures were tested. MALDI-TOF MS was performed using Microflex LT (Bruker Daltonik, Bremen, Germany) and Rotor-Gene Q (Qiagen, Hilden, Germany) was used for real-time PCR-melt curve analysis. The total assay time was <2.5 h. RESULTS: The results revealed 100% concordance with antibiotic susceptibility testing or other reference methods for all culture isolates and enterococcal blood cultures. The percentage of concordance for staphylococcal blood cultures was 97.5%. CONCLUSIONS: The method described herein was fast, economical, reliable and capable of detecting mecALGA251, vanB1 and vanB2 genotypes, which are not included in most commercial assays. Large-scale screening is required to further test the performance of this protocol, especially for genotypes that are infrequently encountered.

Predicting antibiotic susceptibility in urinary tract infection with artificial intelligence-model performance in a multi-centre cohort.

Objective: To develop an artificial intelligence model to predict an antimicrobial susceptibility pattern in patients with urinary tract infection (UTI). Materials and methods: 26 087 adult patients with culture-proven UTI during 2015-2020 from a university teaching hospital and three community hospitals in Hong Kong were included. Cases with asymptomatic bacteriuria (absence of diagnosis code of UTI, or absence of leucocytes in urine microscopy) were excluded. Patients from 2015 to 2019 were included in the training set, while patients from the year 2020 were included as the test set.Three first-line antibiotics were chosen for prediction of susceptibility in the bacterial isolates causing UTI: namely nitrofurantoin, ciprofloxacin and amoxicillin-clavulanate. Baseline epidemiological factors, previous antimicrobial consumption, medical history and previous culture results were included as features. Logistic regression and random forest were applied to the dataset. Models were evaluated by F1-score and area under the curve-receiver operating characteristic (AUC-ROC). Results: Random forest was the best algorithm in predicting susceptibility of the three antibiotics (nitrofurantoin, amoxicillin-clavulanate and ciprofloxacin). The AUC-ROC values were 0.941, 0.939 and 0.937, respectively. The F1 scores were 0.938, 0.928 and 0.906 respectively. Conclusions: Random forest model may aid judicious empirical antibiotics use in UTI. Given the reasonable performance and accuracy, these accurate models may aid clinicians in choosing between different first-line antibiotics for UTI.

Independent effect of weather, air pollutants, and seasonal influenza on risk of tuberculosis hospitalization: An analysis of 22-year hospital admission data.

BACKGROUND: While influenza infections and environmental factors have been documented as potential drivers of tuberculosis, no investigations have simultaneously examined their impact on tuberculosis at a population level. This study thereby made use of Hong Kong's surveillance data over 22 years to elucidate the temporal association between environmental influences, influenza infections, and tuberculosis activity. METHODS: Weekly total numbers of hospital admissions due to tuberculosis, meteorological data, and outdoor air pollutant concentrations in Hong Kong during 1998-2019 were obtained. All-type influenza-like illness positive (ILI+) rate and type-specific ILI+ rates were used as proxies for influenza activity. Quasi-Poisson generalized additive models together with distributed lag non-linear models were used to assess the association of interest. RESULTS: A total of 164,116 hospital admissions due to tuberculosis were notified in public settings over a period of 22 years. The cumulative adjusted relative risk (ARR) of hospital admission due to tuberculosis was 1.07 (95% CI, 1.00-1.14) when the mean ambient temperature increased from 15.1 °C (the 5th percentile) to 24.5 °C (median). Short-term exposure to air pollutants was not found to be statistically significantly related to tuberculosis hospitalization. Accounting for the environmental covariates in the analysis, the cumulative ARR of tuberculosis admission was elevated to 1.05 (95% CI, 1.01-1.08) when the rate of ILI+ total increased from zero to 19.9 per 1000 consultations, the 95th percentile. CONCLUSION: Our findings demonstrated that increased influenza activity and higher temperature were related to a higher risk of tuberculosis admissions. Stepping up the promotion of influenza vaccination, especially before the summer season, may lower the risk of tuberculosis infection/reactivation for vulnerable groups (e.g. elderly born before the launch of Bacillus Calmette-Guérin vaccination programme).

Case series of HIV SARS-CoV-2 co-infection in Chinese adults.

