Literature DB >> 29039373

DisEpi: Compact Visualization as a Tool for Applied Epidemiological Research.

Arriel Benis1, Moshe Hoshen1.   

Abstract

Outcomes research and evidence-based medical practice is being positively impacted by proliferation of healthcare databases. Modern epidemiologic studies require complex data comprehension. A new tool, DisEpi, facilitates visual exploration of epidemiological data supporting Public Health Knowledge Discovery. It provides domain-experts a compact visualization of information at the population level. In this study, DisEpi is applied to Attention-Deficit/Hyperactivity Disorder (ADHD) patients within Clalit Health Services, analyzing the socio-demographic and ADHD filled prescription data between 2006 and 2016 of 1,605,800 children aged 6 to 17 years. DisEpi's goals facilitate the identification of (1) Links between attributes and/or events, (2) Changes in these relationships over time, and (3) Clusters of population attributes for similar trends. DisEpi combines hierarchical clustering graphics and a heatmap where color shades reflect disease time-trends. In the ADHD context, DisEpi allowed the domain-expert to visually analyze a snapshot summary of data mining results. Accordingly, the domain-expert was able to efficiently identify that: (1) Relatively younger children and particularly youngest children in class are treated more often, (2) Medication incidence increased between 2006 and 2011 but then stabilized, and (3) Progression rates of medication incidence is different for each of the 3 main discovered clusters (aka: profiles) of treated children. DisEpi delivered results similar to those previously published which used classical statistical approaches. DisEpi requires minimal preparation and fewer iterations, generating results in a user-friendly format for the domain-expert. DisEpi will be wrapped as a package containing the end-to-end discovery process. Optionally, it may provide automated annotation using calendar events (such as policy changes or media interests), which can improve discovery efficiency, interpretation, and policy implementation.

Entities:  

Keywords:  ADHD; Clustering; Epidemiology; Facilitation; Heatmap; Public Health Informatics; Visual Data Mining

Mesh:

Year:  2017        PMID: 29039373

Source DB:  PubMed          Journal:  Stud Health Technol Inform        ISSN: 0926-9630


  3 in total

1.  A Prerequisite for Patient Centred Care is Empowering and Engaging Patients in the Digital Health: Report from EFMI Special Topic Conference (STC) 2017.

Authors:  Simon de Lusignan
Journal:  Acta Inform Med       Date:  2017-12

Review 2.  Visual Analytic Tools and Techniques in Population Health and Health Services Research: Scoping Review.

Authors:  Jawad Ahmed Chishtie; Jean-Sebastien Marchand; Luke A Turcotte; Iwona Anna Bielska; Jessica Babineau; Monica Cepoiu-Martin; Michael Irvine; Sarah Munce; Sally Abudiab; Marko Bjelica; Saima Hossain; Muhammad Imran; Tara Jeji; Susan Jaglal
Journal:  J Med Internet Res       Date:  2020-12-03       Impact factor: 5.428

3.  Trends in the Prevalence of Chronic Medication Use Among Children in Israel Between 2010 and 2019: Protocol for a Retrospective Cohort Study.

Authors:  Yair Sadaka; Dana Horwitz; Arriel Benis; Leor Wolff; Tomer Sela; Joseph Meyerovitch; Assaf Peleg; Eitan Bachmat
Journal:  JMIR Res Protoc       Date:  2022-08-05
  3 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.