Code for: Cost-Effective Case Finding Strategies for Cognitive Impairment with Digital Twin Simulations

This repository contains the analysis code for the manuscript "Cost-Effective Case Finding Strategies for Cognitive Impairment with Digital Twin Simulations," published in Nature Health. The study uses a 'digital twin' of Singapore’s older adult population (n=753,905), simulated using a deep learning model (Generative Adversarial Network). This digital twin was used to evaluate a comprehensive range of case-finding strategies for cognitive impairment, identifying which approaches can best balance diagnostic reach and cost-effectiveness. The findings provide a tiered framework for policymakers to implement screening programs based on resource availability and willingness-to-pay.