|
Activity Number:
|
439
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #305688 |
|
Title:
|
Comparing Diagnostic Tools: Different Methods, Same Story?
|
|
Author(s):
|
Novie Younger*+ and Audrey Pottinger
|
|
Companies:
|
Tropical Medicine Research Institute, Jamaica and University of the West Indies
|
|
Address:
|
P.O. Box 25580, Miami, FL, 33102-5580,
|
|
Keywords:
|
Agreement ; ROC anaylsis ; Kappa ; Diagnostic tools ; mental health
|
|
Abstract:
|
Diagnostic tools, mostly as symptom check lists, are used to classify mental health states. The Edinburgh Postnatal Depression Scale (EPDS) and the Beck Depression Inventory-II (BDI-II) are two such tools. In many cultures they have been validated based on postnatal data and validation using prenatal data is emerging. With prenatal data gathered in a clinical setting in Jamaica we use different statistical techniques - Bland-Altman plots, receiver-operating curve (ROC) analysis, Kappa statistics - for comparing these diagnostic tools to show that inference can vary with technique and we conclude that a variety of comparison techniques should be applied to data from diagnostic tools prior to determining the validity of a given tool for a particular setting.
|
