Andrea Ripoli1, Emanuela Sozio2, Francesco Sbrana3, Giacomo Bertolino4,5, Carlo Pallotto6,7, Gianluigi Cardinali8,9, Simone Meini10, Filippo Pieralli11, Anna Maria Azzini12, Ercole Concia12, Bruno Viaggi13, Carlo Tascini14. 1. Bioengineering Department, Fondazione Toscana Gabriele Monasterio, Pisa, Italy. 2. Emergency Department, North-West District, Tuscany Health Care, Spedali Riuniti Livorno, Livorno, Italy. 3. U.O. Lipoapheresis and Center for Inherited Dyslipidemias, Fondazione Toscana Gabriele Monasterio, Via Moruzzi,1, 56124, Pisa, Italy. francesco.sbrana@ftgm.it. 4. Pharmaceutical Department, ASSL Cagliari, Cagliari, Italy. 5. Department of Medical Sciences and Public Health, University of Cagliari, Cagliari, Italy. 6. UOC Malattie Infettive, Ospedale San Donato Arezzo, Sud-Est District, Tuscany Health Care, Arezzo, Italy. 7. Sezione Di Malattie Infettive, Dipartimento Di Medicina, Università Di Perugia, Perugia, Italy. 8. Department of Pharmaceutical Sciences-Microbiology, University of Perugia, Perugia, Italy. 9. CEMIN, Centre of Excellence On Nanostructured Innovative Materials, Department of Chemistry, Biology and Biotechnology, University of Perugia, Perugia, Italy. 10. Internal Medicine Unit, Santa Maria Annunziata Hospital, Florence, Italy. 11. Intermediate Care Unit, Azienda Ospedaliera Universitaria Careggi, Florence, Italy. 12. Dipartimento Di Diagnostica E Sanità Pubblica, Sezione Di Malattie Infettive, Università Di Verona, Verona, Italy. 13. Department of Anesthesia, Neuro Intensive Care Unit, Careggi Universital Hospital, Florence, Italy. 14. First Division of Infectious Diseases, Cotugno Hospital, Azienda Ospedaliera Dei Colli, Napoli, Italy.
Abstract
PURPOSE: Candidemia is a highly lethal infection; several scores have been developed to assist the diagnosis process and recently different models have been proposed. Aim of this work was to assess predictive performance of a Random Forest (RF) algorithm for early detection of candidemia in the internal medical wards (IMWs). METHODS: A set of 42 potential predictors was acquired in a sample of 295 patients (male: 142, age: 72 ± 15 years; candidemia: 157/295; bacteremia: 138/295). Using tenfold cross-validation, a RF algorithm was compared with a classic stepwise multivariable logistic regression model; discriminative performance was assessed by C-statistics, sensitivity and specificity, while calibration was evaluated by Hosmer-Lemeshow test. RESULTS: The best tuned RF algorithm demonstrated excellent discrimination (C-statistics = 0.874 ± 0.003, sensitivity = 84.24% ± 0.67%, specificity = 91% ± 2.63%) and calibration (Hosmer-Lemeshow statistics = 12.779 ± 1.369, p = 0.120), markedly greater than the ones guaranteed by the classic stepwise logistic regression (C-statistics = 0.829 ± 0.011, sensitivity = 80.21% ± 1.67%, specificity = 84.81% ± 2.68%; Hosmer-Lemeshow statistics = 38.182 ± 15.983, p < 0.001). In addition, RF suggests a major role of in-hospital antibiotic treatment with microbioma highly impacting antimicrobials (MHIA) that are found as a fundamental risk of candidemia, further enhanced by TPN. When in-hospital MHIA therapy is not performed, PICC is the dominant risk factor for candidemia, again enhanced by TPN. When PICC is not used and MHIA therapy is not performed, the risk of candidemia is minimum, slightly increased by in-hospital antibiotic therapy. CONCLUSION: RF accurately estimates the risk of candidemia in patients admitted to IMWs. Machine learning technique might help to identify patients at high risk of candidemia, reduce the delay in empirical treatment and improve appropriateness in antifungal prescription.
PURPOSE:Candidemia is a highly lethal infection; several scores have been developed to assist the diagnosis process and recently different models have been proposed. Aim of this work was to assess predictive performance of a Random Forest (RF) algorithm for early detection of candidemia in the internal medical wards (IMWs). METHODS: A set of 42 potential predictors was acquired in a sample of 295 patients (male: 142, age: 72 ± 15 years; candidemia: 157/295; bacteremia: 138/295). Using tenfold cross-validation, a RF algorithm was compared with a classic stepwise multivariable logistic regression model; discriminative performance was assessed by C-statistics, sensitivity and specificity, while calibration was evaluated by Hosmer-Lemeshow test. RESULTS: The best tuned RF algorithm demonstrated excellent discrimination (C-statistics = 0.874 ± 0.003, sensitivity = 84.24% ± 0.67%, specificity = 91% ± 2.63%) and calibration (Hosmer-Lemeshow statistics = 12.779 ± 1.369, p = 0.120), markedly greater than the ones guaranteed by the classic stepwise logistic regression (C-statistics = 0.829 ± 0.011, sensitivity = 80.21% ± 1.67%, specificity = 84.81% ± 2.68%; Hosmer-Lemeshow statistics = 38.182 ± 15.983, p < 0.001). In addition, RF suggests a major role of in-hospital antibiotic treatment with microbioma highly impacting antimicrobials (MHIA) that are found as a fundamental risk of candidemia, further enhanced by TPN. When in-hospital MHIA therapy is not performed, PICC is the dominant risk factor for candidemia, again enhanced by TPN. When PICC is not used and MHIA therapy is not performed, the risk of candidemia is minimum, slightly increased by in-hospital antibiotic therapy. CONCLUSION: RF accurately estimates the risk of candidemia in patients admitted to IMWs. Machine learning technique might help to identify patients at high risk of candidemia, reduce the delay in empirical treatment and improve appropriateness in antifungal prescription.
Entities:
Keywords:
Candidemia; Machine learning; Medical ward; Septic patients
Authors: Eyal Zimlichman; Daniel Henderson; Orly Tamir; Calvin Franz; Peter Song; Cyrus K Yamin; Carol Keohane; Charles R Denham; David W Bates Journal: JAMA Intern Med Date: 2013 Dec 9-23 Impact factor: 21.873
Authors: Angela Ahlquist Cleveland; Monica M Farley; Lee H Harrison; Betsy Stein; Rosemary Hollick; Shawn R Lockhart; Shelley S Magill; Gordana Derado; Benjamin J Park; Tom M Chiller Journal: Clin Infect Dis Date: 2012-08-14 Impact factor: 9.079
Authors: Matteo Bassetti; Maria Merelli; Elda Righi; Ana Diaz-Martin; Eva Maria Rosello; Roberto Luzzati; Anna Parra; Enrico Maria Trecarichi; Maurizio Sanguinetti; Brunella Posteraro; Jose Garnacho-Montero; Assunta Sartor; Jordi Rello; Mario Tumbarello Journal: J Clin Microbiol Date: 2013-10-09 Impact factor: 5.948
Authors: Hilmar Wisplinghoff; Tammy Bischoff; Sandra M Tallent; Harald Seifert; Richard P Wenzel; Michael B Edmond Journal: Clin Infect Dis Date: 2004-07-15 Impact factor: 9.079