Abstract
From 1984 to 1994, 40 patients with a subtalar fusion were reexamined in long-term follow-up. The evaluation of the overall results was carried out with three different scoring systems. The calcaneal fractures were primarily treated nonsurgically in 23 patients (57.5%), and 17 patients (42.5%) had undergone ORIF. Complete pain relief was achieved in 52.5% of patients; 65% revealed a restriction in a range of motion in the ankle joint, and in 62% a grade 1 arthritis of the ankle joint was found. The statistical analysis could only reveal a tendency for a better outcome in the operative group compared with the nonsurgical group, although within the operated group, the majority of the os calcis fractures were more severe than in the nonsurgical group.
INTRODUCTION
In the treatment of intra-articular calcaneal fractures, the optimal management of posterior facet comminution continues to be under dispute. 1,2,7,16,23 The successful anatomic reconstruction of the subtalar joint is, in some cases, neither technically possible because of severe comminution nor biologically obtainable because of cartilage damage. Independent from the treatment modality, a subtalar fusion seems to be inevitable as a sequela of this injury. In view of the destiny of these fractures, there are several different treatment concepts that can be taken into account. The therapeutic options include nonsurgical treatment, closed or “semi-open” reduction and minimal invasive osteosynthesis (K-wire fixation), open reduction and internal fixation (ORIF) for reconstruction of length, height, and width of the calcaneus, and ORIF of the calcaneus and primary fusion of the subtalar joint.
The decision to apply the nonsurgical or semi-open reduction is based on the assumption that any remaining deformity (hindfoot varus/valgus or dorsiflexion of the talus) will be correctable at the time of subsequent arthrodeses.
The benefit of an ORIF is the restoration of the hindfoot anatomy, which may prevent subsequent problems such as transfer arthritis to adjacent joints and muscle imbalances caused by persistent malalignment. Furthermore, subsequent subtalar fusion is technically not as challenging, because the hindfoot anatomy is already restored.
A primary subtalar arthrodesis requires a profound experience of the surgeon in calcaneal surgery. The main benefit for the patient is the avoidance of an interim period of significant pain and discomfort and the risk of another operation.
Regarding the long-term results after subtalar fusion, there is no scientific evidence whether all concepts are equivalent. If a subtalar fusion is performed, the functional results are mainly influenced by the following pathobiomechanical aspects:
3,7,14,18,19,21,22
restoration of the anatomic hindfoot axis (varus/valgus correction); talometatarsal axis (dorsiflexion of the talus); resection of the lateral “bulge” (peroneal impingement/fibular abutment); and shortening of the lateral foot column with medial rotation of the talus and forefoot abduction.
The aim of this retrospective study was to compare the long-term results of subtalar fusion for posttraumatic arthritis after an intra-articular calcaneal fracture. One group was primarily treated nonsurgically; the other group was treated with ORIF.
Additionally, the validity of our newly developed questionnaire was evaluated on the basis of comparison with our own clinical and radiological scoring system and the American Orthopaedic Foot and Ankle Society ankle/hindfoot evaluation protocol.

A, Technique of subtalar fusion with two cancellous 7.0 mm screws, B, Correction of the hindfoot with distraction bone-block. C, Lateral weightbearing radiogram 3 years after calcaneal fracture. D, Five-year follow-up radiogram.
MATERIALS AND METHODS
From May 1984 to May 1994 at the Trauma Department of Hannover Medical School, a total of 91 subtalar fusions were performed. In our study we included only patients with a complete radiological series of the primary fracture and the follow-up who met the following criteria: (1) they had been fused because of posttraumatic subtalar arthritis after an intraarticular calcaneal fracture; and (2) if the calcaneal fracture had been treated operatively initially, only patients who had been treated with an ORIF were included. Of these remaining 48 patients, 40 (83.3%) were reexamined in a long-term follow-up, and eight were lost to follow-up. The patients were subdivided into two groups:
Group 1: Nonsurgical treatment of the calcaneal fracture (in those days, using a plaster) consisting of 23 patients (57.5%), 16 (70%) of which had been treated in another hospitals; and Group 2: ORIF of the calcaneal fracture with 17 patients (42.5%); all had been treated at our institution.
The indication for surgery in all patients (38 patients) was persistent pain with hindfoot deformities in combination with radiological signs of subtalar arthritis. In all patients, a consequent nonsurgical treatment with medico-physical therapy and appropriate insoles had failed for pain relief.
