Microbiologia Medica (Jul 2017)
The San Luigi Gonzaga Hospital experience: improving blood and urine culture preanalytical quality by shared protocols
Abstract
Background and aims: Reduction in the number of blood culture and urine culture contamination samples. Materials and methods: We have designed a partly retrospective and partly prospective observational study. On one hand, we have been striving for the creation, dissemination and promotion of shared operational rules in all departments/hospital services to improve the quality of the levy; on the other hand, we analysed data. We considered blood cultures and urine cultures analysed in the laboratory from March to August 2015, and from March to August 2016. The data were processed with R and the incidence of contaminated samples was calculated by dividing the number of blood cultures/urine cultures contaminated by the total. The results of 2015 and 2016 were compared by χ2. To highlight the possible differences between departments and identify those at higher risk of contamination, the data of each year were stratified dividing departments into five groups: Medicine, Surgery, Critical Area, Specialties and ER. To assess the strength of the association, a risk analysis was carried out using the risk ratio (RR). The RR was calculated by dividing the contamination rates of 2015 by the those of 2016. The value of α was set at 0.05. Results: After implementation of the shared protocols, blood culture contamination was substantially reduced (−56.8%, P=1.783e-05), confirmed by an RR of 2.2 (95%CI: 1.54±3.27). The evidence is strengthened by the finding of a lower number of isolates belonging to the group of possible contaminants (−32.7%, P=2.042e-07) and confirmed by an RR of 1.5 (95%CI: 1.27±1.73). Urine culture data analysis showed no change in the incidence of contamination between 2015 and 2016 (P=0.8808), as confirmed by a non-informational RR (95%CI: 0.62±1:46). Even the analysis of the individual areas showed no change in the two semesters, as confirmed by the risk analysis that does not show any association between outcome and group. Conclusions: The results confirm the value of multidisciplinary work and encourage us to continue the path of standardisation and updating of the sampling procedures, as well as the prospective monitoring and comprehensive analysis of the data collected for longer time intervals.
Keywords
