OLAP for health statistics: how to turn a simple spreadsheet into a powerful analytical tool


Artnik, Barbara, Vidmar, Gaj and Javornik, Jana (2002) OLAP for health statistics: how to turn a simple spreadsheet into a powerful analytical tool. In, Tkačik, Boris (ed.) 12th Statistical Days - Proceedings. 12th Conference on Integration of Statistics in the Information Society Radenci, Slovenia, Statistical Society of the Republic of Slovenia, 476-481.

Download

Full text not available from this repository.

Description/Abstract

Over the last ten years, Online Analytical Processing (OLAP) has become a very popular tool for interactive analysis of multidimensional information. Providing online operation and flexible summarising, tabulating and charting options, it has become an essential part of the decision support process in corporate setting. Our aim is to demonstrate how easily applicable and useful OLAP can be in the public sector. To achieve that, we used data compiled from different sources for the purpose of exploring the relations between causes of death (according to ICD-10) and socio-economic characteristics (educational level, marital status, profession, etc.) for selected years 1992, 1995 and 1998 in Slovenia. Using a standard personal computer and the Windows® platform, the application was implemented in Microsoft® Excel 2000, without any programming. After data cleansing (elimination of incorrect entries, duplicates and inconsistencies based on exploratory statistical methods), the case-based spreadsheet data was instantly converted into an OLAP application with the user-friendly pivot table technology. A bonus of this approach is that the results can be made directly accessible over the WWW by publishing the workbook to a web server. Provided that the user has Microsoft® Internet Explorer and Microsoft® Office 2000 installed, all the drill-in, drill-out and dimension-swapping capabilities are accessible within the browser, while the data source remains fully protected. Privacy constraints are respected since all the information is only provided at the aggregate level. Even though our dataset provides exhaustive coverage of mortality at the national level, storage and processing capabilities did not prove to be an issue. Hence, we argue that OLAP methodology should find a place in health statistics. With proper data collection and/or transformations, informative comparisons within the study population and with international databases become readily accessible. The key advantage of OLAP over relational database management systems and ordinary tables is interactive browsing of multidimensional and hierarchical data, while OLAP can also aid data integrity checking and reporting.

Item Type: Book Section
Related URLs:
Subjects: H Social Sciences > HA Statistics
R Medicine > RA Public aspects of medicine
Divisions: University Structure - Pre August 2011 > School of Social Sciences > Sociology and Social Policy
ePrint ID: 42836
Date Deposited: 15 Feb 2007
Last Modified: 27 Mar 2014 18:27
URI: http://eprints.soton.ac.uk/id/eprint/42836

Actions (login required)

View Item View Item