Anaesthesia Coffee Break

Statistics with Stan - the basics of sensitivity, specificity, NPV and PPV and the LMA in obesity

Oct 17, 2021
Dive into the essential world of medical statistics as they unravel sensitivity, specificity, and the importance of predictive values. Discover how likelihood ratios can reshape your understanding of diagnostic tests and the implications of Bayes' theorem. They tackle the relevance of the Mallampati score and its selective use in higher-risk patients, while also discussing airway management for the morbidly obese. With intriguing insights into ROC curves and neck circumference, this discussion is packed with valuable knowledge for practitioners.
Ask episode
AI Snips
Chapters
Transcript
Episode notes
INSIGHT

Likelihood Ratios Change Odds, Not Probabilities

  • Likelihood ratios change odds and are independent of prevalence.
  • Use LR+ = sensitivity / (1 - specificity) and LR- = (1 - sensitivity) / specificity.
ADVICE

Convert Probability → Odds → Probability

  • Convert pre-test probability to odds, multiply by the likelihood ratio, then convert back to probability.
  • Use the Bayes nomogram or calculations to get post-test probability correctly.
INSIGHT

Bayes' Theorem Equals PPV With True Prevalence

  • Bayes' theorem links pre-test odds, likelihood ratio, and post-test odds; with true prevalence the theorem equals PPV.
  • Thus PPV can be derived from Bayes when study prevalence matches target population prevalence.
Get the Snipd Podcast app to discover more snips from this episode
Get the app