‘Public Access Defibrillation’ (PAD) refers to programs for distribution on the territory of Automated External Defibrillators (AED) and for training of volunteers, addressing the necessity, in occurrence of out-of-hospital cardiocirculatory arrests (OHCA), of intervention with Cardiopulmonary Resuscitation (CPR) and early defibrillation before arrival of Emergency Medical Systems (EMS). In Lombardy, a PAD program was started in 2011. Exploiting databases collecting georeferenced data about OHCAs occurred in Lombardy in 2015 and 2016, and about public AEDs, it was possible to analyse the current performance of PAD program in the city of Milan, exploiting Geographic Information Systems (GIS) for visualization and elaboration. In the analysis, catchment areas for each AED were considered: areas reachable, with walking speed, within a limited amount of time (set to 3 minutes) from the position of the AED (geometry defined ‘isochrone’, considering actual streets topology). The number of OHCAs occurring inside catchment areas was computed. The focus was set on OHCAs occurring in residential locations, that represent almost 90% of the total: development of a model estimating the residential people living in each building allowed to compute the amount of resident population living inside AEDs catchment areas. The analysis revealed that PAD program is underperforming, both in use of AEDs (less than 5% of the total cases) and placements distribution (16% of resident population coverage, 18,5% of OHCAs coverage). This study focuses on optimization of the territorial placement. A risk function was developed with a data mining approach, including demographic, socio-economic and territorial factors, in order to evaluate the risk of OHCA occurrence inside statistically meaningful unit areas (200m x 200m squared cells), and was set as a priority ranking for identification of areas where placement of an AED is required. An algorithm was implemented to program new placements according to a target performance, or pursuing the best performance under budget constraints. The novel distribution could triple the performance of the current one, considering the same budget.
Il termine ‘Public Access Defibrillation’ (PAD) indica programmi di distribuzione sul territorio di Defibrillatori Automatici Esterni (DAE) e di addestramento di volontari, al fine di intervenire, in caso di arresto cardiocircolatorio ‘out-of-hospital’ (OHCA), con rianimazione cardiopolmonare (RCP) e defibrillazione prima dell’arrivo dei soccorsi. In Lombardia, un programma PAD è stato avviato nel 2011. Grazie a database per la raccolta di dati georeferenziati riguardanti gli arresti in Lombardia (2015 e 2016) e i DAE pubblici, è stato possibile valutare la performance del PAD nella città di Milano, utilizzando dei Geographic Information Systems (GIS) per l’analisi e l’elaborazione. Nell’analisi, sono state considerate le ‘catchment areas’ dei DAE: area raggiungibile, a piedi, dalla posizione del DAE entro un limite di tempo, posto a 3 minuti (elemento definito ‘isocrona’), tenendo conto della topologia stradale. Si è calcolato il numero di arresti avvenuti all’interno delle catchment areas. L’attenzione è stata posta sugli arresti avvenuti all’interno delle abitazioni, i quali costituiscono quasi il 90% del totale: per calcolare il numero di persone residenti nelle catchment areas, è stato sviluppato un modello di stima del numero di abitanti in ogni edificio. L’analisi ha rivelato una bassa performance: i DAE sono usati in meno del 5% degli arresti, e la distribuzione non è efficiente, coprendo il 16% della popolazione e il 17% degli arresti. Lo studio è stato orientato all’ottimizzazione della distribuzione sul territorio. Con tecniche di data mining, è stata sviluppata una funzione di rischio, basata su fattori demografici, socio-economici e territoriali, per stimare l’insorgenza attesa di arresti in ogni unità di area (celle di 200m di lato); la funzione è stata utilizzata per dare un ordine di priorità alle celle rispetto alla necessità di copertura. È stato implementato un algoritmo per programmare le installazioni di nuovi DAE, a seconda dell’obiettivo posto (raggiungimento di una certa performance, miglior performance con limite di budget). La distribuzione suggerita potrebbe, a parità di budget, triplicare la performance di quella attuale.
A geographic information data-driven decision-making support system for accessibility of public defibrillators in the city of Milan
GIANQUINTIERI, LORENZO
2016/2017
Abstract
‘Public Access Defibrillation’ (PAD) refers to programs for distribution on the territory of Automated External Defibrillators (AED) and for training of volunteers, addressing the necessity, in occurrence of out-of-hospital cardiocirculatory arrests (OHCA), of intervention with Cardiopulmonary Resuscitation (CPR) and early defibrillation before arrival of Emergency Medical Systems (EMS). In Lombardy, a PAD program was started in 2011. Exploiting databases collecting georeferenced data about OHCAs occurred in Lombardy in 2015 and 2016, and about public AEDs, it was possible to analyse the current performance of PAD program in the city of Milan, exploiting Geographic Information Systems (GIS) for visualization and elaboration. In the analysis, catchment areas for each AED were considered: areas reachable, with walking speed, within a limited amount of time (set to 3 minutes) from the position of the AED (geometry defined ‘isochrone’, considering actual streets topology). The number of OHCAs occurring inside catchment areas was computed. The focus was set on OHCAs occurring in residential locations, that represent almost 90% of the total: development of a model estimating the residential people living in each building allowed to compute the amount of resident population living inside AEDs catchment areas. The analysis revealed that PAD program is underperforming, both in use of AEDs (less than 5% of the total cases) and placements distribution (16% of resident population coverage, 18,5% of OHCAs coverage). This study focuses on optimization of the territorial placement. A risk function was developed with a data mining approach, including demographic, socio-economic and territorial factors, in order to evaluate the risk of OHCA occurrence inside statistically meaningful unit areas (200m x 200m squared cells), and was set as a priority ranking for identification of areas where placement of an AED is required. An algorithm was implemented to program new placements according to a target performance, or pursuing the best performance under budget constraints. The novel distribution could triple the performance of the current one, considering the same budget.| File | Dimensione | Formato | |
|---|---|---|---|
|
2017_12_Gianquintieri.pdf
non accessibile
Descrizione: Testo della tesi
Dimensione
20.68 MB
Formato
Adobe PDF
|
20.68 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/137995