V magistrski nalogi smo proučevali povprečni čakalni čas od triažnega pregleda to začetka obravnave na urgencah v Trstu in Vidmu. Podatki o povprečnih čakalnih časih v italijanski pokrajini Furlanija – Julijska krajina so dostopni na javni spletni strani
https://servizionline.sanita.fvg.it/psonline/#/index. Podatki so se zbirali med 24.12.2016 in 19.1.2017. Podatki o čakalnih časih se posodabljajo vsakih 10 minut in vsebujejo informacije o številu pacientov v čakalnici, številu pacientov na obravnavi in povprečni čakalni čas za vsako urgenco v tej pokrajini ter kategorijo ogroženosti znotraj posamezne urgence. Opazili smo, da so povprečni čakalni časi na tej spletni strani napačno izračunani, saj pri izračunu upoštevajo čakalne čase pacientov v čakalnici v enaki meri, kot čakalne čase pacientov na obravnavi. Posledično objavljeni povprečni čakalni časi podcenijo realno vrednost. Problem napačno izračunanega povprečnega čakalnega časa je natančneje opisan že v uvodnem poglavju. Za namen pravilnega izračuna povprečnega čakalnega časa je bilo potrebno dobiti čakalne čase za posamezne paciente. Naredili smo algoritem, ki javno dostopne podatke preoblikuje tako, da imamo za vsakega pacienta na voljo podatke o datumu in uri triažnega pregleda ter začetka in konca obravnave. Na podlagi ocenjenih čakalnih časov posameznikov smo ocenili povprečni čakalni čas. Zaradi pomanjkanja informacij algoritem pri preoblikovanju uporablja določene predpostavke, zaradi česar so ocenjeni povprečni čakalni časi rahlo precenjeni. Precenjenost smo ocenili s simulacijo. V nadaljevanju smo za napoved povprečnih čakalnih časov predlagali regresijske modele. Rezultati regresijskih modelov so potrdili, da obstajajo razlike v čakalnih časih med videmsko in tržaško urgenco, na dolžino čakalnega časa pa najbolj vpliva število pacientov v čakalnici.
| Jezik: | Angleški jezik |
|---|
| Naslov: | Estimation and modelling of waiting times in Emergency Departments of the Friuli Venezia Giulia Region |
|---|
| Izvleček: |
|---|
The master thesis is focused on the analysis of the average waiting time for patients from triage evaluation to treatment in the emergency rooms of Trieste and Udine. The thesis is based on the publicly available data provided by the Region Friuli Venezia Giulia (on the web page https://servizionline.sanita.fvg.it/psonline/#/index); data were collected between 24th of December 2016 and 19th of January 2017. The publicly available data are updated every 10 minutes and include the number of patients in waiting room, the number of patients in treatment and their average waiting time. We observed that the publicly available average waiting times calculations are wrong, as the waiting times of patients in the waiting room and those in treatment are given the same weight. The reported average waiting times thus underestimate the real waiting time. The problem of the wrongly calculated average waiting times is described in detail, as is the data collection process. We propose an algorithm for transforming the publicly available data into suitable data for the correct estimation patients' waiting times. The algorithm transforms the aggregated data into individual records, estimating for each patient the date and time of his or her triage evaluation, beginning of treatment and discharge from emergency room. The algorithm is based on some assumptions that make the estimated average times slightly overestimated. We quantify the magnitude of the overestimation with a series of simulation studies. Based on the reconstructed individual histories we estimate the average waiting times; moreover, we propose a set of regression models that can be used for predicting the average times for new patients. The results of regression models show that there is a difference in waiting times between emergency rooms of Cattinara and Udine. The number of patients in the waiting room has the greatest impact on the waiting time.
|
| Ključne besede: | average waiting time, emergency, triage evaluation, patient flow, statistical analysis of waiting times, modelling of waiting times, linear regression, quantile regression |
|---|
