The accuracy of fully automated algorithms for surveillance of healthcare-associated urinary tract infections in hospitalized patients

Abstract

Background: Surveillance for healthcare-associated infections such as healthcareassociated urinary tract infections (HA-UTI) is important for directing resources and evaluating interventions. However, traditional surveillance methods are resourceintensive and subject to bias. Aim: To develop and validate a fully automated surveillance algorithm for HA-UTI using electronic health record (EHR) data. Methods: Five algorithms were developed using EHR data from 2979 admissions at Karolinska University Hospital from 2010 to 2011: (1) positive urine culture (UCx); (2) positive UCx þ UTI codes (International Statistical Classification of Diseases and Related Health Problems, 10 th revision); (3) positive UCx þ UTI-specific antibiotics; (4) positive UCx þ fever and/or UTI symptoms; (5) algorithm 4 with negation for fever without UTI symptoms. Natural language processing (NLP) was used for processing free-text medical notes. The algorithms were validated in 1258 potential UTI episodes from January to March 2012 and results extrapolated to all UTI episodes within this period (N ¼ 16,712). The reference standard for HA-UTIs was manual record review according to the European Centre for Disease Prevention and Control (and US Centers for Disease Control and Prevention) definitions by trained healthcare personnel. Findings: Of the 1258 UTI episodes, 163 fulfilled the ECDC HA-UTI definition and the algorithms classified 391, 150, 189, 194, and 153 UTI episodes, respectively, as HA-UTI. Algorithms 1, 2, and 3 had insufficient performances. Algorithm 4 achieved better performance and algorithm 5 performed best for surveillance purposes with sensitivity 0.667 (95% confidence interval: 0.594e0.733), specificity 0.997 (0.996e0.998), positive predictive value 0.719 (0.624e0.807) and negative predictive value 0.997 (0.996e0.997).

Risk factors for hospital-acquired infection during the SARS-CoV-2 pandemic

“…Previous studies have shown that the correlation between laboratory parameters and ICD-10 codes may vary by site of infection. 25 , 26 Our study was not designed to assess the relationship between laboratory parameters and ICD-10 codes, and this is a potential limitation, i.e. that our sensitivity analyses did not fully capture the population with true HAIs.…”

Federated Semi-Supervised Multi-Task Learning to Detect COVID-19 and Lungs Segmentation Marking Using Chest Radiography Images and Raspberry Pi Devices: An Internet of Medical Things Application

“…Data can be a wealth of resources if they can be adequately represented to tackle a healthcare problem such as urinary tract infection detection [ 56 ]. The effective surveillance deployment [ 57 ] also depends on the more nuanced representation of it. As IoMT deals with the effective integration of the medical equipment and its associated data, a careful exploratory analysis of this heterogeneous data and their applicability should be investigated together.…”

"Machine Learning Application to Combat Superbugs in Hospitals: A Primer to Infection Prevention Practitioners"

“…ML models could be made for surveillance of Blood Stream Infections (BSI), CD Infections (CDI), Urinary Tract Infections (UTI), pneumonia and Surgical Site Infections (SSI). Vab der Werff, et al [12] developed a fully automated Surviellance algorithm for hospital acquired UTI using electronic health record (EHR) data. This study concluded that a fully automated surveillance algorithm based on artificial intelligence and machine learning to detect UTI symptoms from EHR had acceptable performance HA-UTI compared to manual record review.…”