The optimum treatment protocol for acute total Achilles tendon rupture remains controversial. The options for treatment include the choice of surgical or nonsurgical treatment as well as the method of immobilization in a cast or a functional brace. Many different surgical techniques that can influence the outcome have been described in the literature. Meta-analyses have suggested a decreased rate of re-rupture but an increased risk of infections and wound problems in surgically managed patients1-4. Additionally, the use of a functional brace rather than cast immobilization with early ankle mobilization seems to reduce the rate of re-rupture and other complications in surgically as well as nonsurgically managed patients5-12. To date, studies have mainly focused on complications such as re-rupture, deep-vein thrombosis (DVT), and wound problems and the clinical evaluation of muscle strength, whereas outcomes have not been well evaluated on the basis of patient age and sex.
Two university hospitals in the south of Sweden traditionally have used different strategies for the treatment of acute total Achilles tendon ruptures. At the Orthopaedic Department in Hospital 1, the majority of patients have been managed surgically. In contrast, at the Orthopaedic Department in Hospital 2, the majority of patients have been managed nonsurgically. The purpose of this comparative retrospective case series was to describe the outcomes of treatment of acute total Achilles tendon ruptures in relation to treatment strategy, the rate of complications, function, age, and sex.
The records in the two orthopaedic departments were searched for patients who had had treatment of an Achilles tendon rupture (ICD-10 [International Statistical Classification of Diseases and Related Health Problems, Tenth Revision] code S86.0) between January 1, 2002 and December 31, 2006. Seven hundred and seventy-two records were found and were manually reviewed. Patients with a partial Achilles tendon rupture (101 patients), those in whom the rupture had occurred before January 1, 2002 (fifteen patients), those who had died before the start of study (that is, before December 31, 2007) (twenty patients), those with an incorrect diagnosis (forty-seven patients), and those with a delay between the time of the rupture and the time of treatment (more than two days for those managed nonsurgically and more than four weeks for those managed surgically) (forty patients) were excluded13,14. Also excluded were patients who had been managed partly at some other hospital (forty-three patients), those who had no valid address (eleven patients), and those with insufficient information in the medical record (eight patients). The remaining 487 patients were available for further study (Fig. 1). During the study intake period, there were no private or community-based alternative hospitals available; thus, the studied patients represented all of the patients with a total Achilles tendon rupture who were managed in the region. The diagnosis usually was based on clinical findings, but ultrasonography was used in some cases. Re-rupture was defined as a new Achilles tendon rupture occurring in the same leg by December 31, 2007. Deep-vein thrombosis was diagnosed by means of venography or ultrasonography. Infection was considered to be present when the patient records stated that antibiotics had been given.
To examine outcome, the Achilles tendon Total Rupture Score (ATRS) questionnaire15, a validated, injury-specific, self-reported instrument for patients with total Achilles tendon rupture, was administered to all 487 patients. The ATRS questionnaire consists of ten items (each constructed as a Likert scale from 1 to 10) that evaluate symptoms and physical function. The maximum score is 100, representing full physical function15. ATRS results were excluded for patients who answered fewer than seven questions.
The patients who completed the ATRS questionnaire and still lived in the region (n = 398) were ranked according to rupture date. Every third patient then was selected until 360 patients (180 from each hospital) had been selected. These patients were invited for clinical testing (Fig. 1). The clinical examination, performed by the same physiotherapist (I.Å.), involved measuring the circumference of the thickest part of the calf with a tape measure and counting the maximum number of repeated heel raises (concentric work) that could be performed. The number of heel raises on the injured side was divided by the number of heel raises on the noninjured side; the result was expressed as a percentage and was designated as the Limb Symmetry Index16. The duration of follow-up for complications ranged from one to seven years following the rupture.
Treatment Strategies at the Two University Hospitals (see Appendix)
The primary treatment at Hospital 1 was surgical, and the primary treatment at Hospital 2 was nonsurgical. Nonsurgical treatment involved the use of a functional brace and interaction with a physiotherapist during and after brace treatment. All patients were seen by a physician two weeks after rupture and at the time of removal of immobilization. Because of a discrepancy with respect to treatment indications and rehabilitation between Hospital 1 and Hospital 2, surgically managed patients at Hospital 1 and Hospital 2 and nonsurgically managed patients at Hospital 1 and Hospital 2 were not merged into single surgical and nonsurgical treatment groups.
Ethics
All patients received written information about the purpose and procedure of the study, and written and informed consent was obtained from patients who answered the questionnaire and participated in the clinical evaluation. Ethical approval was obtained from the Regional Ethical Review Board at Hospital 2.
