Skip to main content
Cross Validated

Return to Question

added 37 characters in body; edited title
Source Link
COOLSerdash
COOLSerdash
  • 31.2k
  • 10
  • 104
  • 157

How to calculate + / -the confidence intervals for likelihood ratios from a 2x2 table if there are 0 values (e.g. no.in the presence of false positives =0)cells with zeroes

I am analysing a diagnostic test (against a gold standard, using a 2x2 table). I want to calculate likelihood ratios (sensitivity / (1-specificity) etc) however I have several sets of data with 0 false positives therefore a specificity of 1....

I have come across other data (e.g. in evidence bases for clinical guidelines for investigation of periprosthetic infection of hip or knee by AAOS) where I know that the authors had 0 false positives but have still presented + / - LRsconfidence intervals for the likelihood ratios.

Does anyone know how they have adjusted their calculations to allow this?

Thanks

How to calculate + / - likelihood ratios from a 2x2 table if there are 0 values (e.g. no. of false positives =0)

I am analysing a diagnostic test (against a gold standard, using a 2x2 table). I want to calculate likelihood ratios (sensitivity / (1-specificity) etc) however I have several sets of data with 0 false positives therefore a specificity of 1....

I have come across other data (e.g. in evidence bases for clinical guidelines for investigation of periprosthetic infection of hip or knee by AAOS) where I know that the authors had 0 false positives but have still presented + / - LRs.

Does anyone know how they have adjusted their calculations to allow this?

Thanks

How to calculate the confidence intervals for likelihood ratios from a 2x2 table in the presence of cells with zeroes

I am analysing a diagnostic test (against a gold standard, using a 2x2 table). I want to calculate likelihood ratios (sensitivity / (1-specificity) etc) however I have several sets of data with 0 false positives therefore a specificity of 1....

I have come across other data (e.g. in evidence bases for clinical guidelines for investigation of periprosthetic infection of hip or knee by AAOS) where I know that the authors had 0 false positives but have still presented confidence intervals for the likelihood ratios.

Does anyone know how they have adjusted their calculations to allow this?

Thanks

Embedded link, readability, title
Source Link
edit approved Jun 10, 2013 at 15:11
Andre Silva
edit approved Jun 10, 2013 at 15:11
Andre Silva
  • 3.1k
  • 5
  • 33
  • 57

how How to calculate + / - likelihood ratios from a 2x2 table if there are 0 values (e.g. no. of false positives =0)

I am analysing a diagnostic test (against a gold standard, using a 2x2 table). I want to calculate likelihood ratios (sensitivity / (1-specificity) etc) however I have several sets of data with 0 false positives therefore a specificity of 1....

I have come across other datadata (e.g. in evidence bases for clinical guidelines for investigation of periprosthetic infection of hip or knee by AAOS http://www.aaos.org/research/guidelines/PJIguideline.pdf) where I know that the authors had 0 false positives but have still presented + / - LRs. 

Does anyone know how they have adjusted their calculations to allow this?

Thanks

Chloe

how to calculate + / - likelihood ratios from a 2x2 table if there are 0 values (e.g. no. of false positives =0)

I am analysing a diagnostic test (against a gold standard, using a 2x2 table). I want to calculate likelihood ratios (sensitivity / (1-specificity) etc) however I have several sets of data with 0 false positives therefore a specificity of 1....

I have come across other data (e.g. in evidence bases for clinical guidelines for investigation of periprosthetic infection of hip or knee by AAOS http://www.aaos.org/research/guidelines/PJIguideline.pdf) where I know that the authors had 0 false positives but have still presented + / - LRs. Does anyone know how they have adjusted their calculations to allow this?

Thanks

Chloe

How to calculate + / - likelihood ratios from a 2x2 table if there are 0 values (e.g. no. of false positives =0)

I am analysing a diagnostic test (against a gold standard, using a 2x2 table). I want to calculate likelihood ratios (sensitivity / (1-specificity) etc) however I have several sets of data with 0 false positives therefore a specificity of 1....

I have come across other data (e.g. in evidence bases for clinical guidelines for investigation of periprosthetic infection of hip or knee by AAOS) where I know that the authors had 0 false positives but have still presented + / - LRs. 

Does anyone know how they have adjusted their calculations to allow this?

Thanks

Source Link
Chloe T
Chloe T
  • 91
  • 1
  • 1
  • 2

how to calculate + / - likelihood ratios from a 2x2 table if there are 0 values (e.g. no. of false positives =0)

I am analysing a diagnostic test (against a gold standard, using a 2x2 table). I want to calculate likelihood ratios (sensitivity / (1-specificity) etc) however I have several sets of data with 0 false positives therefore a specificity of 1....

I have come across other data (e.g. in evidence bases for clinical guidelines for investigation of periprosthetic infection of hip or knee by AAOS http://www.aaos.org/research/guidelines/PJIguideline.pdf) where I know that the authors had 0 false positives but have still presented + / - LRs. Does anyone know how they have adjusted their calculations to allow this?

Thanks

Chloe