Purpose To determine which patients respond best to viscosupplementation injections for osteoarthritis of the knee.

Methods and Results We undertook a prospective study of all patients undergoing Hylan G-F 20 injections in the knee recording the indication, severity of symptoms, baseline demographic details and the WOMAC score. Outcome data was collected at 3, 6 and 12 months at an independent telephone interview to determine if patients were improved, the same or worse, and by postal WOMAC score.

100 patients were studied. 3 records were excluded, as there was no follow-up recorded. The demographics on 97 were: mean age 67 (range 37–91), male 56%, mean duration of symptoms 8.8yrs and primary OA in 65%.

Overall 43% were improved at 3 months, 31% at 6 months and 29% at 12 months. When the results were analysed according to indication, patients with ‘moderate non-mechanical osteoarthritic symptoms after failed medical management and not severe enough for arthroplasty’ did best (49% at 3 months and 38% at 6 months). Patients with ‘persisting arthritic symptoms after attempted arthroscopic debridement for mechanical type knee symptoms’ had less predictable results (42% improved at 3 months and 23% at 6 months). Patients with ‘severe or deteriorating symptoms while awaiting knee replacement’ or who were ‘too medically ill for TKR’ had a low rate of improvement (18% at 3 months).

Conclusion Viscosupplementation is unreliable in patients with end stage OA awaiting TKR. This study allows for better targeting of this useful expensive treatment modality.