Reflexion Health

Reflexion Health

Reflexion Health

**Winner of the 2019 Health Value Award for Direct Care Provider – Specialty Musculoskeletal**

Question / Intervention Goal / Objectives 

Reflexion provides a Virtual Exercise Rehabilitation Assistant for patients who have had total knee arthroplasty, as a substitute for in-person physical therapy. The program uses an avatar, motion capture technology, and oversight by physical therapists to deliver physical therapy services.  The research question is whether Reflexion’s program achieves patient outcomes comparable to conventional physical therapy. 


Method/ Calculation 

A randomized controlled trial assigned non-traumatic total knee arthroplasty patients to the intervention or to conventional care.  Measures of patient outcomes included: total healthcare costs for the 12 weeks following the surgery, home health visits, outpatient physical therapy visits, and rehospitalizations.  In addition, the study measured Knee Injury and Osteoarthritis Outcome Score (KOOS) at six and 12 weeks, knee extension, knee flexion, and 10-meter gait speed.  

The service utilization data came from medical claims.  The KOOS is a self-administered survey. The other measures — extension, flexion, gait speed, pain, and falls — were administered by the medical professionals at the six and 12-week point. 


Findings / Metric/ Outcome/ Savings 

The intervention patients had 60% lower billable costs than usual care patients for the 12 weeks after surgery.  The $2,745 savings (mean per patient) stems partly from fewer rehospitalizations on average; fewer home health and outpatient physical therapy visits compared to the usual care group.  

The 6-week knee extension, knee flexion, gait speed and the 6- and 12-week KOOS scores for the intervention and the usual care groups did not significantly differ.  

(Note: patients’ data was assigned to their original assigned group, even if they changed groups during the study period.  An “intent to treat” analysis structure was used.) 



Patients were selected based upon having a Risk Assessment and Prediction Tool score of six or greater prior to surgery, indicating they were expected to be discharged to home after the surgery.  These results may not be generalizable to all types of total knee arthroplasty patients.