fatimanaveeds/health-nutrition-db

Built a PostgreSQL database to track child malnutrition across Mozambique using real UNICEF survey data. Includes an auto-classification trigger (SAM/MAM/Normal), a Python data pipeline, and analytical queries for 7-year trend monitoring. # Mozambique Health & Nutrition Monitoring Database A production-ready PostgreSQL database system for tracking child malnutrition across Mozambique — built to support long-term health monitoring programs like AKF's 7-year nutrition initiative in Cabo Delgado. ## Project Overview Child malnutrition is one of the leading causes of under-5 mortality in Mozambique. This project builds the data infrastructure needed to: - Track individual child assessments over time - Auto-classify nutrition status (Normal / MAM / SAM) using WHO standards - Identify high-risk years, regions, and trends - Support 7-year longitudinal monitoring ## Database Schema 4 relational tables built in PostgreSQL: | Table | Description | |------------------|----------------------------------------------| | `facilities` | Health clinics across Mozambique districts | | `health_workers` | Field staff linked to facilities | | `beneficiaries` | Children registered in the program | | `assessments` | Every visit, measurement, and classification | ### Entity Relationship: ``` facilities ↑ health_workers ──→ assessments ←── beneficiaries ``` --- ## Key Features ### Auto-Classification Trigger Nutrition status is automatically calculated on every INSERT using WHO MUAC thresholds — no manual entry required: ```sql WHEN muac_cm > 12.5 THEN 'Normal' WHEN muac_cm >= 11.5 THEN 'MAM' -- Moderate Acute Malnutrition ELSE 'SAM' -- Severe Acute Malnutrition ``` ### Data Pipeline Real UNICEF/WHO malnutrition survey data sourced from Kaggle, cleaned and loaded via Python: - Handled missing values using column mean imputation - Simulated 93 additional records following real statistical trends - 100 total records with realistic SAM/MAM/Normal distribution --- ## Analytical Queries 5 production-ready queries in `sql/queries.sql`: | Query | Purpose | | …