Diagnostic accuracy (sensitivity, specificity, PPV, NPV, likelihood ratios), post-test probability, and effect sizes (NNT, NNH, ABI, RBI).
Absolute difference in event rates between control and treatment.
Calculates common epidemiological values for diagnostic tests and treatments, including sensitivity, specificity, predictive values, likelihood ratios, and treatment effectiveness measures.
Sensitivity, specificity, PPV, NPV, accuracy, likelihood ratios from a 2x2 table.
FNR = 1 - sensitivity.
Expected false-negative count given sensitivity, prevalence and sample size.
FPR = 1 - specificity.
Expected false-positive count given specificity, prevalence and sample size.
Positive likelihood ratio from sensitivity and specificity: how much a positive test increases disease odds.
Negative likelihood ratio: LR- = (1 − sensitivity) / specificity. A small LR- (<0.1) strongly rules out disease when the test is negative.
Negative predictive value from sensitivity, specificity, and pre-test probability (prevalence): NPV = [specificity × (1 − prevalence)] / [specificity × (1 − prevalence) + (1 − sensitivity) × prevalence]. Merges the duplicate plain-NPV entry.
Number needed to treat (NNT) or harm (NNH) from event rates: enter control and experimental event rates as proportions. If experimental < control → NNT = 1/(CER−EER); if greater → NNH = 1/(EER−CER). Merges the duplicate plain-NNT entry.
Convert a probability (0-1) to odds: odds = p / (1 − p).
Converts a standard normal Z-score to percentile.
LR+ = sensitivity / (1 - specificity); how much a positive test increases disease odds.
Updates disease odds after a test result: post-test odds = pre-test odds x likelihood ratio.
PPV from sensitivity, specificity and disease prevalence (Bayes theorem).
Converts odds to probability.
Relative increase in good outcomes with treatment vs control.
Converts log-odds (logit) to probability: p = 1/(1+e^-x).
Sample size per group to compare two proportions (two-sided).
Sensitivity = true positives / (true positives + false negatives).
Calculates diagnostic test specificity from raw 2x2 table counts: the proportion of truly disease-free individuals correctly identified as negative.
Derives the expected number of true negatives in a screened population from test specificity and disease prevalence: TN = specificity × N × (1 − prevalence). Useful for planning screening programs and understanding absolute false-positive burden.
Derives the expected number of true positives in a screened population from test sensitivity and disease prevalence: TP = sensitivity × N × prevalence. Useful for planning screening programs and estimating case detection.
Clinical Decision Support (CDS) Notice & Legal Disclaimer
All calculations and clinical scoring tools in this directory are intended strictly for educational and decision-support use by licensed medical practitioners. They do not constitute autonomous medical advice or direct treatment orders. See the full DexaMed Medical Disclaimer.