Ganga hospital open injury severity score as a predictor of early failure of limb salvage among gustilo type III A and B tibia fractures in Uganda: a prospective cohort study.

Abstract

Background: Despite advances in antibiotic therapy and microsurgery, the management of Gustilo and Anderson type IIIA and B open tibia fractures continues to pose a considerable challenge in developing countries. This has evolved from historical immediate amputation to modern approaches that prioritize both aesthetics and functional outcomes. Nonetheless, a consensus on limb salvage versus amputation remains elusive, prompting the development of prognostic limb scoring systems. Our study assessed the predictive accuracy of the Ganga Hospital Open Injury Severity Score (GHOISS) for early failure of limb salvage in Gustilo type IIIA and B tibia injuries.

Methods: This was a prospective study that examined open tibia fractures at two tertiary hospitals in the emergency and orthopaedic units between June and October 2023. Fifty-three (26 IIIB and 27 IIIA) satisfied the study inclusion and exclusion criteria. Four injuries (type IIIA) in paediatric patients that had met the inclusion criteria were excluded from analysis to ensure homogeneity and generalizability of the results due to their small numbers. The severity of injury for each Gustilo type III A and B tibia fracture was determined using the GHOISS, and limb salvage decisions were made irrespective of the GHOISS. Follow-up was extended for up to fourteen days to assess the necessity of secondary amputation in salvaged limbs.

Results: Among 49 Gustilo type IIIA and B tibia fractures, 43 were successfully salvaged, while 6 necessitated amputation (4 primary, 2 secondary). A GHOISS of 13 demonstrated maximum specificity (90.7%) and sensitivity (83.3%) in predicting amputation, with an AUC of 0.923 (95% CI 0.804-0.977), indicating strong discriminatory accuracy.

Conclusion: The GHOISS reliably predicted outcomes in patients with Gustilo type IIIA and B tibia fractures, with a score of 13 demonstrating optimal sensitivity and specificity above which early failure of limb salvage is anticipated.