takhona/malaria-drone-cea

Cost-Effectiveness of Drone-Based Commodity Delivery Compared to Land-based Delivery in Remote Madagascar # malaria-drone-cea Cost-Effectiveness of Drone-Based Commodity Delivery Compared to Land-based Delivery in Remote Madagascar # Cost-Effectiveness of Drone-Delivered Malaria Interventions in Madagascar ## Overview This repository contains the Python-based model and analytical code used to evaluate the cost-effectiveness of drone delivery for malaria testing and treatment commodities. The analysis is based on a Markov cohort model evaluating clinical and economic outcomes for children under 5 years of age in a rural Malagasy population. The model compares standard trial-informed supply chains against various drone scale-up scenarios. ## Repository Structure The analysis is modularized into three primary execution files, intended to be run in the following sequence: * `1_Copy_of_Inputs_Drone_CEA.ipynb` * Generates the tables of the key epidemiological, cost, and health-system parameters used in the model. * `2_Copy_of_Deterministic_Drone_CEA.ipynb` * Contains the core deterministic Markov cohort model. * Calculates base-case transition probabilities, tracks the service pathway (testing and prompt treatment), and calculates deterministic Incremental Cost-Effectiveness Ratios (ICERs) and DALYs averted over 1-year and 10-year time horizons. * Includes the epidemiological face-validation exercise comparing model incidence outputs against empirical bounds from routine facility data. * `3_Copy_of_PSA_Drone_CEA.ipynb` * Executes the Probabilistic Sensitivity Analysis (PSA). * Propagates parameter uncertainty through the Markov model via Monte Carlo simulation to generate cost-effectiveness acceptability curves (CEACs) and CEA plane. ## Usage and Reproducibility The code is written in Python 3 and was developed using Google Colab. The easiest way to reproduce these results is to open the `.ipynb` files directly in Google Colab or Jupyter Notebook. **Dependencies:** * `numpy` * `pandas` * `matplotlib` / `seaborn` (for figures) ## Citation If you use this code or model …