Patient Selection
We reviewed the records at our university hospital to identify patients who had sustained a complete closed Achilles tendon rupture. Patients who had had either a rerupture or a deep infection after any form of treatment were selected for the present study. The records of these patients were retrospectively reviewed in order to extract data on age, gender, hometown, symptoms, previous treatments, lifestyle, habits, details of treatment, mechanism of injury, aftercare, and rehabilitation. All of the involved patients were invited to return to our university hospital for a clinical examination and physiotherapy tests, which were performed free of charge.
Clinical Examination
All of the patients completed a questionnaire at the time of the clinical examination and were examined and interviewed by one of the authors (A.P.). The clinical outcome was assessed with use of the scoring method described by Leppilahti et al. 16 , which includes items related to clinical factors such as pain, stiffness, muscle weakness, footwear restrictions, range of active ankle motion, subjective outcome, and isokinetic calf-muscle strength (Appendix).
Physical Performance Tests
The functional tests were carried out by our hospital's physiotherapists, and the calf-muscle strength tests were performed by a special physiotherapist with use of a Lido Multi-Joint II dynamometer that was linked to a microcomputer (Loredan Biomedical, Davis, California). A ten-minute warm-up of ergometric cycling was performed before the testing, and the patient performed some submaximal and maximal repetitions at the isokinetic test velocity in the supine position with the knee straight before the actual testing. The peak torque of plantar flexion and dorsiflexion of the ankle was measured in N-m at velocities of 60°ree;/s, 120°ree;/s, and 180°ree;/s. Four maximal voluntary muscular torque contractions were required at each velocity, and the best result was recorded. After the isokinetic tests, the maximal isometric plan tar flexion strength was measured with the ankle in neutral position. The isokinetic strength scale for scoring peak torque of the ankle during plantar flexion and dorsiflexion at the three test velocities was used to analyze the strength results 16 .
Statistical Analysis
The summary statistics for continuous variables were expressed as the mean and the standard deviation. The Mann-Whitney U test was used to calculate differences between continuous variables, and Fisher's exact test was used to calculate differences between frequencies. Two-tailed p values were reported, and the level of significance was set at p < 0.05. The analyses were performed with use of SPSS (version 10.0; SPSS, Chicago, Illinois).
Patient Information
A total of 409 patients with a complete closed Achilles tendon rupture were treated at our university hospital between 1979 and 2000. Twenty-three patients had a rerupture (Appendix). This group included twenty-one men and two women with a mean age of forty years. These patients represented all social classes and a variety of occupations. Nineteen (83%) of the twenty-three patients had sustained the primary rupture during participation in sports activities, most commonly badminton (seven patients) and volleyball (five patients).
Nine patients had sustained a deep infection after surgery (Appendix). This group included seven men and two women with a mean age of fifty-three years. Only two of these patients had sustained the primary Achilles tendon injury during participation in sports; the other seven had had accidents of various kinds during the normal activities of daily living. In six of these patients the infection had occurred after the primary operation, and in three it had occurred after a second operation was performed to treat a rerupture.
Altogether, twelve patients with a rerupture and seven with a deep infection were available for a follow-up evaluation at a mean of 4.1 years after the primary Achilles tendon rupture. Of the remaining ten patients, one had died, one (who had a well-healed rerupture) was unable to attend because of work obligations, three had a follow-up visit scheduled within less than six months, and five (one of whom was living abroad) did not reply to our questionnaire.
Primary Treatment
Twenty-eight of the twenty-nine patients initially had been treated operatively, and one had been treated nonoperatively with a cast. Our standard protocol for operative treatment was tendon-suturing with absorbable size-0 nonbraided suture. The repair was augmented in twenty patients; specifically, the repair was augmented with one turn-down flap of the gastrocnemius aponeurosis (the Silfverskiöld technique) in fourteen patients, with two flaps (the Lindholm technique) in five patients, and with the plantaris tendon (the Lynn technique) in one patient. For twenty of the twenty-eight patients who had had surgical treatment, the planned postoperative rehabilitation protocol consisted of immobilization in a below-the-knee cast for six weeks, with the ankle in plantar flexion for three weeks and then in a neutral position for three weeks. Weight-bearing was allowed gradually after the first three weeks, with full weight-bearing achieved at six weeks. The other eight patients who had had surgical treatment were enrolled in a clinical trial in which a customized below-the-knee brace was used postoperatively. This device allowed active free plantar flexion of the ankle but restricted dorsiflexion to neutral. Weight-bearing was limited to one-half of body weight until six weeks, at which time active ankle exercises with full weight-bearing and strengthening exercises were allowed. All of the patients who were treated surgically were instructed to begin jogging and controlled sports activities at three months. Jumping sports and professional activities were begun at six months. The one patient who had been treated nonoperatively wore a below-the-knee cast with the ankle in plantar flexion for three weeks and then wore a second cast with the ankle in neutral for an additional three weeks.
Incidence of Rerupture and Deep Infection
During the study period, the population for which our university hospital served as the only unit treating Achilles tendon ruptures increased from 94,000 to 120,000. The annual incidence of these injuries (per 100,000 inhabitants) increased from 4.2 in 1979-1990 to 15.2 in 1991-2000, with the peak annual figure of 19.0 being recorded in 1999 ( Fig. 1 ). Twenty-three (5.6%) of the 409 patients had an Achilles tendon rerupture, and nine (2.2%) a deep Achilles tendon infection. The annual incidence of reruptures (per 100,000 inhabitants) increased from 0.25 in 1979-1990 to 1.0 in 1991-2000, with the peak value of 3.5 being reached in 1999. The annual incidence of deep infections (per 100,000 inhabitants) increased from 0 in the 1980s to 0.63 in the 1990s, with the peak value of 2.6 being reached in 1999. The proportion of reruptures to primary ruptures was 6.0% in 1979-1990 and 6.6% in 1991-2000.
