Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Negative Predictive Value

From   Ronan Conroy <>
To   "" <>
Subject   Re: st: Negative Predictive Value
Date   Mon, 5 Sep 2011 10:27:00 +0100

On 2011 MFómh 4, at 18:17, Jayson Trinchera wrote:

> How do you compute for the minimum required sample size for
> determining the negative predictive values of two diagnostic methods
> with a binary outcome/result (either positive or negative)?

The negative predictive value is a function of the prevalence, so the sample size calculation will critically involve prevalence. 

A more general approach is based on clinical likelihood ratios, which are independent of prevalence. A useful reference is

J Clin Epidemiol. 1991;44(8):763-70.
Likelihood ratios with confidence: sample size estimation for diagnostic test studies.
Simel DL, Samsa GP, Matchar DB.

You might note also that there is a user-written command -fagan- (by Ben Adarkwa Dwamena)which allows you to convert likelihood ratios into post-test probabilities at any given prevalence. 

Ronán Conroy
Associate Professor
Division of Population Health Sciences
Royal College of Surgeons in Ireland
Beaux Lane House
Dublin 2

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index