Session I - Tibia Fractures

Initial Evaluation of The Grade IIIB/IIIC Tibia Fracture: A Prospective Limb Salvage Score Assessment

Michael J. Bosse, MD, Ellen MacKenzie, PhD, and the LEAP Study Group

Carolinas Medical Center, Charlotte, North Carolina, USA

Objective: High-energy lower extremity trauma presents the surgeon with a limb salvage versus amputation treatment dilemma in many cases. Multiple assessment scales have been developed to assist the surgeon in the limb salvage decision-making process. None of these, however, has been evaluated prospectively using a large cohort of patients. The purposes of this study are: (1) to prospectively evaluate the sensitivity, specificity and positive predictive value of four salvage indices in the assessment of Gustilo Grade IIIB and IIIC tibia fractures, (2) to empirically assess the relative contribution of the index components in predicting limb salvage versus amputation, (3) to determine the clinical issues judged by the treating surgeons to be most influential in their decision to salvage or amputate an extremity, and (4) to determine who is participating in the limb-salvage decision-making process.

Methods: A multicenter prospective study was designed to evaluate treatment and outcomes following high-energy lower extremity trauma. All patients were examined on admission, and the injury was characterized according to the Gustilo, Tscherne and the AO classification of soft tissue injuries. All fractures were classified using the OTA fracture compendium. Each patient had the components of the Mangled Extremity Severity Score (MESS), the Predictive Salvage Index (PSI), Limb Salvage Index (LSI), and the NISSSA recorded at the time of initial evaluation. Surgeons were also asked at the time of initial evaluation to rank order multiple injury and patient characteristics as to their importance in making the decision to salvage or amputate. Lastly, the surgeon recorded the degree of participation that the patient, the family and other members of the treating team had in the decision-making process.

Results: A total of 520 patients were enrolled in the Limb Salvage Study. Of these, 248 limbs in 240 patients had Grade IIIB or IIIC tibia fractures. Patients with traumatic amputations (60 legs) were excluded from review; 74 patients underwent amputation during initial hospitalization (66 prior to attempted soft tissue closure and 8 after). The remaining 174 patients proceeded along the reconstruction pathway and were discharged from the hospital without limb amputation. Threshold groupings (above or below) and sensitivity, specificity and positive predictive values of salvage indices are presented in the table.

Score n = 248	Patients >= threshold score	% amputated >= threshold score	% amputated < threshold score	Sensi/Specif for early amputation	Positive Predictive Value
MESS	20.2% (50)	66.0% (33)	20.7% (41)	44.6/90.2%	66.0% (33/50)
PSI	27.0% (67)	58.2% (37)	20.4% (37)	37.8/82.2%	55.2% (37/67)
LSI	19.4% (48)	85.4% (41)	16.5% (33)	55.4/96.0%	85.4% (41/48)
NISSSA	12.5% (31)	90.3% (28)	21.2%(46)	37.8/98.3%	90.3% (28/31)

Multiple logistic regression analysis of the key components of all of the limb salvage indices for this patient group identified the degree of muscle injury, followed by plantar sensation and the fracture pattern to be significant factors in the salvage vs. amputation decision process (P<.01). Analysis of the treating surgeon responses identified the presence of plantar sensation and blood supply as the major determinants in the decision to salvage an extremity. In 30.5% of the cases, the surgeon identified "unsure at present time of outcome" as an important reason to initially select limb salvage. When evaluating the limb and deciding on amputation as the primary procedure, the surgeons identified the severity of muscle damage, vascular injury and absence of plantar sensation as the most critical factors in their assessment.

The decision-making process was heavily influenced by the orthopaedic opinion. In cases selected for limb salvage, the orthopaedic surgeon was listed as "very involved" in the decision process in 98.9% of the cases (general/trauma surgeon = 25.7%, plastic surgeon = 14%, the patient = 11.7%, and the family = 4.7%). In limbs undergoing amputation, the orthopaedic surgeon was "very involved" in 96% of the cases (general/trauma surgeon = 43.8%, plastic surgeon = 14%, the patient = 21.6%, and the family 21.6%).

Discussion and Conclusions: This is the first prospective analysis of the published limb salvage indices using a large cohort of patients who were subjected to a prospective data collection process. In this series, we found a low sensitivity and a relatively high specificity across all of the indices. The percent of patients amputated above the various index threshold scores varied from 58% to 90%. Conversely, between 16% and 22% of patients with scores below this threshold also underwent amputation. At this point in time, our data do not support the use of any of the indices to make an acute limb salvage or amputation decision.

The orthopaedic surgeon appears to have the most decisive role in the acute limb salvage decision making process. Factors identified by the orthopaedic surgeon to heavily influence the decision to amputate, in order of preference, were the severity of muscle damage, the presence of a vascular injury and the absence of plantar sensation. Two of these factors (the degree of muscle injury and plantar sensation) were the same factors identified by the regression analysis as significant predictors of limb amputation. Although our analysis predicts the significance of the severity of muscle damage in the limb salvage decision process, neither the MESS nor the NISSSA score this domain.

Finally, it must be remembered that the ability to predict limb survival and discharge from the hospital does not equate to a long-term good functional outcome. Correlation of late outcome with the initial limb salvage scores must be determined in order to refine the validity assessment of a scoring system.