Subscribe or Manage Preferences

Using Clinical Data for Risk Stratification: An Effective Tool for Care Management

Rheumatology Practice Management April 2017 Vol 5 No 2 - Risk Assessment
Karna W. Morrow, CPC, RCC, CCS-P, PCS
Sheryl A. Riley, RN, OCN, CMCN

The majority of care managers and social workers already know that prioritizing patients by risk is a critical component of effective care management, but risk can mean different things to different people. Therefore, the tool you use or choose to create is vital to proper risk assessment.

“Risk assessment” is a systematic process of evaluating the potential risks that may be involved in a projected activity or undertaking.

“Risk stratification” is the process of identifying the relative risk of patients in a population by analyzing their medical history. It is a key driver for improving the quality of care delivered by health services.

“Acuity” can be defined as the extent of the care that a patient requires from the nursing staff.1 An acuity-based staffing system adjusts how many nurses are working on a shift based on the patients’ needs rather than the actual patients’ numbers.1 Caregivers and health science literature frequently reference patient acuity, but without specific or consistent definitions or measurements.2 Acuity has been used as a reference for estimating nursing staff allocations and for determining budgets. In our programs, we utilize patient acuity to allocate team resources to our patients via a multidisciplinary team. Understanding the increased level of risk raises the patients’ acuity and the intensity of resources. These are the patients we need to find and risk assess to minimize that risk to the best of our ability. In turn, this improves patients’ quality of life and reduces cost.

Risk Stratification Tool

The first steps in improving patient risk stratification are to create a risk assessment tool that has a goal (ie, what you are trying to measure or achieve). The next step is to make certain the tool examines the patients’ current circumstances, such as physical function or disability, which can help predict patient outcomes. This tool should focus on keeping patients safe, and allowing them to stay in their home, avoiding preventable hospital admissions and re­admissions, mitigating unnecessary emergency department visits, and/or reducing mortality rates.

The idea is to work backwards and to focus on the ultimate outcome, which will assist you in identifying the right patient risk assessment tool. Questions that you may want to ask include:

  • What do we want to achieve for our patients?
  • How do we improve patients’ quality of life?
  • How can we save the patient, provider, and payer money?
  • What elements of the assessment can we gather from the data?
  • What else do we need from speaking with the patient directly?

Risk Assessment Tool

Next, a patient risk assessment tool should assist in predicting patient outcomes; therefore, there should be a clear correlation between the patient’s risk factors and the clinical outcomes. There is still significant disagreement regarding the type of information needed to validate risk. As an organization, you must agree on the type of risk and the clinical outcomes, and then decide how to obtain that information, or you will never achieve success.

The goal should be to stratify all patients in a practice or organization, and then stratify them according to their risk level. This allows the care management team to apply the correct level of goal setting and interventions based on the level of risk that the patient is facing (ie, the right patient at the right time with the right intervention).

Having a core set of patient goals and interventions is vital, but the intensity of how you apply them to each risk level should be different. These goals and interventions can only be created if you know what the overall outcomes and objectives are for the program. Everyone must be on the same page, or the process will not be successful.

Gathering Patient Data

Once you have agreed on the data points, you will need to identify where each data point will be gathered from, such as electronic medical records, hospital records, claims, other systems, or a patient conversation or visit. Our current stratification approach contains 3 different levels of data that are weighted based on the process and the point collected.

The first score is derived from demographics, cost, and diagnosis, with the diagnosis weighted more heavily, because you have not yet spoken with the patient. Although cost and diagnosis provide valuable information, there can also be false-positive results from one event or from an old or controlled diagnosis that happened one time and has not been repeated. It is imperative that an educated coder is involved in validating the diagnosis and coding, because these codes play an important role in determining each patient’s risk level.

The second score is accumulated through a telephone interview, in which a care coordinator speaks with the patient and performs a series of disability screenings to validate the patient’s information related to areas of risk. This validation process involves physical, cognitive, functional, behavioral, social, and environmental factors. This screening contributes to the disability score or the level for the patient and carries great weight when assessing patients’ risk or acuity level.

The third score in the acuity process is completed at the patient’s home, or at a clinic with a face-to-face visit with the patient and/or his or her family. This physical assessment and conversation validates all the previous data that were collected and gives the care team member an opportunity to assess the patient and his or her environment in real time.

An important item to remember regarding risk stratification is that it is organic, and it can change with each “touch point” of the patient based on the information entered into the system, and how the patient’s condition can change at any given time.

Applying the Risk Assessment Tool

A comprehensive risk assessment tool should not be designed to ask patients about every facet of their lives. Extensive risk assessments are resource-intensive and may frustrate staff, as well as patients and their families. Instead, care managers and social workers should perform risk assessment in a conversational format with patients and their families over the phone, as well as during their home assessment, to create a working and trusting relationship, which will result in more valid and actionable information.

The best risk assessment tools are easily verifiable and consist of 5 to 10 questions per tool that cover the physical, functional, cognitive, environmental, behavioral, and social concerns in multiple care settings. Also, we must be able to share this information with our providers, so that they can clearly delineate their patients’ risk level and can create plans of care that reflect that risk level.

Identifying high-risk patients should be easy, but because most risk assessment tools evaluate data instead of gathering information directly from the patient, this creates many gaps in data or missed opportunities.

Proper coding, diagnosis, demographics, and cost are critical components of the process, and will give you a good first look at your patients, but if you want to identify the sickest patients, you need to look beyond the data, and speak with your patients. Patient assessment data that include disability, barriers to care, literacy, and social and environmental tools will better equip you to identify high-risk patients.

In our opinion, high-risk patients are getting lost in the concept of population management; however, by identifying and caring for this subset of patients, practices can control costs and improve the quality of life and outcomes for these high-risk patients.

How to Build a Better Risk Assessment Tool

  • Obtain, assess, and validate all International Classification of Diseases, Tenth Revision codes
  • Clarify the scope of the tool in terms of the type of risk being assessed, the targeted patient populations, and the desired health outcomes
  • Ask the right questions by including clinical, demographic, and psychosocial criteria
  • Involve frontline staff early and often by soliciting staff input and feedback to evaluate the tool’s relevance, reliability, and ease of use.


  1. American Sentinel University-Healthcare. How to use the acuity based staffing model in nursing. The Sentinel Watch: Nursing. February 5, 2014. Accessed December 5, 2016.
  2. Habasevich B. Defining acuity. Mediware Re­habilitation Blog. June 26, 2012. Accessed December 5, 2016.
Last modified: May 5, 2017
  • American Health and Drug Benefits
  • Lynx CME
  • Value Based Care in Rheumatology
  • Oncology Practice Management
  • Urology Practice Management