During the first year, the implementation of the Water Safety Plan (WSP) was started in pavilions A and B of the Policlinico of Bari. The second year focused on the evaluation of the effectiveness of the implementation of the WSP in the water supply chains of the two pavilions and the development of a predictive model (ML) of Legionella spp. contamination in the water supply of pavilion A. Finally, two other studies were conducted in the third year. In the first, a cognitive survey was conducted in the Apulia region to assess hospitals' knowledge of D.Lgs 18/23 and the application of WSPs. In the second study, a quantitative microbial risk assessment (QMRA) was applied to estimate the infectious risk of Legionella spp., and a cost-benefit analysis was applied to evaluate the surveillance strategy (total or reduced) of the Policlinico of Bari water supply by routine microbiological monitoring. In the first study, operational monitoring showed a progressive increase in compliance with L. pneumophila and P. aeruginosa parameters in both pavilions due to local treatments on the indoor water distribution systems. In pavilion A, among the ML models tested, the most efficient had an input layer of 56 neurons, a hidden layer of 30 neurons, and an output layer of two neurons, with an accuracy of 93.4%, a sensitivity of 43.8%, and a specificity of 96%. The Poisson regression model had an accuracy of 82.9%, a sensitivity of 20.3%, and a specificity of 97.3%. The most important parameters that influenced the results of the models were type of water supply (hot/cold), filter valve replacement, and atmospheric temperature. In the second study, 38.9% of the questionnaires were collected in Apulia. In 85.7% of the cases, they were aware of Legislative Decree 18/23, the related guidelines (ISTISAN reports 22/32) and the meaning of "priority buildings." Most of them carried out risk analysis of water pipes (92.9%) and chemical-physical and microbiological tests (64.3%). Legionella testing was performed in 100% of the cases, of which 78.6% were entrusted to an accredited private laboratory. In the third study on QMRA, the probability of infection generated by positive specimens in the case of total surveillance (ST) was 2.6*10-5 in high-risk wards, 6.1*10-7 in medium-risk wards, and 2.95*10-7 in low-risk wards. In reduced surveillance (SR), the probabilities of infection were 3.2*10-5, 1.5*10-5, and 3.1*10-6 in the high-, intermediate-, and low-risk groups, respectively. The higher probability of infection in the SR for the intermediate and low-risk wards indicated greater exposure to the pathogen in the absence of complete surveillance of all points. The total cost estimate was 422,900 euros for ST and 276,700 euros for SR. The NB for the SR strategy was found to be 149,959 euros if the WTP was set at 25,000 euros less per case. The NB for ST, on the other hand, was -198,827 euros; therefore, the costs exceed the benefits. The EIB given by the difference in NB between SR and ST was 348,787 euros, showing that the SR strategy was more favorable. SR was more favorable than ST because with SR the costs were lower and the damage caused was insignificant. With ST, there was a gain in effectiveness, but overspending that far exceeded the cost of not monitoring.
Durante il primo anno è stata avviata l’implementazione del Piano di Sicurezza dell’Acqua (PSA) in due padiglioni A e B dell’ospedale Policlinico di Bari. Il secondo anno ha riguardato la valutazione dell’efficacia applicativa del Piano di Sicurezza Dell’Acqua presso le filiere idriche dei due padiglioni succitati ed è stato sviluppato un modello predittivo di contaminazione da Legionella spp. nella rete idrica del padiglione A. Infine, il terzo anno sono stati condotti altri due studi. Nel primo, è stata effettuata un’indagine conoscitiva nella regione Puglia per valutare la conoscenza degli ospedali in materia di D.Lgs 18/23 e applicazione dei PSA. Nel secondo studio è stata applicata una valutazione quantitativa del rischio microbico (QMRA) per stimare il rischio infettivo da Legionella spp. e Analisi Costo-beneficio per valutare la strategia di sorveglianza (totale o ridotta) della rete idrica del Policlinico di Bari tramite monitoraggio microbiologico routinario. Nel primo studio, il monitoraggio operativo ha evidenziato un progressivo aumento della conformità ai parametri L.pneumophila e P.aeruginosa in entrambi i padiglioni A e B grazie ai trattamenti locali sui sistemi di distribuzione idrica interni. Nel padiglione A tra i modelli ML testati, il più efficiente aveva uno strato di input pari a 56 neuroni, uno strato nascosto con 30 neuroni e uno di output con due neuroni con un’accuratezza del 93,4%, sensibilità del 43,8% e specificità del 96%. Il modello di regressione di Poisson ha mostrato un'accuratezza dell'82,9%, una sensibilità del 20,3% e una specificità del 97,3%. I parametri più importanti che hanno influenzato i risultati dei modelli sono stati il tipo di rete idrica (calda/fredda), la sostituzione delle valvole del filtro e la temperatura atmosferica. Nel secondo studio sono stati raccolti in Puglia il 38,9% di questionari. Nell’85,7% dei casi erano a conoscenza del D.Lgs 18/23, delle relative Linee Guida (Rapporti ISTISAN 22/32 ) e del significato di “edifici prioritari”. La maggior parte eseguiva l’analisi del rischio sulla rete idrica (92,9%) e i controlli chimico-fisici e microbiologici (64,3%). La ricerca di Legionella era effettuata nel 100% dei casi, di cui il 78,6% affidata a un laboratorio privato accreditato. Nel terzo studio relativo alla QMRA, la probabilità di infezione generata dai punti positivi nel caso della sorveglianza totale (ST) è risultata 2.6*10-5 nei reparti ad alto rischio, 6.1*10-7 nei reparti a rischio intermedio e 2.95*10-7 nei reparti a basso rischio. Nella sorveglianza ridotta (SR) le probabilità di infezione sono risultate 3.2*10-5, 1.5*10-5 e 3.1*10-6 rispettivamente nel gruppo ad alto, medio e basso rischio. La probabilità di infezione maggiore nella SR per i reparti a intermedio e basso rischio ha evidenziato una maggiore esposizione al patogeno in assenza di un monitoraggio completo di tutti i punti . La stima dei costi totali è risultata 422.900 euro per la ST e 276.700 euro per SR. Il NB per la strategia SR è risultato di 149.959 euro se la WTP è stata fissata a 25.000 euro per ogni caso in meno. Il NB per la ST invece è risultato -198.827 euro; pertanto, il costo supera il beneficio. L’EIB dato dalla differenza di NB tra SR ed ST è risultato di 348.787 euro, mostrando più favorevole la strategia SR. La SR è risultata più vantaggiosa rispetto a ST in quanto con SR i costi sono risultati più bassi e il danno che si è generato irrilevante. Con ST si è avuto un guadagno di efficacia ma spendendo eccessivamente, ben oltre i costi generati dalla non sorveglianza.
