Session IX - Foot & Ankle

Prospective Randomized Clinical Multi-Center Trial: Operative versus Nonoperative Treatment of Displaced Intra-Articular Calcaneal Fractures: A Two- to Eight-Year Follow-Up

Richard E. Buckley, MD; Robert G. McCormack, MD: Ross K. Leighton, MD; Graham C. Pate, MD; David P. Petrie, MD; Robert D. Galpin, MD, Multi Canadian Centers, Canada, Supported by OTA Research Grant

Purpose: A large multi-center prospective randomized clinical trial was conducted to compare operative versus nonoperative treatment of displaced intra-articular calcaneal fractures (DIACF) with 2- to 8-year follow-up.

Methods: This prospective randomized trial took place at 4 level I trauma centers. Between April 1991 and December 1997, 512 patients were treated for a calcaneal fracture. Of these 512 patients, 424 had 469 DIACF's. Seventy-two percent of patients were treated at one center, and 28% were treated at the other centers. Patients were excluded if they were not of adult age, were more than 65 years of age, had open fractures, had medical contra-indication for surgery, or had previous trauma, tumor or infection in the heel bone. They were also excluded if they had other reasons that consent could not be obtained for entrance into a clinical trial. Random numbers were used to assign patients to operative or nonoperative treatment groups. A standard protocol was used for operative care involving a lateral approach with fixation and bone grafting as determined by the 6 surgeons involved. Nonoperative patients had nothing done to the foot (there was no closed reduction performed) but were treated with ice, elevation and rest. After 6 weeks of non-weightbearing, a standardized physiotherapy protocol was initiated for all patients with full weightbearing. Two- to eight-year follow-up was maintained at all study centers by experienced surgeons and study nurse clinicians.

Main Outcome Measures: Radiological classification: Böhler's angle (plain films), Essex-Lopresti classification (plain films), Sanders classification (Computed Tomography). Computed tomography scans (CT) were taken pre-operatively, postoperatively and at 2 years. The quality of reduction of fractures was measured as anatomic, less than 2 mm, greater than 2 mm step or gap, or comminuted reduction (on CT). Clinical scores used the validated visual analogue score (VAS), the SF-36 general health survey score, time to return to work, gait, pain, work capability and complications. All fractures were classified as OTA 73-C₁, 73-C₂, 73-C₃.

Results:

	Patients	Surgeon #1	Other surgeons	SF36	VAS	Fusions
Nonoperative	218	160	58	64.7	64.3	37/222
ORIF	206	145	61	68.7	68.6	7/209
	424	305	119	P = 0.13	P = 0.12	P = 0.00

	Percent	VAS	Standard Deviation
Anatomic reduction	31.3%	72.9	22.0
Step <2mm	50.2%	69.5	24.2
Comminuted reduction	18.5%	67.3	22.7
	100%	P = 0.56	23.2 (ave)

A total of 424 patients were collected in this study with a minimum of 2-year and a maximum of 8-year follow-up. There were 44 (10%) subtalar fusions, 63 (14%) "lost to follow-up", and 7 patients (1.5%) died during the course of the study, leaving 309 of 424 for 2- to 8-year long-term follow up. There was no difference between groups demographically. A combination of plates and screws were used to reduce the fracture in 91% of the patients, with other combinations, including plates ± screws or K-wires, used 9% of the time. Bone graft was used in 28% of the operatively treated patients. There was no difference between treatment centers in relation to scores or quality of reduction. Sanders classification showed a general trend of worsening outcomes from 2A fractures to type 4 fractures (P = 0.14). Patients demonstrated a diminished outcome with a lower Böhler's angle (P = 0.02).

Discussion: This multi-center prospective randomized clinical trial involving displaced intra-articular calcaneal fractures demonstrates that nonoperative care of calcaneal fractures is as good as surgical care. The operative treatment group is marginally better at 4 points out of 100 on a standardized scoring scale (VAS) (P = 0.12). The operative treatment group is marginally better at 4 points out of 100 on the SF36 (P = 0.13). With these numbers, we would need many hundreds more patients in each group to establish a difference between groups. Not only was statistical significance not reached throughout our study, but also all surgical staff involved thought that operative care provided only marginal improvement in patient outcomes. It appears that the quality of reduction is not an important issue relating to patient outcomes (despite 206 operative patients, P = 0.56). Fractures with more comminution, as determined by Sanders classification and Böhler's angle, showed less satisfactory outcomes. No difference was noted between geographic sites in patient outcomes. The fusion rate in the nonoperative group was 6 times the rate in the operative group. Only 10% of all patients entering the study required late fusion. These numbers do not justify ORIF to protect against late fusion. A small learning curve in operating on this fracture was suggested, yet outcomes did not improve significantly through time. We think that this fracture should be treated by experienced surgeons and with careful patient selection. Currently, our data suggest that we should not operate on smokers more than 40 years of age (because of increased infection rate), patients who are non-compliant in initial soft tissue swelling management, worker's compensation patients, older or sedentary patients. Operative care is suggested for younger patients with worse fractures who are active and laboring.

Conclusion: This large prospective randomized study demonstrates that operative treatment provides no improvement over nonoperative treatment of DIACF's. Classification schemes presently used including Sanders and Böhler's angle measurement are prognostic when looking at DIACF's. Higher energy fractures (lower Böhler's angle and more comminution) have less satisfactory results. Nonoperative care more occasionally leads to late fusions. Careful stratification of this huge data bank will provide more exact indications for surgery with this vexing fracture.