Primary care practices use 4 complementary methods to identify high-risk patients

Risk stratified care management—assigning a patient to a risk category on which care is based—is increasingly viewed as a way to improve care and reduce costs.

An analysis of 484 practices in the Comprehensive Primary Care initiative finds that practices used four primary methods to risk stratify their patient populations: a practice-developed algorithm (215 practices), an American Academy of Family Physicians clinical algorithm (155 practices), payer claims/electronic health record (62 practices), and clinical intuition (52 practices).

Practices that developed their own algorithm identified more patients in the highest two risk tiers (mean=286 patients) than practices that used the AAFP algorithm (mean=181 patients), claims/electronic health record-derived algorithm (mean=171 patients), or clinical intuition (mean=218 patients).

However, practices using a practice-developed algorithm had statistically significant lower numbers of patients receiving care management (69 patients) when compared to clinical intuition (91 patients).

Of note, practices that primarily used clinical intuition provided care management to the highest proportion of high-risk patients. The authors suggest that, as payers shift reimbursement from volume-based to value-driven care, more primary care practices will focus on finding the best ways to implement high-risk care management.