Spatio-temporal ecological model of predictors of healthcare-associated infections in intensive care units of São Paulo state

Author(s)

Dayanne Conislla Limaylla

• Document type: Doctoral Thesis
• Press: Botucatu. 2023-05-12.
• Institution: Universidade Estadual Paulista (Unesp). Faculdade de Medicina. Botucatu
• Defense date: 2023-04-26
• Advisor: Carlos Magno Castelo Branco Fortaleza

Abstract

Healthcare-associated infections (HAIs) cause high morbidity and mortality worldwide. Among inpatients, those at greater risk of acquiring HAI are those admitted to Intensive Care Units (ICU). We carried out an ecological study with the aim of identifying predictors of Device-associated hospital-acquired infections (DA-HAIs) in an ICU in São Paulo State, including hospital, socioeconomic and spatial characteristics. Data from the São Paulo HAI surveillance for the period from 2011 through 2018 were collected. The incidence of device-associated infections was georeferenced and submitted to zero-inflated binomial regression analysis. We found that the incidence of all three types of device-associated HAIs under study decreased over time, but there were regional differences in the incidence rates, with Ventilator Associated Pneumonia (VAP) and Urinary Tract Infections (UTI) rates being higher in the upstate regions and Blood-stream Infections (BSI) rates slightly higher in the metropolitan region. In the multivariate analysis, distance from the state capital was positively associated with VAP (IRR:1.06, 95%CI: 1.04-1.08) and UTI (IRR:1.03, 95%CI: 1.01-1.05 ), and negatively associated with BSI (IRR: 0.95, 95%CI: 0.93-0.97) meaning that the factor is indeed significant but affect the infection frequency in opposite ways. Private and non-profit hospitals, as well as those with a greater number of hospital beds, were likely to have lower HAI rates. Indicators such as city Human Development Index (HDI) and mean Income were significant only for VAP, suggesting that better rates of these determinants are associated with lower cases of the disease. During the space-time analysis, several clusters and significant hotspots were identified close to the metropolitan region of the state. However, over time, smaller clusters and fewer hot spots were found, indicating isolated occurrences and reduced incidence in the study areas. These findings show us the need for targeted intervention strategies to reduce infections in ICUs in hospitals located in areas around the metropolitan region and in specific distant areas in the interior of the state (AU)

• FAPESP's process: 19/18775-5 - Space-time ecological model of predictors of healthcare associated infection intensive care units in São Paulo State, Brazil
• Grantee: Dayanne Conislla Limaylla
• Support Opportunities: Scholarships in Brazil - Doctorate