Primary Care Risk Scoring and Stratification – A Guide for Getting Started

By Joy Key, Director of Provider Services, Emtiro Health

2019 Alliance sponsor feature article courtesy of Emtiro Health

One of the goals set forth by Medicaid Managed Care in North Carolina is to build upon previous successes by “innovating and evolving to improve the health of North Carolinians.”1 Stated another way, North Carolina aims to address Population Health at a macro level through innovation at the provider level: using robust data and pioneering service delivery to improve care for Medicaid beneficiaries. More specifically, Tier 3 Advanced Medical Home practices will be responsible for risk scoring and risk stratifying all of their Medicaid patients to achieve these goals and to match patient needs to the optimal level of care management at the right time. Resources can be best allocated using a clear and consistent method of patient stratification.

Accurate risk scoring and risk stratification of patients are two foundational principles of successful population health strategies. At the individual level, accurate risk scoring is the first step in developing a person-centered care plan. At the population level, risk stratification allows providers to develop care models that address the needs of distinct patient subsets with similar complexity and care needs, improving health outcomes for both the individuals involved as well as the overall population.

As our health care system moves away from fee-for-service payment and toward value-based reimbursement, providers and payers will have to fully understand the risks of their patient population in order to optimize care delivery. A risk score is simply a metric that attempts to predict aspects of a patient’s care (i.e. cost, Emergency Department utilization, inpatient readmission, etc.) based upon his/her clinical history and compared to a larger, average population. Sounds simple enough. However, risk scoring and patient stratification are data-driven, iterative processes that should involve providers, data analysts, and care management professionals to be the most successful. Methodologies need near constant evaluation to improve accuracy and increase value to the overall Care Management process and Population Health strategies.

A simple search online will reveal a large number of risk scoring and risk stratification methodologies – some very complex and expensive. While appropriate in some systems or populations, bigger isn’t always necessarily better and one size doesn’t fit all. The first step in choosing or designing risk models is a thorough understanding of the population being served.

Understanding the population will help determine how many risk “levels” are appropriate and what factors to include. In the earliest days of risk scoring, health systems used age and gender as simple factors. As large amounts of clinical data became more accessible, chronic disease measures and utilization rates were added to the risk formulas.

Simple scoring could be “Level 1, Level 2, Level 3”; “Tier A, Tier B, Tier C”; etc. A simple methodology might include condition counts, prescriptions, ED utilization, social determinants of health screening, etc. Again, the key to developing an effective risk scoring and risk stratification system is first understanding the patient population.

The second step in designing appropriate models is access to accurate, reliable patient data. Providers and their staff must use high quality data sources to ensure data integrity. Also, understanding subtle differences in types of data is essential – mortality versus morbidity, claims versus diagnostic data, etc. Risk scoring is only as good as the underlying data. A simple cautionary example is diagnostic data. Ample clinical and diagnostic data elements are likely already readily available from a practice’s own electronic health record (EHR), but any variation among providers in clinical documentation and HCC coding from standardized practices could skew results. Missing or inaccurate clinical data will cause the risk scoring and risk stratification to also be inaccurate and inconsistent.

Other considerations for effective Tier 3 advanced Medical Home risk scoring and risk stratification include:

  1. Conduct a thorough and realistic assessment of internal resources and capabilities. Developing and implementing risk scoring and risk stratification methodologies will require significant staff time as well as significant expertise in analytics and statistics. If your practice plans to rely upon risk scoring modules within the EHR, Quality, IT, and Care Management staff members must still thoroughly understand the algorithms and understand how to modify them to address your practice’s unique patient population. Additionally, each iteration of the risk models will require staff retraining.
  2. Conduct an inventory of the reporting capabilities of your EHR. All of the data you need to risk score and risk stratify attributed patients may be available from payers or it may lie within the clinical documentation system. If staff cannot retrieve actionable data, the risk scoring and risk stratification models will become stagnant and ineffective. Custom reporting may be available, but it should be determined if that strategy is sustainable for the long run.
  3. A Tier 3 practice’s risk scoring and risk stratification methodology must be able to incorporate risk scores from each contracted Prepaid Health Plan. Combined with a practice’s own clinical perspective, the risk stratification system yields a broader view of patients’ needs and health status.
  4. Whatever risk scoring and risk stratification methodology is selected, it must be used consistently by the entire care team. Everyone must be onboard with the process and understand how to use the information.
  5. Evaluate the published aims and objectives of North Carolina’s Medicaid Managed Care Quality Strategy in order to have a complete understanding of what is to be achieved and which enrollees fall into “priority populations.”2

Ultimately, risk scoring and stratification improve outcomes for both providers and patients, which is why it has become a requirement for Tier 3 Advanced Medical Homes. While risk scoring and stratification are complex undertakings, with thoughtful examination of available data, resource capabilities and care delivery goals your organization can succeed in developing a workable system that best suits your patients’ needs and your overall vision for improved population health.

1North Carolina’s Quality Strategy for Medicaid Managed Care, pg 5, March 20, 2018.
2The Department defines “priority populations” as: enrollees with Long Term Services and Supports (LTSS) needs; Adults and children with Special Health Care Needs; Individuals identified by the Prepaid Health Plan (PHP) as at Rising Risk; Individuals with high unmet resource needs; At-risk Children (0-5); High-Risk Pregnant Women; and other priority populations as determined by the PHP.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: