Creating A Novel Fracture Research Animal Models Database

Introduction: Research animal models expedite effective, safe interventions for fracture patients but ethical considerations and failure to meet critical review (e.g., translation) are concerns. Understanding range, selection, and validity of models is crucial but resources challenge users wanting comprehensive information. This study describes development of a novel, fracture-focused database designed to address this need.

Materials and Methods: English language publications (1980–2018) were derived from PubMed using the search-term “bone and fracture and animal.” Clinical case series, reviews, and cadaver studies were excluded. Qualifying papers reporting fracture animal models had the following data transcribed to the new database: author, journal, abstract, summary data (e.g., anaesthetic), animal data (e.g., species), bone, focus (e.g., allograft), and model (e.g., articular fracture). Publications were quantitatively scored (1 star [very poor] – 5 stars [excellent]) for reproducibility, clinical translation and animal welfare.

Results: A total of 4,032 papers were derived from 612 journals from 165 publishers. Number of annual publications progressively increased each year (e.g., 26 [1980]–272 [2018]). Descriptors (low to high) included 15 species (frog [1]–rat [1586]), 24 bones (phalanx [1]–femur [1646]), 134 research foci (bioprinting [4]–fracture healing [3533]), and 37 fracture models (avulsion [4]–diaphyseal [2113]). Percent of total publications scoring 1 or more stars for reproducibility, clinical translation and animal welfare ranged from: 1.0–5.8% (1 star), 5.9–30.6% (2 star), 21.3–42.8% (3 star), 19.2–44.4% (4 stars), and 1.3–26.7% (5 stars).

Discussion/Conclusion: The new database provides a dedicated resource enhancing researchers’ ability to select fracture animal models that are clinically relevant, reproducible and humane.

Acknowledgment: AO Foundation Strategy Fund.