Risk stratification, or separating patient populations based on specific variables, is a central precept in managing value-based contracts like the Medicare Shared Savings Program (MSSP) or Medicare Advantage (MA) plans. The process of risk stratification helps identify patients based on conditions like congestive heart failure (CHF), chronic obstructive pulmonary disease (COPD), or diabetes. Stratifying patients can also be deployed based on predicted resource utilization (cost of healthcare) over a determined period.
For those still familiarizing themselves with population health management and data analytics, comprehensive risk stratification requires different measurements to create a full, accurate picture of a patient’s health status and see if there are any quality deficiencies or unnecessary expenditures. As part of this crash-course on risk stratification, the following models are available to help provider and payer organizations to separate the patients they serve into various classifications of risk and begin care management initiatives.
- Hierarchical Condition Category (HCC) Coding
- The Johns Hopkins ACG® System
- The Elder Risk Assessment (ERA)
- The Charlson Comorbidity Index
- The NYU Algorithm
- Milliman Advanced Risk Adjusters (MARA)
- The Lightbeam ATI Score
Hierarchical Condition Category
Hierarchical Condition Category (HCC) codes use ICD-10 coding to document the health statuses of patient populations. Each chronic condition is assigned a payment value based on the cost of care for a patient with that condition. When providers document HCC codes correctly, payers can calculate their risk-adjustment factor (RAF) score, which is the patient demographics (age, gender, etc.) + sum of HCC codes (factor). The Centers for Medicare & Medicaid Services (CMS) and private payers reimburse MA physicians based on their RAF score, as it reflects the health status of their patients.
The Johns Hopkins ACG® System
The Johns Hopkins Adjusted Clinical Groups (ACG)™ models have been in use for over 30 years. These solutions pull data from electronic health records (EHRs) to create a snapshot of a patient’s future health based on their history, claims, and demographics. Each patient is assigned a score in the ACG®; the average for the full population is 1. A patient with a risk score of 0.5 is predicted to cost half of the average, while a patient with a risk score of 4 is predicted to cost four times the average patient.
The Elder Risk Assessment (ERA)
The Elder Risk Assessment (ERA) was developed by the Mayo Clinic in Minnesota. Taking into account age, gender, marital status, comorbidities, and the number of days the patient spent in the hospital over the previous two years, the ERA calculates a score to indicate a patient’s likelihood of readmission.
The Charlson Comorbidity Index
The Charlson Comorbidity Index was developed in 1987 and assesses patients with comorbid conditions and their mortality-hazard ratio. In short, the index advises healthcare professionals when and where they should focus the majority of their resources. There are 14 conditions measured in the most current Charlson Comorbidity Index, and each condition is weighted with a score from 1 to 6. The person’s probability of living for the next ten years is predicted in the total score, taking their age into consideration.
The NYU Algorithm
The NYU algorithm within the platform assesses the probability of patients utilizing emergency medical services. Due to the high costs associated with ambulances and emergency room visits, this metric provides clinicians with the insight to deploy care before it comes to utilizing emergency departments. It separates patients into four categories: Non-Emergent, Emergent/Primary Care Treatable, Emergent- ED Care Needed – Preventable/Avoidable, and Emergent – ED Care Needed – Not Preventable/Avoidable. With these classifications, providers can act accordingly to prevent readmission.
Milliman Advanced Risk Adjusters (MARA)
The Milliman Advanced Risk Adjusters (MARA)™ score’s objective is to offer the provider a better idea of patient utilization by taking inpatient, outpatient, emergency room, physician, medication, and additional services data and scoring them. The total sum of the scores in these categories equals the MARA score, taking their age into account.
The Lightbeam ATI Score
Lightbeam created the ATI score, or our “Ability to Impact” algorithm. The algorithm is based on insights from discrete clinical data, claims data, and social factors to help determine the probability of positively affecting health outcomes and lowering costs. Patients score between 1 and 10, and the higher the score, the greater the probability of impacting their results with strategic outreach.
See the results of risk stratification methods applied to real beneficiary data by watching the webinar “Population Management: Using Data to Intervene with High-Risk Members,” as part of our Thought Leadership Series.
Jennifer Harkey is an Account Manager at Lightbeam.