An Analysis of Hospital data from 130 US Hospitals from 1998-2008 looking at factors affecting length of hospital stay.

Why This Project?

This analysis represents a study of  130 US hospitals from 1998- 2008 and looked at medical records of patients diagnosed with Diabetes who underwent laboratory, medications, and stayed up to 14 days. The goal is to determine the early readmission of the patient within 30 days of discharge.

Here’s why a reader should read the rest…

By reading on one will discover 1) that race did not significantly effect variability in the procedures done, the amount of time (in days) spent in the hospital correlates with the frequency of procedures,  the number of days spent in the hospital was mostly 5 or less and trails off significantly after the first week through the second week. Thoracic surgery was the most commonly performed of the medical specialties, but cardiology was involved with the most patients.

What and Where of the Dataset

Data set was taken from Kaggle.com and named  diabetic_data.csv.

The data set used was found on Kaggle.com and is titled diabetic_data.csv. There are two tables, Demographics and health in the patient database.

All Analysis

From a histogram, we can see that most hospital stays were 5 days or less and then trailed off significantly during the second week up to 14 days.

Out of nearly 10,000 patients, roughly 47 % were readmitted.

hidden 1st step is Select

Average Number of Procedures by race showed only slight variation

Average procedures per specialty where count > 50

Average time in days in Hospital vs frequency of procedures performed

I used MySQL to analyze this data.

Main Take Aways

In this Project I used MySQL to analyze the data in the patient data set (health and demographic tables) from diabetic_data.csv. The goal was to complete exercises in module 5 of the Data Analyst Accelerator Boot camp.

During the analyses, it was found that:

1) procedure frequency increases with the amount of time spent in hospital or, if you will, decreases with less time spent in the hospital. There were Many procedures performed for those spending an average time of 5.7 days and few procedures performed for those spending only 3.3 days probably because those in hospital longer were sicker.

2) Most patients were in the hospital for 1-5 days and this trailed off till the fewest patients spent 14 days in the hospital. This trend can be seen in the histogram.

3) Race does not appear to be a significant factor in the number of procedures performed. However, for what it is worth, most procedures were performed on African Americans, just not by much.

4) About half the patients, 47% were readmitted, the remaining 53% were not readmitted.

5) Surgery Thoracic was the most common procedure performed (avg 3.5) while most patients ( 5352) were sent to cardiology.

Call to Action

Diabetes is a serious disease for many, and it cuts across racial lines (race is not a determining factor). Many (47 percent) will be or were readmitted. Given that most spend less than a week in the hospital, it might be a good idea to check and see if patients were released to early. This study did not appear to look at diet and amount of exercise patients performed and it would be useful to look into this, especially diet, to see if this might have bearing on procedures that needed to be performed and who got readmitted and who did not.