In 27 patients (67.5%), the Oilier approach, in seven (17.5%) the lateral approach according to Palmer, and in six (15%), the extended posterolateral approach was applied. In cases where a substantial bony defect after resection of the cartilage and osteosclerotic bone was seen, an autogenous corticocancelleous bone graft was harvested from the anterior iliac crest (Fig. 1). In only a few cases was a bone grafting or a “molding graft” not necessary. In one nonsurgically treated patient, an osteotomy according to the technique described by Romash 17 was performed (Fig. 2).
For osteosynthesis, two cannulated cancellous screws were located in a dorsoplantar direction in the earlier years, and they were placed plantardorsally since 1990. The aftertreatment was performed in a therapy boot (Fig. 3), which was functional in all cases without plaster. The physical therapy was comprised of isometric, proprioceptive coordinative exercises. Immediately after the operation, a continuous passive motion machine was applied for ankle motion. Partial weightbearing was allowed as tolerated. Full weightbearing was started at 6 weeks and after 12 weeks in patients with considerable autogenous corticocancellous bone grafts.

Nonsurgically treated os calcis fracture with a nonunion in the primary fracture line. A and B, Preoperative radiogram (lateral weightbearing; Broden view). C and D, Two-year follow-up radiograms (lateral weightbearing; Broden view).
Retrospectively, the calcaneal fractures were analyzed and classified according to Sanders' system. 20 At follow-up, the clinical evaluation included the assessment of the soft tissues, the range of motion (ROM) of the adjacent joints, gait, and function. The radiological examination consisted of bilateral weightbearing x-rays of the foot and ankle with dorsoplantar and lateral views. Arthritis in the adjacent joints was defined radiologically as grade zero with a normal joint space without degenerative cysts or sclerosis; first degree was defined as having subchondral sclerosis and narrowing of the joint space; second degree was defined as having osteophytes and cyst formation; and third degree was defined as having joint destruction with complete loss or ankylosis of joint space. The patient's subjective assessment included standardized questions. The evaluation of the overall results was carried out with three different scoring systems.
The following scoring systems were applied:
Clinical Scoring System (Hannover Scoring System) (Table 1B). Outcome Questionnaire: hindfoot rating symptoms and functional status adapted for the foot from the Symptoms Severity Scale and Functional Status, according to Levine et al.,
11
developed for carpal tunnel syndrome (Table 1A). Clinical Rating System (CRS) according to Kitaoka et al.
10
The Hannover Scoring System summarizes in a 100-point system the objective, functional, and radiological exams (weightbearing radiograms of the foot in two planes, ankle anteroposterior, Broden 20°). The Outcome Questionnaire evaluates subjectively, the success of the treatment by means of 20 questions (pain, [five questions]; swelling, [five questions]; and function, [10 questions]). Every question offers five graduated answers. For reproducibility, the questionnaire was evaluated at the day of reexamination and 3 weeks later by the patient himself. The Clinical Rating System separately evaluates the hindfoot and the forefoot with regard to pain, function, and foot axes.
The statistical evaluation was performed by analysis of variance. The level of significance was based on P = 0.05, and the calculation of the Pearson's correlation coefficient.

Aftertreatment with the therapy boot (Variostabil, AD-IPROMED-Orthotech, Munich, Germany).
RESULTS
There were 33 male and 7 female patients included in this study. The subtalar fusion was performed after a mean of 3.5 years (3 months-20 years) after the calcaneal fracture. In total, 13 surgeons had operated on 40 patients. The interval between subtalar fusion and follow-up was 5.2 years (range, 2–11 years) on average. The mean age at the time of follow-up was 47 years (range, 26–61 years). The indication for the fusion in 22 (55%) patients was subtalar arthritis alone, in 14 patients (35%) subtalar arthritis with a significant hindfoot deformity, and in 4 patients (10%) a nonunion in the primary fracture line with subsequent subtalar arthritis. For realignment of the hindfoot foot axis, autogenous bone grafting from the iliac crest had been applied in 33 (82.5%) patients. With 21 (91%) patients, additional bone grafting had to be performed more often in the nonsurgical group compared with 12 (70.6%) patients of the ORIF group. Postoperative complications occurred in 8 (20%) patients; in 4 patients a hematoma needed to be evacuated, and in 2 patients a superficial infection could be cured by nonsurgical means. As a major complication there was one nonunion and one deep infection that healed after an additional surgical procedure. In one patient, a tenolysis of the tibialis posterior tendon and of the peroneal tendons had to be performed 2 years postoperatively. The implant removal of the screws was performed in 18 (45%) of the patients.
