Predicting Patient Re-Admission within 30 days

 In

PREDICTING PATIENT RE-ADMISSION
WITHIN 30 DAYS

About the Client

The client is a a leading provider of comprehensive health informatics solutions for more than 900+ hospitals in the US.

Business need

The customer wanted ALTEN Calsoft Labs to identify a patient’s chance of getting re-admitted upon discharge within 30 days.

Solution delivered

ALTEN Calsoft Labs’ Big Data Analytics team did the following to come up with a predictive analytics solution.

  • Identify the Predictors
    • Demographics – Age, Sex
    • Lab data – Includes lab tests
    • Vitals – Includes BP, Sugar, Weight
    • Visit types – Emergency, In-patient, outpatient
    • Diagnosis & Diseases/ ailments – Heart, Pneumonia
    • Previous hospital visit
    • Length of stay
  • Process the Dataset
    • Date fields – Validate the date
    • Calculate the 30 days stay
    • Refine the dataset for building Model/Analysis
  • Identify the Suitable Algorithm
    • Gradient Boosting
    • Random Forest
    • Support Vector Machines
    • Logistic Regression
    • K-Nearest Neighbour
    • Ridge
  • Data for new evaluation
    • New data sets are analysed
    • Available in .CSV
  • Data Visualization
    • .csv is converted into JSON format
    • Displayed using Alten Calsoft UI framework
  • Build/Evaluate the Model
    • Model is maintained dynamically
    • Split the data into
      • Training set
      • Test set
    • Build the model on training set
      • Model is serialized
      • Ready for new datasets

Business benefits

Clinicians can be prepared to provide better post-discharge care for patients who are likely to get re-admitted.

Technology stack

  • Business Platform: Hadoop – Hortonworks
  • Paradigm: Hive, MapReduce
  • Machine Learning Tools: Scikit-Learn,R
  • Language: Python, R

Let’s fast-track your next big idea

Recent Posts
Contact Us

If you’d like us to contact you, please fill out the form.

Not readable? Change text. captcha txt