Treatment of Rerupture
The twenty-three reruptures occurred at a mean of seventy-nine days (range, two to 209 days) after the primary operation. Nineteen were treated operatively, and four were treated nonoperatively with six weeks of immobilization in a below-the-knee cast. The repair of the rerupture was augmented in fifteen patients; specifically, the repair was augmented with one turn-down flap of the gastrocnemius fascia in nine patients, with two turn-down flaps in two patients, and with the plantaris tendon in three patients. In one patient, the rerupture was reconstructed with exogenous material (Leeds-Keio; Howmedica, Rutherford, New Jersey). The operative technique was not accurately described in the records of two patients.
Two patients had a second rerupture. The first patient had had two previous surgical repairs that had been performed with the end-to-end technique; the third repair was performed with use of one turn-down flap for augmentation. The other patient had had two previous surgical repairs with plantaris tendon augmentation; the third repair was done with use of two turn-down flaps for augmentation.
Treatment of Deep Infection
There were a total of nine deep infections. Six occurred after the primary repair, and three occurred after the repair of a rerupture. Wound revision was performed in all cases, and necrotic Achilles tendon tissue was totally removed from five patients during the surgical treatment. The results of bacterial cultures, recorded for six of the nine patients, were positive for Staphylococcus aureus (four patients), Staphylococcus epidermidis (one patient), and Propionibacterium acnes and diphtheroid species (one patient).
Six patients were treated with repeat débridement and primary split-thickness skin-grafting. Four of these patients lost the Achilles tendon totally during the course of treatment; in the other two, the infection resolved before all tendon tissue was removed. Two patients were managed with a microvascular radial forearm flap and a tensor fasciae latae graft after débridement. One of these microvascular reconstructions was successful, but the other failed because of thrombosis. The latter patient ultimately had total loss of the tendon and was eventually treated with split-thickness skin-grafting. One patient needed a local two-tail full-thickness transposition flap to cover the exposed tendon after débridement.
Isokinetic Calf-Muscle Strength
The mean relative peak torque deficits for plantar flexion in the affected limb at velocities of 60°ree;/s, 120°ree;/s, and 180°ree;/s were 11.6%, 12.2%, and 7.1%, respectively, in the rerupture group and 39.5%, 39.5%, and 25.9%, respectively, in the infection group. The differences between the two groups were significant (p < 0.001, p < 0.001, and p = 0.010, repectively). In the rerupture group, the isokinetic strength outcome was good for two patients, fair for seven, and poor for three. In the infection group, the isokinetic strength outcome was fair for one patient and poor for six (p = 0.036) (Appendix).
Isometric Strength
The mean relative isometric strength deficit for plantar flexion in the injured limb was 14.4% in the rerupture group (p = 0.012) and 42.1% in the infection group (p = 0.07) (Appendix).
Pain
In the rerupture group, six Achilles tendons were painless, five were mildly painful, and one was moderately painful. In the infection group, only one Achilles tendon was painless whereas four were mildly painful and two were moderately painful (p = 0.27).
Stiffness
Eight patients in the rerupture group and two in the infection group reported no stiffness in the Achilles tendon region. Four patients in the rerupture group and five in the infection group reported occasional mild stiffness (p = 0.17).
Subjective Calf-Muscle Weakness
In the rerupture group, six patients had no subjective calf-muscle weakness and six had mild weakness. In the infection group, one patient had mild subjective calf-muscle weakness, five had moderate weakness, and one had severe weakness (p < 0.001).
Footwear Restrictions
None of the patients in the rerupture group had footwear restrictions. In the infection group, three patients had no footwear restrictions and four had mild restrictions but tolerated most shoes (p = 0.009).
Range of Ankle Motion
In the rerupture group, eleven patients had a normal range of ankle motion and one had mild limitation. In the infection group, two patients had a normal range of motion, four had mild limitation, and one had moderate limitation (p = 0.01).
Subjective Outcome
In the rerupture group, three patients were very satisfied, eight were satisfied but had minor reservations, and one was satisfied but had major reservations. In the infection group, one patient was satisfied but had minor reservations and six were satisfied but had major reservations (p = 0.004). Two patients reported a definite loss of sensation in the foot: one of these patients had undergone an unsuccessful reconstruction with a radial forearm flap and a tensor fasciae latae graft, and the other had sustained damage to a branch of the sural nerve at the time of the repair of the rerupture.
Overall Result
All of the patients in the rerupture group and one of the patients in the infection group were able to walk normally. Five patients in the infection group had a mild limp, and one needed crutches. All of the patients in the rerupture group were able to participate at least in recreational sports, whereas six in the infection group had a restricted ability to participate in sports.
In the rerupture group, the ankle performance score was classified as excellent or good in eight patients and as fair in four. In the infection group, the score was classified as fair in two and as poor in five (p = 0.004).
Risk Factors
The number of patients with an age of more than sixty years who were treated for an Achilles tendon rupture increased from two in 1980-1989 to twenty-four in 1990-1999. We reviewed the patient records for other risk factors, such as corticosteroid use, smoking, symptoms in the tendon before the injury, diabetes, and delay in treatment. Nine (39%) of the patients in the rerupture group and two (22%) of the patients in the deep infection group had none of these risk factors ( Table I ). Five (56%) of the patients in the infection group had three risk factors or more, compared with only four (17%) of the patients in the rerupture group (Appendix). The patients in the infection group were significantly older than those in the rerupture group (p = 0.030), and they received corticosteroid medication more often than those in the rerupture group. The ruptures in the infection group were associated with the activities of daily living more often than they were associated with recreational sports. A delay before treatment was more common in the infection group than in the rerupture group.