Donor site morbidity after autologous bone graft harvesting for fracture nonunion: a systematic review and comparative analysis of four techniques.

Abstract

INTRODUCTION Autologous bone grafting is the reference standard for fracture nonunion; however, donor site morbidity remains a primary concern across all harvesting techniques. SOURCES OF DATA A systematic search of PubMed and Embase was performed from January 2011 to March 2026, chosen to capture the 15-year period since the landmark 2011 review of autograft harvest complications by Dimitriou et al., in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. A total of 49 studies comprising 3029 graft sites across four autograft harvesting techniques were included. AREAS OF AGREEMENT Complication profiles are highly technique-dependent. ICBG demonstrated the lowest overall complication rate (5.83%). Fibular strut grafts demonstrated similarly favourable morbidity (4.88%). Significant differences in overall complication rates were identified between ICBG and reamer-irrigator-aspirator (RIA) (P = .0008) and between ICBG and PK (P < .0001). AREAS OF CONTROVERSY RIA has historically been considered lower morbidity than ICBG; however, the present analysis demonstrated a higher overall complication rate for RIA (9.40%), driven primarily by transfusion requirements (7.88%). This finding likely reflects broader use of RIA in more complex nonunion presentations and improved recognition of transfusion-related risk. GROWING POINTS Periarticular knee harvesting demonstrated the highest overall complication rate (13.64%), challenging its perception as a uniformly low-morbidity technique. Future studies with larger per-subtype cohorts are needed to compare these independently. AREAS FOR DEVELOPING RESEARCH Prospective, multicentre cohort studies employing standardized complication definitions and adequate follow-up durations are needed to overcome the heterogeneity inherent to the current retrospective literature.