Fracture Classification
The retrospective analysis of the calcaneal fractures revealed that 94% of the patients of the ORIF group had had severely displaced fractures Sanders type III/IV, whereas in the nonsurgically treated group, only 39% were classified as Sanders type III/IV, although most patients of this group had sustained a minor joint destruction (Table 2).
Subjective Assessment
At the time of follow-up, 21 patients had complete relief of pain (58% ORIF; 50% nonsurgical). The subjective assessment of patients' function indicated restrictions in daily living in 8 patients (three ORIF, five nonsurgical). The subjective overall assessment revealed excellent or good results in 21 patients (60% ORIF, 45% nonsurgical).
Clinical Evaluation
The clinical examination revealed hindfoot deformities in 9 patients (22.5%) (a varus or valgus malalignment). Five (22%) primarily nonsurgically treated patients had a hindfoot valgus; one had a hindfoot varus. Four (23.5%) of the operated patients had a hindfoot varus deformity. Fourteen patients (24% ORIF, 41 % nonsurgical) had no or only minor restriction of ROM in the ankle joint; 26 patients (76% ORIF, 59% nonsurgical.) have shown >5° restriction of ROM in the ankle joint compared with the uninjured side. Twenty patients were able to walk without any restriction (59% ORIF, 44% nonsurgical). Thirty-one patients (83% ORIF, 73% nonsurgical) could use normal shoe wear.
Radiological Results
Radiological analysis postoperative
The radiological evaluation revealed no significant difference between the two groups. In 21 patients (53%), an anatomical or almost anatomical reconstruction of the foot axes had been achieved. Anatomical alignment was achieved in 59% of the ORIF group and in 48% of nonsurgically treated patients. There was no change in the foot axes over the course until the time of follow-up. In 36 patients (90%), the results of the radiological analysis of the foot axes correlated with the clinical findings at the time of follow-up. In four patients (all nonsurgical) there was no correlation. In two patients the foot axes were anatomical, but the patients reported pain, and the hindfoot scores were poor. In those patients, a transfer arthritis in the adjacent joints was found (one patient was grade 2 in the talonavicular and calcaneocuboid joint; another was grade 1 in the ankle and grade 2 in the calcaneocuboid joint). In another two patients, the radiological analysis revealed a nonanatomical reconstruction with anterior talar impingement and a broadened hindfoot in varus position, but both patients had excellent results for the hindfoot score (89 and 92 points).
Hannover Questionaire: Symptoms-Severity Scale and Functional Status (Foot and Ankle)
Hannover 100-Point Scoring System (Foot and Ankle)
Classification of the Calcaneal Fractures According to Sanders 20
16 patients of this group were treated in other hospitals. N = 23.
All patients were treated in our hospital. N = 17.
Radiological analysis follow-up
After a mean follow-up period of 5.2 years, there was 1st or 2nd degree arthritis in the ankle joint in 25 patients (62.5%; 59% ORIF, 65% nonsurgical), in the talonavicular joint in 24 patients (60%; 59% ORIF, 60% nonsurgical), and in the calcaneocuboid joint in 16 patients (40%; 53% ORIF, 30% nonsurgical). The differences of the calcaneocuboid joint arthritis correspond to the higher fracture frequency involving this joint. In one patient (nonsurgically treated Sanders type III fracture), a triple fusion (talonavicular joint/calcaneocuboid joint) had to be performed 3 years after the subtalar fusion.
In 23 patients with arthritis in the adjacent joints, seven had a good or an excellent result according to all scoring systems. There was no correlation between the transfer arthritis and the time of fusion or follow-up.
Scoring
Fourteen patients (30% ORIF, 35% nonsurgical) reported a significant improvement and were very satisfied at the time of follow-up; 21 patients (60% ORIF, 45% nonsurgical) reported an improvement compared to the time before the fusion; five patients (10% ORIF, 20% nonsurgical) were not satisfied and classified their result as poor (Table 3). The comparison of the overall results, analyzed by the different scores, showed for the ankle/hindfoot score (CRS) a mean of 69 points (range, 26–87 points) in the ORIF group versus 65 points (range, 28–94 points) in the nonsurgical group. Using the Hannover score, a mean of 62 points (range, 33–89 points) was found in the ORIF group versus 61 points (range, 24–89 points) in the nonsurgical group. The forefoot scores (CRS) according to Kitaoka et al. were good and excellent in nearly all patients: hallux score 86 points (63–100 points) and phalangeal score 85 points (range, 71–100 points) (Table 3).
