Novel Strategies for Personalized Clinical Decisions in Knee Arthroplasty

Novel Strategies for Personalized Clinical Decisions in Knee Arthroplasty

This study will compare the postoperative recovery of two patient cohorts who attended outpatient physical therapy at two clinics in the Greenville, South Carolina area. The Usual Care cohort received care according to each clinic's pre-existing care guidelines. The CDS Cohort received care informed by a new clinical decision support (CDS) tool.

In this cohort study, the investigators will use regression to examine whether CDS tool exposure was associated with patient outcomes after total knee replacement. Linear mixed effect models will be used for continuous outcomes that have multiple assessment points per patient. For count outcomes (e.g., visit utilization), the investigators will use Poisson/negative binomial regression or linear regression with appropriate outcome transformations.

Inclusion Criteria:

• Patient underwent TKA for primary knee osteoarthritis
• Patient underwent postoperative TKA rehabilitation at one of the two participating physical therapy clinics in the Greenville, SC area

Exclusion Criteria:

• Patients < 40 years old
• Patients whose total episode of care length in outpatient physical therapy was < 21 days
• Patients who attended < 3 outpatient physical therapy visits throughout their episode of care
• Outcomes collected > 6 months after surgery (except for visit utilization, which can include data up to 1 year after surgery)
• Patients who were not exposed to the CDS tool, but were treated by a clinician who was trained to use the CDS tool