Understanding Risk Stratification, Comorbidities, and the Future of Healthcare
Editor’s Note: This article, written in November 2016, contains some content that is no longer current or is not offered by Health Catalyst. To view updated information about the Health Catalyst patient stratification strategy, please visit: https://www.healthcatalyst.com/product/population-builder-stratification/
As value-based care delivery models—like accountable care organizations (ACOs)—enter the healthcare mainstream, managing population health and risk stratification is more important than ever. Healthcare organizations working to change their cost structure and improve outcomes must design interventions that target high-risk, high-cost patients who need to be carefully and proactively managed.
Risk Management in Healthcare Begins with Stratification
The foundational step of targeting these high-risk patients is, of course, to identify them. For example, ACOs have to be able to pinpoint which heart failure patients are at high risk for readmission. Armed with this knowledge, clinicians can schedule follow-up appointments and ensure those patients understand their medications and other aspects of the care plan. Likewise, ACOs must be able to identify patients with a rising risk index that could be triggered by sudden weight gains or hemoglobin A1c values that are trending upward.
The process of separating patient populations into high-risk, low-risk, and the ever-important rising-risk groups is called risk stratification. Having a platform to stratify patients according to risk is key to the success of any population health management initiative.
Overview of Risk Stratification Methods
Several different methods are available for stratifying a population by risk:
Hierarchical Condition Categories (HCCs): Part of the Medicare Advantage Program for CMS, HCC contains 70 condition categories selected from ICD codes and includes expected health expenditures.
Adjusted Clinical Groups (ACG): Developed at Johns Hopkins University, ACG uses both inpatient and outpatient diagnoses to classify each patient into one of 93 ACG categories. It is commonly used to predict hospital utilization.
Elder Risk Assessment (ERA): For adults over 60, ERA uses age, gender, marital status, number of hospital days over the prior two years, and selected comorbid medical illness to assign an index score to each patient.
Chronic Comorbidity Count (CCC): Based on the publicly available information from Agency for Healthcare Research and Quality (AHRQ)’s Clinical Classification Software, CCC is the total sum of selected comorbid conditions grouped into six categories.
Minnesota Tiering (MN): Based on Major Extended Diagnostic Groups (MEDCs), MN Tiering groups patients into one of five tiers: Tier 0 (Low: 0 Conditions), Tier 1 (Basic: 1 to 3), Tier 2 (Intermediate: 4 to 6), Tier 3 (Extended: 7 to 9), and Tier 4 (Complex: 10+ Conditions).
Charlson Comorbidity Measure: The Charlson model predicts the risk of one-year mortality for patients with a range of comorbid illnesses. Based on administrative data, the model uses the presence/absence of 17 comorbidity definitions and assigns patients a score from one to 20, with 20 being the more complex patients with multiple comorbid conditions. It is effective for predicting future poor outcomes. This method is explained in further detail below
One thing all of these models have in common is that they are based, in some degree, on comorbidity. Understanding comorbid conditions is a critical aspect of population health management because comorbidities are known to significantly increase risk and cost. In fact, a study from the Agency for Healthcare Research and Quality reports that care for patients with comorbid chronic conditions costs up to seven times as much as care for those with only one chronic condition.
Applying Comorbidity to Healthcare Analytics
Click to view Risk Model Analyzer
What is the best way to use comorbidity to more accurately predict risk? The truth is that the industry as a whole hasn’t developed standard processes for doing this. Using technology to predict risk while incorporating comorbidities new and evolving. The best analytics vendors provide platforms that allow healthcare systems to tailor risk stratification models to their specific populations.
Calculating a Comorbidity Index
At Health Catalyst, we’ve developed an analytics application called the Risk Model Analyzer. The application indexes every patient for a variety of comorbid conditions. It performs the indexing based on prebuilt comorbidity models taken from peer-reviewed literature. A significant strength of this application is that it is flexible enough to enable clients to either define their own comorbidity models or modify the out-of-the-box models.
The flexibility of the system to adapt to any comorbidity model is key to successful implementation across a wide variety of healthcare organizations. For example: the prebuilt models in the system were developed for comorbidities in adults, but we have multiple children’s hospital clients.. To address these pediatric hospitals’ needs we found a peer-reviewed pediatrics model and plugged it into the architecture—where development, validation, and implementation only took two days.
One model that can be used in the Risk Model Analyzer is the Charlson/Deyo model (a commonly used variation of the Charlson Comorbidity Measure is explained above). The Charlson/Deyo model allows calculation of an index score that summarizes risk based on patient age and the number and types of comorbid conditions a patient