The evaluation of the questionnaire has revealed for the ORIF group a mean of 66 points (range, 34–89 points) and for the nonsurgical group, 64 points (range, 32–90 points). The analysis of variance revealed statistical difference between neither the two different treatment groups (operative versus nonsurgical) nor within the three scoring systems. The good reproducibility of the questionnaire (repeating the questionnaire after 3 weeks) could be reconfirmed by a correlation coefficient of r = 0.91. The good results according to the forefoot scores (Kitaoka) revealed that the patients' complaints derived from the ankle joint, hindfoot, and midfoot. Ten patients were able to perform some sporting activities at a low level.
DISCUSSION
Painful subtalar arthritis is a common sequela after displaced intra-articular os calcis fractures. The time of onset cannot be predicted and is highly variable, depending in part on the degree of comminution, cartilage destruction, and persistent hindfoot deformity. 5,8,9,13,15 However, the indication for fusion today is more dependent on the patient's symptoms and the radiological findings than on existing hindfoot deformities.
The high rate of 80% bone grafting in our series reflects the severity of the calcaneal fractures and the long interval between fracture and fusion of 3.5 years, on average. In recent years, we have modified our concept with a higher rate of primary fusions in the Sanders type IV fractures or an early fusion within 1 year, according to the symptoms or persistent deformities.
Results of the Scoring Systems
CRS, Clinical Rating System; op, operative; non, nonoperative.
Hindfoot Score of the Clinical Rating System. 10
Forefoot Score of the Clinical Rating System. 10
Repeated after 3 weeks for internal consistency.
Five years after the fusion, 55% of our patients were painfree and 83% could wear normal shoes, but 60% of the patients had arthritis in the adjacent joints. It is remarkable that these results were independent of the primary treatment, although the calcaneal fractures in the ORIF group had been much more severe than in the nonsurgically treated group.
The high rate of arthritis in the adjacent joints has not been described in the literature yet. This might be related to the late timing of fusion and to the high rate of 75% of severe intra-articular fractures Sanders type III/IV. Most of the authors report a rate of arthritis in adjacent joints of 20% but didn't find a correlation to the overall clinical result. 6,12,19 Thirty percent of our patients with arthritis in the adjacent joint had a good or excellent result in the hindfoot scores.
Nonoperatively treated intra-articular calcaneal fractures may lead to a severe hindfoot deformity that only can be realigned through a calcaneal osteotomy in the primary fracture line. 17 Therefore, in a primary fracture reduction, at least a gross alignment of the hindfoot axis (i.e., semi-open technique) is required to obtain a better starting point for a later fusion.
In 80% of our subtalar fusions, an autogenous corticocancellous bone graft was performed, reflecting the major deformities and the bone loss with subsequent gaps for realigning the hindfoot in the sagittal and frontal plane.
For fixation we used 2–3 cancellous screws (7.0 mm) because, in our opinion, the fixation with a single cancellous screw 6 is not sufficient for fusion in posttraumatic subtalar arthritis because of the frequency of poor bone stock. Since 1991, we have changed our technique with a screw placement from the plantar heel (i.e., the calcaneus) into the talus. The advantages are minor risk of neurovascular lesions, no anterior impingement of the screw heads in ankle dorsiflexion, and a better biomechanical placement and compression of the screw in the talus instead of in the calcaneus.
A high rate of 80% to 90% good and excellent results are reported for subtalar fusions for degenerative arthropathies. 4,18,19 But in posttraumatic arthritis after calcaneal fractures, these rates drop to approximately 50% to 60%, 12,24 comparable to our series.
CONCLUSION
In our study, the comparison of the outcomes after subtalar fusion has revealed equivalent results within both groups. However, the operated calcaneal fractures had been classified as more severe compared to the less severe fractures, which had been treated nonoperatively. We could not find a correlation for the overall outcome and the timing of the arthrodesis. However, we now perform fusions earlier, either primarily in Sanders type IV fractures or in symptomatic cases, within 12 months after injury. A prospective study with more patients is necessary to evaluate the benefits of an earlier fusion.
The comparison of the different scoring systems revealed no significant difference between the newly developed questionnaire and the clinical scoring systems. The slightly poorer results in the our clinical scoring systems reflect the poor radiological outcome in most of the cases. In summary, although in the group of primary operative treatment the calcaneal fractures had been more severe, the results were equivalent within all three scoring systems.
