Udacity-data-visualization

This is my final project on Udacity Data Analyst Nanodegree.

Domain

This analysis is focused on Healthcare Analytics.

Objectives

Objectives

1. Practice the core principles of data analysis

2. Gather information from the data using univariate, bivariate, and multivariate analysis

3. Use question, code, and observation methods.

4. Generate 15 insights from the dataset.

Dataset Link

The dataset link can be found on kaggle - https://www.kaggle.com/datasets/anmolkumar/janatahack-healthcare-analytics-part-2

Data Dictionary

Column	Description
case_id	Case_ID registered in Hospital
Hospital_code	Unique code for the Hospital
Hospital_type_code	Unique code for the type of Hospital
City_Code_Hospital	Code of the Hospital
Hospital_region_code	Region Code of the Hospital
Available Extra Rooms in Hospital	Number of Extra rooms available in the Hospital
Department	Department overlooking the case
Ward_Type	Code for the Ward type
Ward_Facility_Code	Code for the Ward Facility
Bed Grade	Condition of Bed in the Ward
patientid	Unique Patient Id
City_Code_Patient	City Code for the patient
Type of Admission	Admission Type registered by the Hospital
Severity of Illness	Severity of the illness recorded at the time of admission
Visitors with Patient	Number of Visitors with the patient
Age	Age of the patient
Admission_Deposit	Deposit at the Admission Time
Stay	Stay Days by the patient

Problem Statement

The recent Covid-19 Pandemic has raised alarms over one of the most overlooked area to focus: Healthcare Management. While healthcare management has various use cases for using data science, patient length of stay is one critical parameter to observe and predict if one wants to improve the efficiency of the healthcare management in a hospital.

This parameter helps hospitals to identify patients of high LOS risk (patients who will stay longer) at the time of admission. Once identified, patients with high LOS risk can have their treatment plan optimized to minimize LOS and lower the chance of staff/visitor infection. Also, prior knowledge of LOS can aid in logistics such as room and bed allocation planning.

Suppose you have been hired as Data Scientist of HealthMan – a not for profit organization dedicated to manage the functioning of Hospitals in a professional and optimal manner. The task is to accurately predict the Length of Stay for each patient on case by case basis so that the Hospitals can use this information for optimal resource allocation and better functioning. The length of stay is divided into 11 different classes ranging from 0-10 days to more than 100 days.

Dataset summary

1. The dataset has a total rows of 318438 and total columns of 18

2. It consists of 9 numeric and 9 categorical features.

3. There are 2 columns with missing values but they are less than 5% of the entire data set rows.

4. The data is Tidy because we did not have any structural issues.

Data Wrangling process

1. The missing rows are drop from the dataset

2. The categorical variables that are ordinal in nature are converted using pandas Categorical Dtype

Plot used

In the exploratory data analysis process, Bar-chart, Histogram, Violin plot, Heatmap, Box plot, Line plot, Grid plot are used.

Questions asked from the dataset

Univariate Analysis

1. For How long did Patient Stay?
2. Which Ward Type admit patient the most and least in time of their stays in the hospital?
3. Which Department has the most admitted patient?
4. Which Severity of Ilness brought most patient to the Hospital?
5. Top 5 most visting patient code admitted.
6. Which type of hospital is visted the most?
7. Age group frequency.
8. Admission Deposit Distribution

Bivariate Analysis

9. Which type of Severity of Ilness as the most Stay
10. Average Admission deposit per Severity of Ilness in the Hospital.
11. Department distribution in Age and Hospital_region_code
12. Type of admission per Stay

Multi-Variate Analysis

13. Age distribution per Admission Deposit and Severity of Illness.
14. Stay vs Ward type vs Admission Deposit.
15. Correlation analysis of the numeric data.

Findings to Above Questions

Question 1

• 21-30 days and 11-20 days has the highest days of Patient Stay.

• 61-70 days and 91-100 has the least days of Patient Stay.

Question 2

• Most patient are admitted to R ward type
• Least patient are admitted to U ward type

Question 3

• Most pateint are admitted to Gynecology department.

• Radiotherapy and anesthesia has relative the same admission into the department.

• Surgery department as the least admission.

Question 4

Most the patient visit for Moderate and Minor Severity of Ilness.

Question 5

• Patient 66714 visted the hospital 50 times, which is the patient with highest number of occurence.

Question 6

• Hospital code 26 is visted more than 30,000 times in the dataset.

Question 7

• Most Age group falls between 31-40 and 41-50 Age Group

• The Age group distribution looks Uni-model

Question 8

• Most patient admission deposit falls between 4000 - 5000.

• The admission deposit distribution is right skewed.

Question 9

• 21-30 days stays has 16% of Moderate Ilness and 4.9% of Extreme Ilness

• 11-20 days has 8.5% of Minor Ilness.

Question 10

• On Average the admission deposit per severity of ilness are the same. They are around about 4500.

Question 11

• The Gynecology department dominate both Age and Hospital region in the dataset.

Question 12

• At 0-10 days of stay most patient are admitted into Emergency.

• Other days of stay most patient are admitted into Trauma.

Question 13

• The Minor Severity of Ilness has the largest proportion with respect to age and Admission deposit.

Question 14

• Most patient did not require admission deposit at U ward type. Only stay 0-10, 11-20, 21-30 and 31-40 use the ward U, while 11-20 and 31-40 have better distribution.

Question 15

• City code Hospital and Hospital code has 13% correlation.

• City code Hospital and Hospital code are positive weakly correlated.

• Admission deposit and Available Extra rooms in Hospital are negative weakly correlated.

Conclusion

1. It is observed that most people visited the hospital for sexual reproduction reasons because gynecology is the most visted department with about 79%.
2. Patient are most likely to stay 11 - 40 days maximum
3. The Age shows most patient in the hospital are Adults.