HealthConnect appointment analysis project , AnalystLab Africa Week 4
# AnalystLab Africa – Week 4 HealthConnect
## Project Overview
This project is part of the **AnalystLab Africa Experience Lab Internship Programme – Week 4**.
The project focuses on **HealthConnect**, a healthcare appointment analysis case study. The main objective is to understand appointment attendance patterns, identify factors that may contribute to missed appointments, and explore opportunities to improve patient support.
## Project Objective
The analysis is designed to help HealthConnect:
* Understand appointment attendance patterns.
* Identify factors associated with missed appointments.
* Monitor important appointment performance indicators.
* Identify patient groups or appointment situations that may require additional support.
* Support data-driven decisions aimed at reducing missed appointments.
## Dataset Overview
The HealthConnect appointment dataset contains:
* **5,000 appointment records**
* **18 variables**
* **1,696 unique patients**
* Primary identifier: `appointment_id`
The dataset includes information related to patient characteristics, appointment scheduling, previous appointment history, reminders, distance to the clinic, waiting time, and appointment outcomes.
## Data Quality
An initial data quality assessment identified:
* **0 duplicate rows**
* **0 duplicate appointment IDs**
* **90 missing values** in `distance_to_clinic_km`
* **60 missing values** in `waiting_time_minutes`
* **1,366 missing values** in `reminder_channel`
* **0 missing values** in `appointment_outcome`
The missing values in `reminder_channel` appear to be related to appointments where no reminder was sent and therefore require contextual interpretation rather than being automatically treated as errors.
## Key Business Questions
The project focuses on questions such as:
1. What percentage of scheduled appointments are missed by patients?
2. Which factors are most strongly associated with missed appointments?
3. Does the time between booking and the appointme …