Dataset and Code for Risk-Informed Data Analytics for Sustainable Pharmaceutical Supply: A Governance Framework for Public Oncology Hospitals

This repository contains all raw operational data layers and reproducible code used in the study “Risk-Informed Data Analytics for Sustainable Pharmaceutical Supply: A Governance Framework for Public Oncology Hospitals.” The materials hosted here enable full transparency and end-to-end replication of the analyses, figures, and before–after inventory parameter tables presented in the manuscript.

Contents

The repository includes the following components:

1. Raw operational datasets (Excel)

These files correspond to real institutional data from a public oncology hospital and cover multiple layers of the medication supply chain:

• ABC/XYZ classifications and institutional stock parameters (base sss, SSSSSS, QQQ)
• Monthly consumption data (2023–2024)
• Purchase lead times and arrival patterns by procurement channel
• Inventory levels and overstock valuations
• Stock maximum/replenishment settings (as defined by institutional governance)
• Item master data including price, formulation, and subunit
• Records of stockouts, mermas, and operational events when applicable

All data are strictly operational and contain no patient-level or identifiable information.

2. Reproducible code

A minimal, executable pipeline is provided to regenerate the preprocessing steps and the illustrative before–after table of the continuous-review inventory parameters (s,SS,Q)(s, SS, Q)(s,SS,Q) used in Module 7 of the manuscript. Contents include:

• zenodo_example_pipeline.py — main Python script for data harmonization, merging, prioritization, and computation of illustrative (snew,SSnew,Qnew)(s_{new}, SS_{new}, Q_{new})(snew,SSnew,Qnew).
• requirements.txt — exact dependencies (pandas, numpy, openpyxl).
• README_zenodo_example.md — instructions for executing the pipeline.
• Optional merging of user-supplied optimized parameters through Parámetros de stock.xlsx.

The code is designed to be robust to inconsistent column names, missing optional files, and heterogeneous operational sources.

3. Documentation

Additional notes explain:

• Expected schema of each dataset
• How the data layers interact
• How the code regenerates the tables reported in the paper
• How to extend or adapt the pipeline for alternative prioritizations or governance rules

Purpose

The goal of this repository is to support:

• Transparency of data-driven public-sector decision-making
• Reproducibility of the risk-informed analytical workflow
• Validation of the governance framework proposed in the manuscript
• Future methodological extensions in sustainable pharmaceutical supply chain management

License and ethical considerations

All files are released for scientific and academic use. The dataset contains administrative operational records only; no patient information is included.