equiflow

What problem does it solve?

Generates equity-focused cohort flow diagrams that track demographic, socioeconomic, and outcome variables at each exclusion step and compute standardized mean differences to detect selection bias.

Core Features & Use Cases

• Auto-detection of equity variables from your dataset (gender/age/race/etc.)
• Produce CONSORT-style diagrams and SMD bias checks across exclusion steps
• Works with clinical datasets (e.g., MIMIC-IV, eICU) and supports custom cohorts
• Python API: CohortFlow wrapper to build, filter, and visualize the cohort with PDF output

Quick Start

Create a CohortFlow from your dataframe, apply exclusions with exclude(), and generate the visualization by calling plot().