Variation in Implant Selection for Ankle Fractures: Identifying Cost Drivers

BACKGROUND: Implant selection is the first opportunity for surgeons to control costs of fracture fixation. The current literature has demonstrated surgeons' poor understanding of implant costs. Our study evaluated implant cost variability for surgically treated ankle fractures and distal tibia fractures. Our hypothesis was that significant cost variation exists among providers. The goal was to identify cost drivers and determine whether specialty training is linked to implant selection.

METHODS: A retrospective 2010-2017 chart review was performed for 1281 patients at a Level I trauma center. Patients were excluded for skeletal immaturity, open fractures, polytrauma, and concurrent surgeries. Variables were assessed included age, sex, body mass index, OTA/AO classification, Weber classification, 1-year reoperation status, surgeon specialty, and use of syndesmotic screws, locking plates, and cannulated screws. Construct cost was determined by using electronic medical record implant model numbers cross-referenced with the chargemaster database. Statistical analysis involved intergroup comparative tests, regression analysis, and goodness-of-fit analyses.

RESULTS: Implant cost was different among OTA patterns (P < 0.01), highest among 43C ($3771) and lowest with 44A ($819). Construct costs of OTA 43 fractures varied from $2568 to 3771, whereas OTA 44 ranged from $819 to $1474. Costs were comparable across Weber patterns (P = 0.15), with Weber B having the highest ($1494). Costs were highest among reconstructive, podiatry, and spine surgeons, with mean costs of $1804, $1404, and $1396, respectively. Traumatologist constructs had the lowest overall price ($987). A total of 433 (33.8%) procedures used locking plates with 512 (40.0%) using at least one cannulated screw. Locking plates averaged a larger total implant cost ($1947) than nonlocking plates ($1313) but had a comparable reoperation rate (18.5% vs. 17.7%, P = 0.81). Use of a cannulated screw presented a higher total cost ($2008 vs. $1435) with comparable reoperation rates (17.4% vs. 18.8%, P = 0.72). A total of 401 (31.5%) patients received syndesmotic fixation and a significantly higher reoperation rate (17.0% vs. 11.0%, P < 0.01). Overall, 199 patients underwent elective hardware removal, 23 were infected, 7 required revision, and 3 were identified with a nonunion.

CONCLUSIONS: Our study demonstrated significant variability in implant costs for ankle fracture fixation and identified the key cost drivers as locking plates and cannulated screws. Surgical management of ankle fractures could be an ideal setting to pilot economic alignment between physicians and hospitals to drive value.

LEVEL OF EVIDENCE: Level III. Retrospective Cohort.