Nursing workload: use of artificial intelligence to develop a classifier model.

OBJECTIVE: to describe the development of a predictive nursing workload classifier model, using artificial intelligence. METHOD: retrospective observational study, using secondary sources of electronic patient records, using machine learning. The convenience sample consisted of 43,871 assessments carried out by clinical nurses using the Perroca Patient Classification System, which served as the gold standard, and clinical data from the electronic medical records of 11,774 patients, which constituted the variables. In order to organize the data and carry out the analysis, the Dataiku® data science platform was used. Data analysis occurred in an exploratory, descriptive and predictive manner. The study was approved by the Ethics and Research Committee of the institution where the study was carried out. RESULTS: the use of artificial intelligence enabled the development of the nursing workload assessment classifier model, identifying the variables that most contributed to its prediction. The algorithm correctly classified 72% of the variables and the area under the Receiver Operating Characteristic curve was 82%. CONCLUSION: a predictive model was developed, demonstrating that it is possible to train algorithms with data from the patient's electronic medical record to predict the nursing workload and that artificial intelligence tools can be effective in automating this activity.

Hospital-acquired infections surveillance: The machine-learning algorithm mirrors National Healthcare Safety Network definitions.

BACKGROUND: Surveillance of hospital-acquired infections (HAIs) is the foundation of infection control. Machine learning (ML) has been demonstrated to be a valuable tool for HAI surveillance. We compared manual surveillance with a supervised, semiautomated, ML method, and we explored the types of infection and features of importance depicted by the model. METHODS: From July 2021 to December 2021, a semiautomated surveillance method based on the ML random forest algorithm, was implemented in a Brazilian hospital. Inpatient records were independently manually searched by the local team, and a panel of independent experts reviewed the ML semiautomated results for confirmation of HAI. RESULTS: Among 6,296 patients, manual surveillance classified 183 HAI cases (2.9%), and a semiautomated method found 299 HAI cases (4.7%). The semiautomated method added 77 respiratory infections, which comprised 93.9% of the additional HAIs. The ML model considered 447 features for HAI classification. Among them, 148 features (33.1%) were related to infection signs and symptoms; 101 (22.6%) were related to patient severity status, 51 features (11.4%) were related to bacterial laboratory results; 40 features (8.9%) were related to invasive procedures; 34 (7.6%) were related to antibiotic use; and 31 features (6.9%) were related to patient comorbidities. Among these 447 features, 229 (51.2%) were similar to those proposed by NHSN as criteria for HAI classification. CONCLUSION: The ML algorithm, which included most NHSN criteria and >200 features, augmented the human capacity for HAI classification. Well-documented algorithm performances may facilitate the incorporation of AI tools in clinical or epidemiological practice and overcome the drawbacks of traditional HAI surveillance.