| Author/Editor | Malkić, Jasmin; Sarajlić, Nermin; Rejec Smrke, Barbara Urška; Smrke, Dragica | |
| Title | Cluster analysis of association of weather, daily factors and emergent medical condition | |
| Translated title | Klasterska analiza vremenskih i dnevnih čimbenika povezanih uz brojnost hitnih intervencija | |
| Type | članek | |
| Vol. and No. | Letnik 37, št. 1 | |
| Publication year | 2013 | |
| Volume | str. 189-194 | |
| ISSN | 1848-9486 - Collegium antropologicum | |
| Language | eng | |
| Abstract | The goal of this study was to evaluate associations between the meteorologicalconditions and the number of emergency cases for five distinctive causes of dispatch groups reported to SOS dispatch centre in Uppsala, Sweden. Center¼s responsibility include alerting to 17 ambulances in whole Uppsala County, area of 8,209 kmŽ with around 320,000 inhabitants representing the target patient group. Source of the medical data for this study is the database of dispatch data for the year of 2009, while the metrological data have been provided from Uppsala University Department of Earth Sciences yearly weather report. Medical and meteorological data were summoned into the unified data space where each point represents a day with its weather parameters and dispatch cause group cardinality. DBSCAN data mining algorithm was implemented to five distinctive groups of dispatch causesafter the data spaces have gone through the variance adjustment and the principal component analyses. As the result, several point clusters were discovered in each of the examined data spaces indicating the distinctive conditions regarding the weather and daily cardinality of the dispatch cause, as well as the associations between these two. Most interesting finding is that specific type of winter weather formed a cluster only around the days with the high count of breathing difficulties, while one of the summer weatherclusters made similar association with the days with low number of cases. Findings were confirmed by confidence level estimation based on signal to noise ratio for the observed data points. | |
| Keywords | emergency department visit primary health care meteorology data mining Uppsala |