Biostatistics Analysis Tool

Perform common biostatistical analyses for medical research including survival analysis (Kaplan-Meier, Cox proportional hazards), logistic and linear regression, ROC curve analysis with AUC calculation, sample size and power calculations, and diagnostic test performance metrics. The skill guides researchers through appropriate test selection, assumption checking, result interpretation, and reporting in accordance with published statistical reporting guidelines.

Quick Install

npx skills add Open-Medica/open-medical-skills --skill biostatistics-analyzer

What It Does

• Statistical test selection: Given a research question, data structure (continuous/categorical, paired/unpaired, parametric/non-parametric), and study design, recommends the appropriate statistical test and explains why alternatives were not selected
• Survival analysis: Performs Kaplan-Meier estimation with log-rank test comparison between groups, Cox proportional hazards regression with hazard ratio calculation and proportional hazards assumption testing (Schoenfeld residuals)
• Regression modeling: Conducts logistic regression (binary, ordinal, multinomial) and linear regression with model diagnostics, effect size reporting (odds ratios with 95% CI for logistic; beta coefficients for linear), multicollinearity assessment (VIF), and model fit statistics (C-statistic, R-squared, Hosmer-Lemeshow)
• Diagnostic test evaluation: Calculates sensitivity, specificity, positive/negative predictive values, positive/negative likelihood ratios, and generates ROC curves with AUC and optimal threshold determination (Youden's index)
• Sample size and power: Computes required sample sizes for common study designs (two-group comparison of means, two-group comparison of proportions, survival analysis, non-inferiority/equivalence trials) and performs post-hoc power analysis for completed studies

Clinical Use Cases