Planning a sample for an epidemiological survey

Tomčíková, Daniela ; Cocchi, Daniela ; Bordoni, Barbara ; Marzocchi, Antonio (2011) Planning a sample for an epidemiological survey. Bologna, IT: Dipartimento di Scienze Statistiche "Paolo Fortunati", Alma Mater Studiorum Università di Bologna, p. 38. DOI 10.6092/unibo/amsacta/3220. In: Quaderni di Dipartimento. Serie Ricerche ISSN 1973-9346.
Full text available as:
[img]
Preview
PDF
License: Creative Commons Attribution Non-commercial No Derivatives

Download (1MB) | Preview

Abstract

This work illustrates the joint use of a pilot study and an administrative data base for designing a probabilistic sample for an epidemiological survey. The target is to estimate the prevalence of an asymptomatic disease, the aortic valve stenosis (AS), in the elderly population of the city of Bologna, Italy. The novelty of the study is to reach the target population of elderly patients via a sample of their general practitioners (GPs). The pilot study was conducted in San Giovanni in Persiceto, a town in the province of Bologna. Overall information on patients and their GPs are available in the Azienda Unità Sanitaria Locale di Bologna (AUSL) data sets. Since the disease is asymptomatic, the sampling plan is designed to estimate the number of suspected patients that will be sent to further echocardiographic (ECO) examination. The probabilistic sampling plan aims at controlling the sources of randomness, via an appropriate clustering of the population of GPs. The number of practitioners to sample is fixed in advance. The subpopulations of patients to screen are also defined in advance and assigned to doctors. In this way the potential sources of randomness, due to the individual choices of doctors out of the definition of the experiment, are avoided. The number of elderly patients per doctor has been identified, from the pilot study, as an important factor able to influence the proportion of suspected patients sent to further examination. This feature is the leading factor of the sampling design, together with the clustering of the AUSL Bologna territory in NCPs, which emerges from the AUSL data set.

Abstract
Document type
Monograph (Working Paper)
Creators
CreatorsAffiliationORCID
Tomčíková, Daniela
Cocchi, Daniela
Bordoni, Barbara
Marzocchi, Antonio
Keywords
Survey sampling, ratio and regression sampling estimator, finite population, epidemiology, cardiology
Subjects
ISSN
1973-9346
DOI
Deposit date
14 Feb 2012 11:18
Last modified
14 Feb 2012 11:18
URI

Other metadata

Downloads

Downloads

Staff only: View the document

^