Il piano di Sicurezza dell’Acqua (PSA) negli edifici per il controllo e la prevenzione delle waterborne diseases
SPAGNUOLO, VALENTINA
2025
Abstract
During the first year, the implementation of the Water Safety Plan (WSP) was started in pavilions A and B of the Policlinico of Bari. The second year focused on the evaluation of the effectiveness of the implementation of the WSP in the water supply chains of the two pavilions and the development of a predictive model (ML) of Legionella spp. contamination in the water supply of pavilion A. Finally, two other studies were conducted in the third year. In the first, a cognitive survey was conducted in the Apulia region to assess hospitals' knowledge of D.Lgs 18/23 and the application of WSPs. In the second study, a quantitative microbial risk assessment (QMRA) was applied to estimate the infectious risk of Legionella spp., and a cost-benefit analysis was applied to evaluate the surveillance strategy (total or reduced) of the Policlinico of Bari water supply by routine microbiological monitoring. In the first study, operational monitoring showed a progressive increase in compliance with L. pneumophila and P. aeruginosa parameters in both pavilions due to local treatments on the indoor water distribution systems. In pavilion A, among the ML models tested, the most efficient had an input layer of 56 neurons, a hidden layer of 30 neurons, and an output layer of two neurons, with an accuracy of 93.4%, a sensitivity of 43.8%, and a specificity of 96%. The Poisson regression model had an accuracy of 82.9%, a sensitivity of 20.3%, and a specificity of 97.3%. The most important parameters that influenced the results of the models were type of water supply (hot/cold), filter valve replacement, and atmospheric temperature. In the second study, 38.9% of the questionnaires were collected in Apulia. In 85.7% of the cases, they were aware of Legislative Decree 18/23, the related guidelines (ISTISAN reports 22/32) and the meaning of "priority buildings." Most of them carried out risk analysis of water pipes (92.9%) and chemical-physical and microbiological tests (64.3%). Legionella testing was performed in 100% of the cases, of which 78.6% were entrusted to an accredited private laboratory. In the third study on QMRA, the probability of infection generated by positive specimens in the case of total surveillance (ST) was 2.6*10-5 in high-risk wards, 6.1*10-7 in medium-risk wards, and 2.95*10-7 in low-risk wards. In reduced surveillance (SR), the probabilities of infection were 3.2*10-5, 1.5*10-5, and 3.1*10-6 in the high-, intermediate-, and low-risk groups, respectively. The higher probability of infection in the SR for the intermediate and low-risk wards indicated greater exposure to the pathogen in the absence of complete surveillance of all points. The total cost estimate was 422,900 euros for ST and 276,700 euros for SR. The NB for the SR strategy was found to be 149,959 euros if the WTP was set at 25,000 euros less per case. The NB for ST, on the other hand, was -198,827 euros; therefore, the costs exceed the benefits. The EIB given by the difference in NB between SR and ST was 348,787 euros, showing that the SR strategy was more favorable. SR was more favorable than ST because with SR the costs were lower and the damage caused was insignificant. With ST, there was a gain in effectiveness, but overspending that far exceeded the cost of not monitoring.File | Dimensione | Formato | |
---|---|---|---|
tesi definitiva dottorato Valentina Spagnuolo.pdf
accesso solo da BNCF e BNCR
Dimensione
3.18 MB
Formato
Adobe PDF
|
3.18 MB | Adobe PDF | |
tesi definitiva dottorato Valentina Spagnuolo_1.pdf
accesso solo da BNCF e BNCR
Dimensione
3.18 MB
Formato
Adobe PDF
|
3.18 MB | Adobe PDF |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/209529
URN:NBN:IT:UNIBA-209529