Applying a hierarchical clustering on principal components approach to identify different patterns of the SARS-CoV-2 epidemic across Italian regions
Abstract Italy has experienced the epidemic of Severe Acute Respiratory Syndrome Coronavirus 2, which spread at different times and with different intensities throughout its territory. We aimed to identify clusters with similar epidemic patterns across Italian regions. To do that, we defined a set o...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2021-03-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-021-86703-3 |
| _version_ | 1818420369840668672 |
|---|---|
| author | Andrea Maugeri Martina Barchitta Guido Basile Antonella Agodi |
| author_facet | Andrea Maugeri Martina Barchitta Guido Basile Antonella Agodi |
| author_sort | Andrea Maugeri |
| collection | DOAJ |
| description | Abstract Italy has experienced the epidemic of Severe Acute Respiratory Syndrome Coronavirus 2, which spread at different times and with different intensities throughout its territory. We aimed to identify clusters with similar epidemic patterns across Italian regions. To do that, we defined a set of regional indicators reflecting different domains and employed a hierarchical clustering on principal component approach to obtain an optimal cluster solution. As of 24 April 2020, Lombardy was the worst hit Italian region and entirely separated from all the others. Sensitivity analysis—by excluding data from Lombardy—partitioned the remaining regions into four clusters. Although cluster 1 (i.e. Veneto) and 2 (i.e. Piedmont and Emilia-Romagna) included the most hit regions beyond Lombardy, this partition reflected differences in the efficacy of restrictions and testing strategies. Cluster 3 was heterogeneous and comprised regions where the epidemic started later and/or where it spread with the lowest intensity. Regions within cluster 4 were those where the epidemic started slightly after Veneto, Emilia-Romagna and Piedmont, favoring timely adoption of control measures. Our findings provide policymakers with a snapshot of the epidemic in Italy, which might help guiding the adoption of countermeasures in accordance with the situation at regional level. |
| first_indexed | 2024-12-14T12:53:23Z |
| format | Article |
| id | doaj.art-d85f1c2647a14bc1ad59c8efc94a4782 |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-12-14T12:53:23Z |
| publishDate | 2021-03-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-d85f1c2647a14bc1ad59c8efc94a47822022-12-21T23:00:38ZengNature PortfolioScientific Reports2045-23222021-03-011111910.1038/s41598-021-86703-3Applying a hierarchical clustering on principal components approach to identify different patterns of the SARS-CoV-2 epidemic across Italian regionsAndrea Maugeri0Martina Barchitta1Guido Basile2Antonella Agodi3Department of Medical and Surgical Sciences and Advanced Technologies “GF Ingrassia”, University of CataniaDepartment of Medical and Surgical Sciences and Advanced Technologies “GF Ingrassia”, University of CataniaDepartment of General Surgery and Medical-Surgical Specialties, University of CataniaDepartment of Medical and Surgical Sciences and Advanced Technologies “GF Ingrassia”, University of CataniaAbstract Italy has experienced the epidemic of Severe Acute Respiratory Syndrome Coronavirus 2, which spread at different times and with different intensities throughout its territory. We aimed to identify clusters with similar epidemic patterns across Italian regions. To do that, we defined a set of regional indicators reflecting different domains and employed a hierarchical clustering on principal component approach to obtain an optimal cluster solution. As of 24 April 2020, Lombardy was the worst hit Italian region and entirely separated from all the others. Sensitivity analysis—by excluding data from Lombardy—partitioned the remaining regions into four clusters. Although cluster 1 (i.e. Veneto) and 2 (i.e. Piedmont and Emilia-Romagna) included the most hit regions beyond Lombardy, this partition reflected differences in the efficacy of restrictions and testing strategies. Cluster 3 was heterogeneous and comprised regions where the epidemic started later and/or where it spread with the lowest intensity. Regions within cluster 4 were those where the epidemic started slightly after Veneto, Emilia-Romagna and Piedmont, favoring timely adoption of control measures. Our findings provide policymakers with a snapshot of the epidemic in Italy, which might help guiding the adoption of countermeasures in accordance with the situation at regional level.https://doi.org/10.1038/s41598-021-86703-3 |
| spellingShingle | Andrea Maugeri Martina Barchitta Guido Basile Antonella Agodi Applying a hierarchical clustering on principal components approach to identify different patterns of the SARS-CoV-2 epidemic across Italian regions Scientific Reports |
| title | Applying a hierarchical clustering on principal components approach to identify different patterns of the SARS-CoV-2 epidemic across Italian regions |
| title_full | Applying a hierarchical clustering on principal components approach to identify different patterns of the SARS-CoV-2 epidemic across Italian regions |
| title_fullStr | Applying a hierarchical clustering on principal components approach to identify different patterns of the SARS-CoV-2 epidemic across Italian regions |
| title_full_unstemmed | Applying a hierarchical clustering on principal components approach to identify different patterns of the SARS-CoV-2 epidemic across Italian regions |
| title_short | Applying a hierarchical clustering on principal components approach to identify different patterns of the SARS-CoV-2 epidemic across Italian regions |
| title_sort | applying a hierarchical clustering on principal components approach to identify different patterns of the sars cov 2 epidemic across italian regions |
| url | https://doi.org/10.1038/s41598-021-86703-3 |
| work_keys_str_mv | AT andreamaugeri applyingahierarchicalclusteringonprincipalcomponentsapproachtoidentifydifferentpatternsofthesarscov2epidemicacrossitalianregions AT martinabarchitta applyingahierarchicalclusteringonprincipalcomponentsapproachtoidentifydifferentpatternsofthesarscov2epidemicacrossitalianregions AT guidobasile applyingahierarchicalclusteringonprincipalcomponentsapproachtoidentifydifferentpatternsofthesarscov2epidemicacrossitalianregions AT antonellaagodi applyingahierarchicalclusteringonprincipalcomponentsapproachtoidentifydifferentpatternsofthesarscov2epidemicacrossitalianregions |