Patient Hotspotting
Your name: Lee Arthur
Your name: Diana King
Project Overview:
The purpose of this project is to identify high-cost/high-need patients, and risk stratify them
based on patient data to understand how to best develop a tailored chronic care management plan
with the goal of improving quality of care and population health while decreasing costs. We used
aggregated patient claims data on 14,215 randomly selected Monroe Health Plan Medicaid patients
who had an encounter in 2016, and used regression analysis tools such as pivot tables to compare
them across health variables. Furthermore, we used this data to compare high cost patients (top 10%
most expensive patients with an encounter in 2016) relative to all others (other 90%) on 8 metrics (age
group frequency, gender frequency, race frequency, ethnicity frequency, top 5 most prevalent primary
risk conditions, ED utilizations, inpatient utilizations, and total costs), allowing us to best tailor our
chronic care management plans.
Risk Stratification and Comparison:
Comparing high cost patients (top 10%) relative to all others (other 90%) on 8 metrics:
1. Age group frequency (%): High cost patients are more likely to be ages 20-54, while low cost
patients are more likely to be ages 0-19.
2. Gender frequency (%): The highest cost and lowest cost patients are more likely to be women.
3. Race group frequency (%): High cost and low cost patients are more likely to be White.
1
Your name: Lee Arthur
Your name: Diana King
Project Overview:
The purpose of this project is to identify high-cost/high-need patients, and risk stratify them
based on patient data to understand how to best develop a tailored chronic care management plan
with the goal of improving quality of care and population health while decreasing costs. We used
aggregated patient claims data on 14,215 randomly selected Monroe Health Plan Medicaid patients
who had an encounter in 2016, and used regression analysis tools such as pivot tables to compare
them across health variables. Furthermore, we used this data to compare high cost patients (top 10%
most expensive patients with an encounter in 2016) relative to all others (other 90%) on 8 metrics (age
group frequency, gender frequency, race frequency, ethnicity frequency, top 5 most prevalent primary
risk conditions, ED utilizations, inpatient utilizations, and total costs), allowing us to best tailor our
chronic care management plans.
Risk Stratification and Comparison:
Comparing high cost patients (top 10%) relative to all others (other 90%) on 8 metrics:
1. Age group frequency (%): High cost patients are more likely to be ages 20-54, while low cost
patients are more likely to be ages 0-19.
2. Gender frequency (%): The highest cost and lowest cost patients are more likely to be women.
3. Race group frequency (%): High cost and low cost patients are more likely to be White.
1
