Legionella pneumophila in un sistema idrico ospedaliero: applicazione della geostatistica a funzioni aleatorie di ordine k (faik) quale stimatore della distribuzione

Translated title of the contribution: [Autom. eng. transl.] Legionella pneumophila in a hospital water system: application of geostatistics to random functions of order k (faik) as estimator of the distribution

Umberto Moscato, Andrea Poscia, Daniele Ignazio La Milia, P Laganà, E Avventuroso, S. Delia

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

[Autom. eng. transl.] It is now common acquisition that Legionellosis is contracted by inhalation of Legionelle from a contaminated environmental source. The water systems of large buildings, such as hospitals, are often contaminated by the germ, representing a potential risk of infection. The objective of this study was to carry out a comparison between classical, inferential and geostatistic descriptive analysis on data from microbiological analyzes carried out on water samples, to assess whether the application of geostatistics, for random K order functions , is a more effective estimator of descriptive or inferential statistics capable of explaining the phenomena of environmental diffusion of Legionella and identifying possible sources of contamination and, consequently, the risk of exposure both to health personnel and to patients. Method: The descriptive epidemiological study is of the cross sectional type. For the survey, data from microbiological analyzes carried out for the research of L. pneumophila were used, during the monitoring activity between 2004 and 2011, on water samples taken at various points distributed on two contiguous floors ( ground floor and first floor) of a pavilion of the AOU 'G. Martino di Messina. The information was collected, stored in a database and geo-referenced through the use of the FAIPACK Software, in which each parameter was identified with the coordinates of the water pipes coinciding with the withdrawal points. Results: Over 50% of the samples tested were positive for Legionella pneumophila sgr 1 and 2-14. L. pneumophila 1 was isolated in 69% of the samples coming from the ground floor and in 60% from those of the first floor; L. pneumophila 2-14 was found in 36% of samples taken from the ground floor and 24% from those on the first floor. From the application of the Shapiro-Wilk test it was deduced that the variables followed a non-normal distribution, for which the non-parametric Wilcoxon test of the sum sum was used for the comparison between two independent groups, obtaining a value p = 0.005 for each serogroup / per floor. Analyzing the iso-estimation maps created by the FAIPACK software, we clearly delineate the most contaminated pipelines and the difference in the diffusion of the different serogroups. Conclusion: The experimental investigation has underlined the importance of the use of geostatistical analysis, a discipline that studies the behavior of the variables, their correlation and their spatial structure allowing a reliable estimate of values that are not directly measurable: the iso- maps show the distribution of Legionella pneumophila in the water system and allow us to estimate the concentration of the germ in the pipelines of the entire plane and not only at the points where the samples were taken. The study has also shown that the geostatistical method applied to microbiological data deriving from water analysis, integrated with classical statistical analysis, allows a more in-depth interpretation of the phenomenon, provides a more powerful tool for managing 'Legionella' risk by offering the possibility of making more prudent choices in the management of preventive acts to be implemented.
Translated title of the contribution[Autom. eng. transl.] Legionella pneumophila in a hospital water system: application of geostatistics to random functions of order k (faik) as estimator of the distribution
Original languageItalian
Title of host publicationAtti del 46° Congresso Nazionale SItI
PagesN/A
Publication statusPublished - 2013
Event46° Congresso Nazionale SItI - Taormina
Duration: 17 Oct 201320 Oct 2013

Conference

Conference46° Congresso Nazionale SItI
CityTaormina
Period17/10/1320/10/13

Keywords

  • Geostatistica
  • Legionella

Fingerprint

Dive into the research topics of '[Autom. eng. transl.] Legionella pneumophila in a hospital water system: application of geostatistics to random functions of order k (faik) as estimator of the distribution'. Together they form a unique fingerprint.

Cite this