A machine learning approach to predict healthcare-associated infections at intensive care unit admission: findings from the SPIN-UTI project.

BACKGROUND: Identifying patients at higher risk of healthcare-associated infections (HAIs) in intensive care units (ICUs) represents a major challenge for public health. Machine learning could improve patient risk stratification and lead to targeted infection prevention and control interventions. AIM: To evaluate the performance of the Simplified Acute Physiology Score (SAPS) II for HAI risk prediction in ICUs, using both traditional statistical and machine learning approaches. METHODS: Data for 7827 patients from the 'Italian Nosocomial Infections Surveillance in Intensive Care Units' project were used in this study. The Support Vector Machines (SVM) algorithm was applied to classify patients according to sex, patient origin, non-surgical treatment for acute coronary disease, surgical intervention, SAPS II at admission, presence of invasive devices, trauma, impaired immunity, and antibiotic therapy in 48 h preceding ICU admission. FINDINGS: The performance of SAPS II for predicting HAI risk provides a receiver operating characteristic curve with an area under the curve of 0.612 (P<0.001) and accuracy of 56%. Considering SAPS II along with other characteristics at ICU admission, the SVM classifier was found to have accuracy of 88% and an AUC of 0.90 (P<0.001) for the test set. The predictive ability was lower when considering the same SVM model but with the SAPS II variable removed (accuracy 78%, AUC 0.66). CONCLUSIONS: This study suggested that the SVM model is a useful tool for early prediction of patients at higher risk of HAIs at ICU admission.

Cluster analysis identifies patients at risk of catheter-associated urinary tract infections in intensive care units: findings from the SPIN-UTI Network.

BACKGROUND: Although preventive strategies have been proposed against catheter-associated urinary tract infections (CAUTIs) in intensive care units (ICUs), more efforts are needed to control the incidence rate. AIM: To distinguish patients according to their characteristics at ICU admission, and to identify clusters of patients at higher risk for CAUTIs. METHODS: A two-step cluster analysis was conducted on 9656 patients from the Italian Nosocomial Infections Surveillance in Intensive Care Units project. FINDINGS: Three clusters of patients were identified. Type of admission, patient origin and administration of antibiotics had the greatest weight on the clustering model. Cluster 1 comprised more patients with a medical type of ICU admission who came from the community. Cluster 2 comprised patients who were more likely to come from other wards/hospitals, and to report administration of antibiotics 48 h before or after ICU admission. Cluster 3 was similar to Cluster 2 but was characterized by a lower percentage of patients with administration of antibiotics 48 h before or after ICU admission. Patients in Clusters 1 and 2 had a longer duration of urinary catheterization [median 7 days, interquartile range (IQR) 12 days for Cluster 1; median 7 days, IQR 11 days for Cluster 2] than patients in Cluster 3 (median 6 days, IQR 8 days; P<0.001). Interestingly, patients in Cluster 1 had a higher incidence of CAUTIs (3.5 per 100 patients) compared with patients in the other two clusters (2.5 per 100 patients in both clusters; P=0.033). CONCLUSION: To the authors' knowledge, this is the first study to use cluster analysis to identify patients at higher risk of CAUTIs who could gain greater benefit from preventive strategies.