Operative treatment of displaced intraarticular calcaneal fractures: long-term (10-20 Years) results in 108 fractures using a prognostic CT classification.

OBJECTIVE: The primary purpose of this study was to determine whether the Sanders computed tomography (CT) scan classification was still prognostic for outcome when long-term (10-20 years) radiographic and functional data of patients after open reduction and internal fixation for Sanders type II versus type III displaced intra-articular calcaneal fractures (DIACFs) were compared. The secondary purpose was to assess whether a bone graft or a locked plate was needed to maintain a reduction over time.

DESIGN: Prognostic case-control study.

SETTING: Level I trauma hospital.

PATIENTS: Patients with operatively treated Sanders type II/III DIACF managed between January 1, 1990, and December 31, 2000, by a single surgeon were identified from a prospectively gathered database. Skeletally mature patients with a closed isolated DIACF and a minimum of 10-year follow-up were included in this analysis. All fractures were classified according to Essex-Lopresti and Sanders. Of 638 fractures, 208 met the inclusion criteria.

INTERVENTION: Surgery consisted of a lateral extensile approach, posterior facet reduction, and lag screw fixation, followed by reduction of the anterior process and tuberosity with the application of a nonlocked lateral plate. Neither bone graft nor locking plates were used.

MAIN OUTCOME MEASURES: Articular congruity and overall reduction were assessed by CT scan and plain radiography (Böhler and Gissane angle) immediately postoperatively and at the final follow-up examination in all patients. Functional assessment and outcome scores were obtained [AOFAS-AHS, the Maryland Foot Score, Short Form-36 (SF-36), Ankle Osteoarthritis Score (AOS), and Visual Analog Scale (VAS)], and all complications and/or subsequent surgeries were noted. A subtalar (ST) arthrodesis was considered a treatment failure and was used as the determining outcome variable for comparing the 2 groups (II vs.

III) RESULTS: One hundred eight fractures in 93 patients were available for follow-up at a minimum of 10 years (52%). Average follow-up was 15.22 years (range, 10.5-21.2 years). Eighty were joint depression (J) and 28 were tongue-type (T) fractures. There were 70 Sanders type II and 38 Sanders type III fractures. On immediate postoperative CT scan, posterior facet reduction was anatomic in 103 fractures (95%), near anatomic in 3 fractures (1-3 mm), and approximate in 2 fractures (3-5 mm step). There were no failed reductions (>5 mm step). Long-term results indicated that only 3 fractures settled, but no plates failed. There was 1 missed peroneal tendon dislocation. Seven patients had sural neuritis. Twelve fractures (11%) required local wound care for apical necrosis. One patient had a dehiscence resulting in osteomyelitis, requiring a ST fusion. Thirty-one fractures (29 patients) developed ST arthritis, requiring an arthrodesis (30 ST, 1 triple) for unrelenting pain (VAS, 8-10) during the follow-up period, resulting in an overall long-term failure rate of 29%. Further breakdown by fracture type revealed that an ST fusion was performed in 47% of type III fractures (18/38) versus only 19% of type II (13/70) fractures (P = 0.002). Type III fractures were 4 times more likely to need a fusion compared with type II fractures (relative risk = 3.94; 95% confidence interval, 1.64-9.48). The remaining 66 patients (77 fractures) who did not require a fusion were evaluated for long-term functional outcome. Of these, only 1 patient used a cane and had a limp. Seventy-seven percent of the nonfused group (51/66) were within the US norm for the SF-36 PCS, with 46% (30/66) above the norm. The average AOFAS-AHS was 75. The average VAS was 1.75, with scores of 0-1 (very little or no pain) seen in 56% of this subset of patients (37/66).

CONCLUSIONS: Based on the results of this comparative analysis, the Sanders classification remains prognostic; after a minimum of 10 years, type III fractures were 4 times more likely to need a fusion than type II fractures. Secondarily, it seems that neither a locked plate nor a bone graft is required to maintain a reduction over time, as virtually no loss of reduction was seen in this series (3/108, 0.9%). The "joint first" surgical treatment did not adversely affect calcaneocuboid joint outcome. Based on these results, if severe posttraumatic ST arthritis does not occur, long-term (10-20 years) functional results with mild pain, minimal alterations in activities of daily living or work, and essentially normal shoe wear can be expected from a properly performed open reduction and internal fixation. Patients must be counseled regarding difficulty with uneven ground and an inability to return to vigorous sports activities.

LEVEL OF EVIDENCE: Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.