Keywords

case management, patient outcomes, predictive modeling, targeted outreach

 

Authors

  1. Hodgman, Stacey B. BS, MSc, RN, CPUM

ABSTRACT

Purpose/objectives: The intent of this article is to explain predictive modeling-a statistical tool-as it applies to the practice of case management. While actuaries and financial experts focus on the statistical relevance of predictive risk scores, case managers will benefit from knowing what these scores mean and how interpreting and applying them into meaningful action can lead to improved patient outcomes.

 

Primary practice setting(s): Predictive modeling can be used by physician practice groups, managed care organizations, worksite wellness programs, and any organization desiring to identify the most actionable population for targeted outreach, education, and management.

 

Findings/conclusions: Predictive modeling is a technological tool that functions as an electronic claims canvasser searching for predefined variables of interest. This tool is used to identify high-cost diagnoses that, in turn, provide a risk score indicative of the likelihood to utilize more healthcare resources and dollars than persons of the same age and gender. By targeting specific diagnoses or conditions, clinicians can define precise patient interventions such as appointment reminders, weight checks, and dietary compliance; assess the intended results of prescribed medications, or simply to provide education and support. The validation of predictive modeling's true value lies in the thorough evaluation of the outcomes of these interventions. The success of predictive modeling can be demonstrated only by a combination of specific data and evidence-based intervention leading to improved models of healthcare delivery and improved patient outcomes.

 

Implications for case management practice: Using predictive models, case managers will be able to target the most actionable patients who will benefit from targeted outreach and education. Case managers will gain an understanding of how a numerical value can lead to the development of a comprehensive collaborative care plan with the patient and other members of the interdisciplinary team to not only improve the patient's overall health status but to engage the patient in his or her own care, empowering the patient to take responsibility for his or her own health status.