Evidence Assessment
GRADE methodology and the ASH evidence classification system
Last updated: November 1, 2026
Rigorous evidence assessment is essential for credible medical communication. ASH applies the internationally recognized GRADE framework alongside a purpose-built 3-tier classification system to ensure that every clinical claim on this website is transparently linked to the quality of its supporting data. This page explains how to read, interpret, and critically evaluate the evidence presented throughout our resources.
GRADE System — Certainty of Evidence
The Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) system categorizes evidence certainty into four levels. RCTs begin as “high” and may be downgraded; observational studies begin as “low” and may be upgraded.
High Certainty
Further research is very unlikely to change confidence in the effect estimate. Typically requires multiple large RCTs with consistent results.
Moderate Certainty
Further research is likely to have an important impact on confidence and may change the estimate. One or more RCTs with some limitations.
Low Certainty
Further research is very likely to change confidence. Well-conducted observational studies or RCTs with significant methodological limitations.
Very Low Certainty
Any estimate of effect is very uncertain. Case series, case reports, expert opinion, or studies with critical flaws.
Study Design Hierarchy
Not all study designs carry equal weight. The hierarchy below reflects the inherent susceptibility of each design to bias:
| Level | Study Design | Strengths | Limitations |
|---|---|---|---|
| I | Systematic review / Meta-analysis | Synthesizes all available evidence | Dependent on quality of included studies |
| II | Randomized controlled trial (RCT) | Minimizes selection bias, confounding | Expensive, blinding difficult for leeches |
| III | Controlled cohort / Comparative study | Real-world data, larger samples possible | Selection bias, confounders |
| IV | Case series / Case-control | Documents clinical experience | No comparator, high bias risk |
| V | Case report / Expert opinion | Generates hypotheses | Anecdotal, not generalizable |
ASH 3-Tier Evidence Classification
Classification System
Tier A — FDA-Cleared
Applications included in the FDA 510(k) clearance for medicinal leeches. The 2004 clearance (K040187) rested on the device's historical use as a pre-amendment device, with no predicate device. Current Tier A indications: microsurgery salvage (pedicle flaps, digit replantation) and venous congestion.
Tier B — Evidence-Based
Published RCTs demonstrating clinical benefit, but not included in the FDA clearance. Key Tier B indications: osteoarthritis (Michalsen 2003, Andereya 2008), lateral epicondylitis (Backer 2011).
Tier C — Investigational
International clinical experience — case series, cohort studies, or expert consensus without RCT support. Covers cardiology, neurology, and most specialties. Presented for educational purposes with explicit disclaimers.
Key Randomized Controlled Trials in Hirudotherapy
| Study | n | Indication | Comparator | Primary Outcome |
|---|---|---|---|---|
| Michalsen et al. (2003) | 51 | Knee OA | Topical diclofenac | WOMAC pain — significantly superior at day 7 and 28 |
| Andereya et al. (2008) | 113 | Knee OA | Sham (artificial leech) | KOOS/WOMAC/VAS improved in all groups; significant in the leech-treated groups |
| Backer et al. (2011) | 68 | Epicondylitis | Topical diclofenac | VAS pain — significantly greater decrease with leeches at day 7 (P=0.0075); group difference no longer significant at day 45 (P=0.110) |
| Michalsen et al. (2008) | 32 | Thumb base OA | Topical diclofenac | Pain and grip strength — significant benefit |
What Makes Good Clinical Evidence
Design Features
- Randomization — eliminates selection bias
- Blinding — minimizes expectation effects (challenging with leeches; sham designs address this)
- Adequate sample size — sufficient statistical power (a priori calculation)
- Validated outcome measures — WOMAC, VAS, SF-36, tissue salvage rate
- Intention-to-treat analysis — preserves randomization
Reporting Standards
- CONSORT compliance for RCTs
- Pre-registration (ClinicalTrials.gov or equivalent)
- Conflict of interest disclosure
- Peer-reviewed publication
- Long-term follow-up (≥6 months preferred)
Current Evidence Gaps
Despite promising results, hirudotherapy evidence has critical gaps. No multi-center RCTs have been conducted. Existing trials are relatively small (n=51–113). Follow-up is mostly 2-3 months (Michalsen 2003: day 91; Stange 2012: day 63), with the longest 26 weeks (Andereya 2008) and 180 days (Isik 2017). There is no head-to-head comparison with gold-standard treatments (e.g., intra-articular corticosteroids for OA). FDA-recognized endpoints for expanded indications have not been established. ASH advocates for larger, longer, and more rigorously designed trials to advance the field from Tier B toward potential regulatory recognition.
ASH transparency: how grades are assigned
For the full editorial workflow behind every Tier A / B / C assignment — including PMID verification status, reviewer attribution, and the dispute-resolution process — see /how-evidence-graded. Current import-baseline transparency (9,786 PubMed/DOI records in the research library, against a 10,000-article target) is tracked at /research-roadmap and /pmid-audit-status. For clinician-facing decision support keyed to these tiers, see /clinical-decision-support.