Thesis title: TESTING OF A MODEL FOR THE INTEGRATION BETWEEN CLINICAL RISK AND INFECTIOUS RISK AIMED AT MANAGEMENT OF ADVERSE EVENTS CLASSIFIABLE AS HAI IN A SECOND-LEVEL HOSPITAL
Hospital-acquired infections, also known as healthcare-associated infections (HAI), are an infection that develops during hospitalization, not present, or in incubation, upon entry into the hospital. Despite the possibility to access new information systems and the large amount of data offered by the automatic warning systems - "big data", the management of information flows related to nosocomial infections is still very compartmentalized. The aggregation of information flows requires a model that involves the construction of a common taxonomic platform where can be placed (merged) data come from different sources (epidemiological clinical and forensic). In the literature, there are some international and national experiences of using reference taxonomies for the classification of events and dangers related to patient safety. The International Classification of Patient Safety (ICPS) developed by the World Health Organization (WHO) appears to be a further important step towards a comprehensive conceptual framework of patient safety, which is able to facilitate the comparison of results from different sources of information, both within health organizations and between institutions, at local, national and international level.
Objectives:
This study aimed to evaluate a model for integrating clinical risk and infectious risk to manage data flow from adverse events classified as healthcare-associated infections (HAI) in a secondary-level hospital. The validation of the ICPS classification system introduced by the WHO within a second-level Hospital, allows the structuring of a common platform that allows the integration of data from the management of litigation, voluntary reporting systems (IR) of forensic interest, and related to usually serious events (with death or serious injury to patients), with data from administrative and health surveillance systems of epidemiological clinical interest.
Materials and Methods
We conducted a retrospective study based on the quantitative analysis of litigation data from a second-level hospital healthcare, during the period from january 1st to 2017 to june 30th 2025. We implemented a new event classification system based on the ICPS classification platform, available by the WHO, identifying each classification subgroup and focusing specifically on the subgroup of healthcare-associated infections (HAI). Within this subgroup of HAIs that were consistent according to inclusion criteria, we proceeded to identify the site of infection according to the classification established by the ICPS. Through an audit review technique with a multidisciplinary study group (Infectious Disease Specialists, Intensive Care Specialist), we identified the contributing factors for each infectious case and assessed the level of adherence to available best practices, using the sepsis bundle made available since 2016 as a process indicator. To verify the reliability of the data from the electronic health records (EHRs), we checked for the presence of ICD-9 codes for infection/sepsis in cases previously identified as hospital-acquired infections (HAIs).
Results:
From the quantitative analysis of litigation data (2017–2025, n. 566 cases), we identified 112 cases in which the claim concerned a healthcare-associated infection.The HAIs described according to the ICPS classification showed that the most represented categories were: 37% pneumonia, 23% surgical site infections, 21% urinary tract infections, 10% central venous catheter-related infections, and 9% other infections. A multidisciplinary analysis of the cases was carried out in order to highlight the right application of sepsis bundles, with the following findings: lactate levels were measured in 36% of suspected sepsis cases; in 53% of cases, cultures were performed before the administration of antibiotics; in 92% of cases, broad-spectrum antibiotics were administered; and in 90% of cases, intravenous crystalloids were given in the presence of hypoperfusion or lactate values higher than 4 mmol/l. We selected the claims HAIs that were consistent according inclusion criteria (no. 87) and analyzed the hospital discharge records (EHR). We analyzed the degree of agreement with the ICD-9 codes reported among the secondary diagnoses in the hospital discharge records. In 45% of cases, at least one infection code was indicated in the secondary diagnosis field; in 27% of cases, there was no infection/sepsis code in secondary diagnosis field; and in 23% of cases, a sepsis code was present.
The implementation of the ICPS platform was an opportunity to share information deriving from the analysis of litigation in the field of healthcare-associated infections. The data obtained from ICPS platform easily allow for multidisciplinary analyses, with the opportunity to assess the degree of adherence to good clinical practices and guidelines.