🔒 This topic is solved and locked. Need further help from the community? Please sign in and ask a new question.
Obsidian | Level 7

## Sensitivity specificity LR

Hi all

is there a macro/code that mimics in SAS what pROC and bootLR do in R?

my aim is to calculate bootstrap threshold, sensitivity, specificity, LT+ and LR- along with 95% CI

Tks!

1 ACCEPTED SOLUTION

Accepted Solutions
Diamond | Level 26

## Re: Sensitivity specificity LR

@Rick_SAS has written the definitive guide to bootstrapping in SAS

https://blogs.sas.com/content/iml/2018/12/12/essential-guide-bootstrapping-sas.html

--
Paige Miller
8 REPLIES 8
Diamond | Level 26

## Re: Sensitivity specificity LR

@antor82 wrote:

is there a macro/code that mimics in SAS what pROC and bootLR do in R?

What do pROC and bootLR do?

--
Paige Miller
Obsidian | Level 7

## Re: Sensitivity specificity LR

bootLR Computes the (positive/negative) likelihood ratio with appropriate, bootstrapped confidence intervals

pROC computes threshold sensitivity and specificity for a given value with appropriate, bootstrapped confidence intervals
Diamond | Level 26

## Re: Sensitivity specificity LR

Normally, a likelihood ratio is computed to test to see if a statistic is significantly different than some quantity ... what statistic is being tested, and in what context?

Normally, bootstrap confidence intervals are computed to see some measure of variability of a particular statistic ... what statistic is being tested, and in what context?

--
Paige Miller
Obsidian | Level 7

## Re: Sensitivity specificity LR

I’m testing if a new parameter is predictive of an outcome.
My sample size is small (60 patients).
Other authors did the same and provided AUC, cut-off (with corresponding sensibility, specificity, LR+ and -) calculated after bootstrapping
Diamond | Level 26

## Re: Sensitivity specificity LR

I’m testing if a new parameter is predictive of an outcome.

Some sort of regression??? A decision tree??? A Neural Network??? What is the exact type of model?

Other authors did the same and provided AUC, cut-off (with corresponding sensibility, specificity, LR+ and -) calculated after bootstrapping

Logistic regression?? Decision tree?? Neural Network?? Other???

--
Paige Miller
Obsidian | Level 7

## Re: Sensitivity specificity LR

Logistic regression
Diamond | Level 26

## Re: Sensitivity specificity LR

@Rick_SAS has written the definitive guide to bootstrapping in SAS

https://blogs.sas.com/content/iml/2018/12/12/essential-guide-bootstrapping-sas.html

--
Paige Miller
Obsidian | Level 7

## Re: Sensitivity specificity LR

Tks!
Discussion stats
• 8 replies
• 1245 views
• 0 likes
• 2 in conversation