Dextrose Prolotherapy for Symptomatic Knee Osteoarthritis: Feasibility, Acceptability, and Patient-Oriented Outcomes in a Pilot-Level Quality Improvement Project

Knee osteoarthritis (KOA) is a growing health problem with limited nonsurgical treatment options. Prolotherapy is an injection-based technique for chronic KOA pain; health plan coverage is limited, presenting an access barrier. A local health plan recently included coverage for prolotherapy for KOA, but uptake and treatment response in routine care are unknown. The authors conducted a pilot-level quality improvement (QI) project to explore the feasibility, acceptability, and effects of prolotherapy for painful KOA in a primary care setting. Design: QI prospective case series. Setting/Location: Outpatient: invitation letters were sent to symptomatic KOA patients with a primary care provider whose health plan covered prolotherapy. Subjects: Primary care patients with KOA. Intervention: Intra- and extra-articular prolotherapy injections: patients received up to six prolotherapy sessions. Outcome measures: Primary: Feasibility: response rate to invitation to utilize prolotherapy. Acceptability: patient adherence to, and satisfaction with, three or more prolotherapy sessions. Secondary: Survey based (the Western Ontario McMaster University Osteoarthritis Index, WOMAC, 0-100; EuroQOL 5-D). Objectively assessed: function (30-sec chair stand, 4 × 10 m walk, 9-step stair-climb), overall activity (ActiGraph wGT3X accelerometer), treadmill gait analysis, and preferred walking speed. Outcome analysis (paired t-test) was per protocol, comparing follow-up and baseline outcome data at ∼8 months. Results: Thirty-nine patients were invited, 11 responded, and 7 patients (59.6 ± 9.3 years, 6 female) received 5.0 ± 1.1 prolotherapy sessions. Satisfaction was high. Their WOMAC scores improved by 27.6 ± 19.5 points (p = 0.02) at 8.4 months. Functional testing improved by 8.0 ± 3.6 sec (p = 0.003) in the 4 × 10 m walk. There were no group differences between baseline and follow-up in chair stand, stair-climb, accelerometry, or gait outcomes. Five patients increased their preferred walking speed (p = 0.001). Conclusions: These data suggest that prolotherapy in this primary care clinic is feasible and acceptable. Self-reported improvement is similar to that of efficacy studies; office-based, objectively assessed functional assessment can be performed. Further evaluation is warranted.