The goal of this methodological, retrospective and prospective study is to; it is a tool to develop a risk estimator tool to detect risk gaps in individuals using artificial intelligence technology that is dangerous for those with CVC in adult intensive care patients, to test risk level estimation frameworks and to evaluate outcomes in the clinic. In our study, it is also our aim to protect, to present the security measures to prevent the risk of CVC with an artificial intelligence model, in an evidence-based way.

The main question[s]it aims to answer are:

• Can the risk of CVC-related infection be determined in adult intensive care patients using artificial intelligence?
• To what degree of accuracy can the risk of CVC-associated infection be determined in adult intensive care patients using artificial intelligence?
• What are the nursing practices that can reduce the risk of CVC-related infections?

Methodology to develop an artificial intelligence-based CVC-associated infection risk level determination algorithm, retrospective using data from Electronic Health Records (EHR) patient data and manual patient files between January 2018 and December 2022 to create the algorithm and test the model accuracy, and the development stages of the model After the completion of the model, up-to-date data were collected for the use of the model and it was planned to be done prospectively.

The applications of artificial intelligence-based technologies in nursing are still in their infancy, and it is emphasized that nurses have limited participation in these processes. We think that our study will be supportive in determining the focal points in the clinical practice of nurses in terms of determining the risk level and will contribute to the development of patient goals. It is also important in terms of creating evidence for making the effects of nursing practices visible. In this context, the aim of our study is to develop a risk estimation tool to determine the risk levels of individuals in terms of CVC-related infection in adult intensive care patients by using artificial intelligence technology, to test the accuracy of the risk level estimation and to apply the tool to evaluate the results in the clinic. In addition, our secondary aim is to present the effects of nursing care in preventing the risk of CVC-related infection with the artificial intelligence model in an evidence-based manner.

Inclusion Criteria:

• Received at least 48 hours of treatment in the GICU,
• Age ≥ 18,
• CVC inserted,
• No existing infection before hospitalization, patient data will be included in the dataset for designing and training the artificial intelligence model.

Exclusion Criteria:

• Age <18,
• Those receiving immunosuppressive therapy,
• Those with multiple organ failure,
• Patients undergoing organ transplantation,
• Patients with a diagnosis of chronic kidney failure, will not be included in the dataset.

All patients admitted to the GICU and meeting the research criteria will be included in the study.