Spatial and sociodemographic factors associated with surgical site infection rates in hospitals in inner São Paulo State, Brazil.

Low-to-middle-income countries often have high incidence of surgical site infection (SSI). To assess spatial and sociodemographic predictors of SSI rates, this study analysed and georeferenced governmental surveillance data from 385 hospitals located in inner São Paulo State, Brazil. In multi-variate models, SSI rates were positively associated with distance from the state capital [incidence rate ratio (IRR) for each 100 km 1.19, 95% confidence interval (CI) 1.07-1.32], and were lower for non-profit (IRR 0.95, 95% CI 0.37-0.85) and private (IRR 0.47, 95% CI 0.31-0.71) facilities compared with public hospitals. Georeferencing results reinforced the need to direct SSI-prevention policies to hospitals located in areas distant from the state capital.

Implementation of tailored interventions in a statewide programme to reduce central line-associated bloodstream infections.

BACKGROUND: There have been few studies exploring implementation strategies to central line-associated bloodstream infections (CLABSIs) in low- or middle-income countries. AIM: To implement tailored interventions to reduce CLABSI rates in adult intensive care units. METHODS: The implementation strategy of the State Health Department was performed in São Paulo State, Brazil, over two cycles. Cycle 1 (56 hospitals) was exploratory and cycle 2 (77 hospitals) was designed to confirm the hypothesis generated by the first cycle, with three phases each (pre-intervention, intervention, post-intervention). Cycles included: evaluation of healthcare workers' knowledge, observation of practices, and CLABSI rates monthly report. In cycle 1, a log-normal mixed model was used to select variables significantly associated with the reduction of CLABSI. In cycle 2, CLABSI rates were evaluated. FINDINGS: Healthcare workers' practices improved after intervention. In cycle 1, reduction of CLABSI rates was more pronounced in hospitals with initial CLABSI rates >7.4 per 1000 catheter-days (P < 0.001) and those that introduced the use of peripherally inserted central catheters (P = 0.01). For hospitals with high CLABSI initial rates, simulation demonstrated that the rates were expected to decrease by 36% (95% CI: 9-63), no matter the type of intervention. In cycle 2, there was an overall decrease in CLABSI rates during the intervention period; whereas the mean rate fell further post-intervention, rates at the 90th percentile increased. CONCLUSION: The implementation strategy may have had an effect on infection rates independently of the specific interventions implemented; however, the sustainability of reduction in the post-intervention period remains a challenge.

Surveillance Programme for Healthcare Associated Infections in the State of São Paulo, Brazil. Implementation and the first three years' results.

Governmental programmes should be developed to collect and analyse data on healthcare associated infections (HAIs). This study describes the healthcare setting and both the implementation and preliminary results of the Programme for Surveillance of Healthcare Associated Infections in the State of São Paulo (PSHAISP), Brazil, from 2004 to 2006. Characterisation of the healthcare settings was carried out using a national database. The PSHAISP was implemented using components for acute care hospitals (ACH) or long term care facilities (LTCF). The components for surveillance in ACHs were surgical unit, intensive care unit and high risk nursery. The infections included in the surveillance were surgical site infection in clean surgery, pneumonia, urinary tract infection and device-associated bloodstream infections. Regarding the LTCF component, pneumonia, scabies and gastroenteritis in all inpatients were reported. In the first year of the programme there were 457 participating healthcare settings, representing 51.1% of the hospitals registered in the national database. Data obtained in this study are the initial results and have already been used for education in both surveillance and the prevention of HAI. The results of the PSHAISP show that it is feasible to collect data from a large number of hospitals. This will assist the State of São Paulo in assessing the impact of interventions and in resource allocation.

Outbreak of surgical infection caused by non-tuberculous mycobacteria in breast implants in Brazil.

SUMMARY: We investigated an outbreak caused by non-tuberculous mycobacteria (NTM) related to breast implant surgery in the city of Campinas, Brazil, by means of a retrospective cohort and molecular epidemiological study. A total of 492 records of individuals having breast surgery in 12 hospitals were evaluated. Twelve isolates were analysed using four different molecular typing methods. There were 14 confirmed cases, 14 possible cases and one probable case. One probable, nine possible and 12 confirmed cases were included in a cohort study; all occurred in eight of the hospitals and the confirmed cases in five. Univariate analysis showed that patients who had had breast reconstruction surgery in hospitals A and B were more likely to have NTM infections. No risk factor was independently associated with NTM infection in the multivariate model. The isolates obtained from patients at each hospital showed different molecular patterns, excluding isolates from hospital C that repeatedly showed the same genotype for approximately one year. In conclusion, this outbreak was caused by polyclonal strains at different institutions, and in one hospital a unique genotype caused most cases. No specific risk factors were found.