Objectives: Little is known whether differences exist in virus shedding, immune and inflammatory response related to SARS-CoV-2 in people living with human immunodeficiency virus (PLWH). We assessed viral RNA and cytokine profiles of HIV and SARS-CoV-2 coinfection in Hong Kong. Methods: PLWH hospitalized with SARS-CoV-2 infection in Hong Kong were included, compared with age-matched and disease severity-matched SARS-CoV-2 infected controls (ratio of 1:5) from February 1st 2020 to July 31st 2020. SARS-CoV-2 infection was confirmed by public health laboratory and virus concentration was quantified by an in-house real-time reverse transcription-quantitative polymerase chain reaction. A panel of cytokines and chemokines were performed. Results: HIV patients had a similar respiratory shedding profile compared to controls. Duration of faecal shedding of patient A, B, C and D were at least 9, 10, 33, and 11 days, respectively. HIV patients had lower plasma levels of IL-10 and NT-pro-BNP. All 4 PLWH cases showed seroconversion to SARS-CoV-2 with anti-SARS-CoV-2 S antibodies detected in serum collected between day 18 and 30 after symptom onset. Conclusions: PLWH behaves similarly with HIV-negative controls in respiratory viral load, but with decrease in IL-10 and NT-proBNP. PLWH may have a lower risk of immunostimulatory effect due to lower IL-10.

Effectiveness of antimicrobial hospital curtains on reducing bacterial contamination-A multicenter study.

OBJECTIVE: To determine the efficacy of 2 types of antimicrobial privacy curtains in clinical settings and the costs involved in replacing standard curtains with antimicrobial curtains. DESIGN: A prospective, open-labeled, multicenter study with a follow-up duration of 6 months. SETTING: This study included 12 rooms of patients with multidrug-resistant organisms (MDROs) (668 patient bed days) and 10 cubicles (8,839 patient bed days) in the medical, surgical, neurosurgical, orthopedics, and rehabilitation units of 10 hospitals. METHOD: Culture samples were collected from curtain surfaces twice a week for 2 weeks, followed by weekly intervals. RESULTS: With a median hanging time of 173 days, antimicrobial curtain B (quaternary ammonium chlorides [QAC] plus polyorganosiloxane) was highly effective in reducing the bioburden (colony-forming units/100 cm2, 1 vs 57; P < .001) compared with the standard curtain. The percentages of MDRO contamination were also significantly lower on antimicrobial curtain B than the standard curtain: methicillin-resistant Staphylococcus aureus, 0.5% vs 24% (P < .001); carbapenem-resistant Acinetobacter spp, 0.2% vs 22.1% (P < .001); multidrug-resistant Acinetobacter spp, 0% vs 13.2% (P < .001). Notably, the median time to first contamination by MDROs was 27.6 times longer for antimicrobial curtain B than for the standard curtain (138 days vs 5 days; P = .001). CONCLUSIONS: Antimicrobial curtain B (QAC plus polyorganosiloxane) but not antimicrobial curtain A (built-in silver) effectively reduced the microbial burden and MDRO contamination compared with the standard curtain, even after extended use in an active clinical setting. The antimicrobial curtain provided an opportunity to avert indirect costs related to curtain changing and laundering in addition to improving patient safety.

Implementation and new insights in molecular diagnostics for HIV infection.

INTRODUCTION: Acquired immunodeficiency syndrome (AIDS) is a kind of acquired disease that breaks down the immune system. Human immunodeficiency virus (HIV) is the causative agent of AIDS. By the end of 2016, there were 36.7 million people living with HIV worldwide. Early diagnosis can alert infected individuals to risk behaviors in order to control HIV transmission. Infected individuals are also benefited from proper treatment and management upon early diagnosis. Thanks to the public awareness of the disease, the annual increase of new HIV infections has been slowly declining over the past decades. The advent of molecular diagnostics has allowed early detection and better management of HIV infected patients. Areas covered: In this review, the authors summarized and discussed the current and future technologies in molecular diagnosis as well as the biomarkers developed for HIV infection. Expert Commentary: A simple and rapid detection of viral load is important for patients and doctors to monitor HIV progression and antiretroviral treatment efficiency. In the near future, it is expected that new technologies such as digital PCR and CRISPR-based technology will play more important role in HIV detection and patient management.