gms | German Medical Science

Objectives/Interrogation: The mangled extremity severity score (MESS) is utilized to decide whether to try limb salvage or perform an amputation for the severely mangled extremity, but its utility in the treatment of the mangled upper extremity is not documented. To demonstrate its accuracy, we conducted a meta-analysis of studies comparing the diagnostic accuracy of the MESS in patients with mangled upper extremities.

Methods: We retrieved PubMed and EMBASE. We defined sensitivity as the probability that limbs requiring amputation had the MESS of 7 and above. Specificity was defined as the probability that salvaged limbs had the MESS <7. Cases where reconstruction failed and eventual revision amputation was performed were counted as amputation. The heterogeneity and publication bias were evaluated. Pooled sensitivity and specificity and positive and negative likelihood ratios were calculated using a bivariate random effects model for meta-analysis of diagnostic tests.

Results and Conclusions: Most studies dealt with injuries of lower and upper extremities collectively. Therefore, after full-text reading, we verified the availability of the information associated only with upper extremity injuries from the studies. Ultimately, nine studies (N=285) were included in the meta-analysis. The I-square test showed heterogeneity (I=0.60), caused by the difference in the injury characteristics between studies. Allowing for potential heterogeneity among studies, a hierarchical summary receiver operating curves model and bivariate model were used.

The pooled sensitivity and specificity were 97.7% (95%CI; 72.5% to 99.9%) and 89.2% (95%C; 75.6% to 95.7%), respectively. The positive and negative likelihood ratios were 9.04 (95%CI; 3.81 to 21.5) and 0.25 (95%CI, 0.00 to 0.38), respectively. The area under the hierarchical summary receiver operating curve was 0.98 (95%CI; 0.97 to 0.99). Deeks' funnel plot asymmetry test suggested a little potential publication bias (p=0.09).

The pooled sensitivity and specificity was not worse than those of the lower extremity; the sensitivity means an extremely high probability of achieving limb salvage in cases where the MESS was under the threshold. The gained specificity suggests that limb salvage was achieved in more than 10% of the patients whose MESS was 7 and above. Limb salvageability of the mangled upper extremity does not depend on the MESS; amputation cannot be justified by the MESS. Development of a new scoring system that has excellent specificity is indispensable for surgeons.