mugabepatricia/Maternal-Child-Health-Data-Analysis

This project demonstrates an end-to-end data analysis workflow applied to a realistic Maternal and Child Health (MCH) dataset simulating clinical records from 8 public health facilities across Kenya # Maternal & Child Health Data-Analysis (End-to-End Excel Project) ## Project Overview This project demonstrates an end-to-end data analysis workflow applied to a realistic Maternal and Child Health (MCH) dataset simulating clinical records from 8 public health facilities across Kenya. The dataset covers 250 patient admissions across 2022–2023 and was intentionally constructed with real-world data quality issues — the kind routinely encountered in health information systems. The project was completed entirely in Microsoft Excel, covering data auditing, cleaning, transformation, and interactive dashboard development — without the use of any programming language or BI tool. ## Dataset Overview - Source: Simulated clinical MCH records — Kenya, 2022–2023 - Records: 250 patient admissions (post-deduplication: 245) - Facilities: 8 county and national referral hospitals - Counties: Nairobi, Mombasa, Kisumu, Nakuru, Kakamega, Uasin Gishu, Garissa - Variables: 29 columns covering maternal demographics, ANC, delivery, and neonatal outcomes - Time Period: January 2022 – December 2023 ## Data Quality Issues Identified and Resolved - Delivery_Mode: 15+ entry variants --> Standardised to SVD / C-Section / Assisted Vaginal using IF/OR - Baby_Gender: M, F, Boy, Girl, blanks --> Standardised to Male / Female / Unknown - PPH, Eclampsia, Stillbirth, NICU, Maternal Death: YES, yes, Y, 1, 0, blanks --> Recoded to Yes / No / Unknown using nested IF - Mother_Age_Yrs: Outliers (0, 9, 55), blanks --> Flagged Invalid using IFS with range 14–45 - Parity: Blanks, N/A strings --> Derived from Gravida–1 using IF/OR(ISBLANK) - Birth_Weight_g: 0, 9999 entry errors, blanks --> Flagged Invalid outside 400–5500g range - Haemoglobin_gdL: 0, 'pending', blanks --> Labelled Not Recorded using OR(ISBLANK, =0, ='pending') - ANC_Visits: Negatives, N/A, blanks --> Flagged Unknown for any value Flagged Invalid; cross-validated against delivery mode - Duplicate Rows: 5 exact duplicates --> Identified and re …