Recently Published
STAT W4702 - UNI-MM3557 - Assignment 10
STAT W4702 - UNI-MM3557 - Assignment 10
Columbia University - Statistical Inference and Modeling - W4702 - Chapter 4 ISLR
Assignment 9
Columbia University - Statistical Inference and Modeling - W4702 - Chapter 4 ISLR
Columbia University - Statistical Inference and Modeling - W4702 - Chapter 3 ISLR
Columbia University - Statistical Inference and Modeling - W4702 - Chapter 3 ISLR
Assignment 8
Columbia University - Statistical Inference and Modeling - W4702 - Chapter 2 ISLR
Columbia University - Statistical Inference and Modeling - W4702 - Chapter 2 ISLR
Assignment 7 - Rcode and output
Physician Affinity Graph
Network Analysis of Healthcare Providers based on Medicare Claims data
What do we hope to answer?
In our work, we hope to create the entity relationship graph based on comman Medicare claim counts (referrals) with additional considerations to shared HCP specialization, shared drug prescriptions, propensity to prescribe propritory versus generics and geographic area. Specifically, our aim is to understand the strength of the relationship between HCPs that can, in turn, help answer the following questions:
Who are the most influential doctors for a particular disease area in a specific geography?
Can we apply ranking or an influence score to the healthcare provider?
Is there a way, using these data, to develop a 'Key Opinion Leader' KOL recommendation engine for the health care provider community?
We see payers, pharmaceutical companies, patients, researchers and academia as potential stakeholders interested in answers to the above questions.
The insights will be aggregated at the following levels:
Providers (NPI, name, city, state) view (by Geo and Specialty)
Geography (city and state) view (by HCP and Specialty)
Specialties view (by Geo and HCP)