1668. Clinical Risk Score Model for Predicting Extended Spectrum Beta Lactamase (ESBL)-Positive Urinary Tract Infection among Hospitalized Filipino Patients: a Retrospective Single-Center Study

• Published in: Open forum infectious diseases Vol. 7; no. Supplement_1; pp. S820 - S821
• Main Authors: Edding, Sherida, Cadeliña, Jekrylei, Leaño, Danice Romagne, Panaligan, Mario, Hufano, Ma Charmian
• Format: Journal Article
• Language: English
• Published: US Oxford University Press 31.12.2020
• Subjects: Poster Abstracts
• ISSN: 2328-8957; 2328-8957
• DOI: 10.1093/ofid/ofaa439.1846

Abstract Background Infections caused by Extended-Spectrum B-lactamases (ESBL) organisms are an emerging health concern worldwide. In the background of progressive rise of antibiotic resistant organisms, efforts should be started to minimize antibiotic resistance. The development of clinical risk score models to predict the possibility of infection with these drug resistant organisms will bridge the gap between treating too late and treating too early and too much. Methods This is a single-center cross-sectional study of ESBL positive urinary tract infection (UTI) among patients admitted at St Luke’s Medical Center - Global City from January 2018 to December 2018. A total of 93 patients positive for ESBL-positive UTI and 186 patients with ESBL-negative UTI were included in the study. Clinical characteristics, medical and medication histories were obtained from the computerized medical database. To develop the risk score, clinical risk indicators measured at the time of recruitment were built by logistic regression. Regression coefficients were transformed into item scores and added up to a total score. A risk-scoring scheme was developed from clinical predictors. Results The following factors were observed to be predictive of ESBL UTI: hospital-acquired or healthcare-associated infection, prior ESBL infection, cardiac comorbidity and history of carbapenem intake. Scoring for the predictive model is as follows: having age at least 50 years old=1.5, hospital-acquired or healthcare-associated infection=1, having prior ESBL infection=2, having cardiac comorbidity (NYHA I-II)=1, having cardiac comorbidity (NYHA III-IV)=2.5, and history of carbapenem intake=3. ROC analysis showed that the optimum cut point in the model predictive of ESBL is 3.5/10. The risk prediction model for ESBL had high sensitivity of 87%, medium specificity of 68%, and good predictive accuracy of 0.78. Conclusion A simple and non-invasive scoring scheme of three predictors provides good prediction indices for presence of ESBL organisms in patients diagnosed with UTI. However, a large sample study is needed to improve the power of the study. Validation studies are also needed. Disclosures All Authors: No reported disclosures