Assessment of leech therapy for knee osteoarthritis: a randomized study
Andereya S, Stanzel S, Maus U, Mueller-Rath R, Mumme T, Siebert CH, Stock F, Schneider U (2008) · Acta Orthopaedica · n=113
Study Profile
- Design
- single-center, three-arm randomized study with patient blinding via 'artificial leech' simulated control (Aachen University Hospital)
- Sample size (n)
- 113
- Intervention
- Single application of leeches (group I, n=38) or double application 4 weeks apart (group II, n=35); Hirudo medicinalis periarticularly at the symptomatic knee
- Comparator
- Control group (n=40) treated with an 'artificial leech' simulation device, with treatment blinded as far as feasible
- Primary endpoint
- KOOS and WOMAC scores plus VAS pain, monitored over 26 weeks with intake of pain medication tracked
- Primary result
- All three groups showed improvement in KOOS, WOMAC, and VAS pain after intervention; improvements were statistically significant for treatment groups I and II throughout the 26-week follow-up; greatest improvement seen in the double-treatment group with long-term reduction of joint stiffness and improved activities of daily living
- Follow-up duration
- 26 weeks
- PMID
- 18484250
Key Findings
- First leech-OA RCT to include patient blinding via an 'artificial leech' simulated control arm
- Three-arm design (single application n=38, double application n=35, artificial-leech control n=40) tested both dose-response and blinding
- Repeated leech treatment (group II) produced the greatest long-term improvement in joint stiffness and activities of daily living
- Authors explicitly acknowledged that the positive outcomes could reflect active substances, placebo effect, or expectation effects — methodologically honest about residual uncertainty
- Reduction in pain medication intake statistically significant for both leech treatment arms over 26 weeks
Limitations
- Sham control fidelity uncertain — 'artificial leech' design may not have fully blinded patients to allocation
- Single center (Aachen) — same investigators as 2006 pilot and 2008 hyaluronic-acid trial, raising investigator-site dependence concerns
- Authors explicitly note inability to distinguish biological from expectancy effects
- 26-week follow-up still leaves long-term durability beyond 6 months unanswered
- Three-arm allocation reduced per-arm power versus a two-arm superiority design
Clinical Implications
Andereya 2008 (Acta Orthopaedica) is methodologically the strongest of the early German leech-OA RCTs because it included an 'artificial leech' control intended to address the blinding limitation of Michalsen 2003 and the Aachen 2008 HA comparison. The trial confirms a robust effect of repeated leech therapy on KOOS, WOMAC and VAS pain in moderate-to-advanced knee OA at 26 weeks, while honestly acknowledging that the trial cannot fully separate biological from placebo effects. The result motivated the subsequent Hohmann 2014 plantar-fasciitis sham trial and the ongoing ELECT trial which uses a fully validated sham device. Clinicians should cite Andereya 2008 (Acta Orthopaedica) when discussing the question of whether the leech-OA effect persists under blinded conditions — the answer suggested here is yes, but with the methodological caveat that perfect blinding remains aspirational.
Related Trials
Effectiveness of leech therapy in osteoarthritis of the knee: a randomized, controlled trial
Michalsen A, Klotz S, Lüdtke R, Moebus S, Spahn G, Dobos GJ (2003)
Leech therapy for symptomatic treatment of knee osteoarthritis: results and implications of a pilot study
Andereya S, Stanzel S, Maus U, Mueller-Rath R, Mumme T, Miltner O (2006)
Comparison of modern leech therapy with intra-articular hyaluronic acid injections for symptomatic relief of knee osteoarthritis
Andereya S, Stanzel S, Maus U, Mueller-Rath R, Mumme T, Miltner O, Andereya S (2008)
Effectiveness of home-based cupping massage compared to progressive muscle relaxation in patients with chronic neck pain — a randomized controlled trial (Note: companion knee OA study)
Lauche R, Cramer H, Langhorst J, Dobos G, Michalsen A (2014)