Statistical Methods
The Fisher exact test was used to compare the groups in terms of the rates of re-rupture, infection, and DVT; the Mann-Whitney test was used to compare the groups in terms of the ATRS, heel-raise, and calf-circumference results; and the Spearman rho was used to calculate correlations between the ATRS and heel-raise results and age. The level of significance was set at p < 0.05.
Source of Funding
The study was funded by the Swedish Medical Research Council, Medical Faculty of Lund University, Herman Järnhardt Foundation, Österlund Foundation, and Swedish National Center for Research in Sports (CIF). The funds were used for the salary for the physiotherapist, mailing, and telephone costs.
Demographic Characteristics
The mean age of all patients at the time of the injury was forty-five years (Fig. 2). Eighty-four percent of the patients were male. The reported causes of rupture were mainly sport activities (76%; 370 of 487), including badminton (21%), soccer (16%), floorball (indoor hockey) (9%), and other sports (30%). Lower-energy traumatic injuries, such as a stumble on an indoor stair or a slip on the pavement, were seen in seventy-eight (16%) of the 487 cases. The median time to re-rupture after any treatment was eighty-eight days (range, forty-four to 1315 days). The mean age at the time of the index rupture among patients who had a re-rupture (43.8 ± 14.1 years) was similar to that among patients who did not have a re-rupture (45.1 ± 13.1 years). Eighty percent of the re-ruptures occurred within six months after the index injury. No patient had a re-rupture during the six to eight-week treatment period, and three re-ruptures (all in patients who were managed nonsurgically) occurred more than one year after the index injury.
Three hundred and ninety-eight patients (82%), with a mean age of forty-six years, answered the ATRS questionnaire. The mean time between the injury and the return of the questionnaire was 3.6 ± 1.4 years (range, 1.0 to 6.6 years).
Of the 360 patients who were invited for clinical testing, 168 accepted. Because of an injury or other functional defect (such as knee arthritis) involving either leg during the testing situation, thirty-three of these 168 patients were excluded (Fig. 1). The remaining 135 patients who were studied clinically were similar to the entire group of patients in terms of age (forty-seven compared with forty-five years), male sex (78% compared with 84%), surgical treatment (56% compared with 45%), and time between the rupture and clinical testing (3.6 compared with 3.6 years).
Of the 241 patients from Hospital 1, 201 (83%) underwent surgery and were immobilized in a cast for two weeks followed by four to six weeks with a cast or a functional brace. Of the 246 patients from Hospital 2, 227 (92%) underwent nonoperative treatment and were immobilized in a functional brace for eight weeks (Fig. 2).
Complications
Of the 201 patients who were managed surgically at Hospital 1, six (3%) had a re-rupture, three (1.5%) had an infection (all superficial), and two (1%) had a DVT. Of the 227 patients who were managed nonsurgically at Hospital 2, fifteen (6.6%) had a re-rupture and four (1.8%) had a DVT. The differences between these two main groups, with respect to all three complications, were not significant (Table I).
ATRS, Heel-Raise, and Calf Circumference
Comparison of the results for the patients who were managed surgically at Hospital 1 with those for the patients who were managed nonsurgically at Hospital 2 revealed no significant difference in terms of the ATRS (81.7 versus 78.9 points; p = 0.1). However, both the difference in the heel-raise test (Limb Symmetry Index) (p = 0.01) and the difference in calf circumference (p = 0.01) were better for the surgical treatment group (Table I).
Correlation Between ATRS, Heel-Raise, And Age
A combined analysis of all surgically managed patients showed no correlation between the ATRS or the heel-raise Limb Symmetry Index and age. Likewise, an analysis of nonsurgically managed men showed no correlation between the ATRS or the heel-raise Limb Symmetry Index and age. However, an analysis of nonsurgically managed women (n = 47) showed significant worsening in both the ATRS and the heel-raise Limb Symmetry Index with increasing age at the time of injury (see Appendix).
In spite of several studies on Achilles tendon rupture, there is still no consensus regarding patient selection for surgical or nonsurgical treatment. The main issue for patient selection is the elevated risk of infection associated with surgery compared with the elevated risk of re-rupture associated with nonsurgical treatment. The present study evaluated a sizeable cohort of patients with an Achilles tendon rupture, which allowed for an analysis of surgical versus nonsurgical treatment as well as the roles of age and sex.
In our analysis, we could not confirm a clear advantage associated with the surgical treatment of acute Achilles tendon ruptures. However, at long-term follow-up, the heel-raise test and the difference in calf circumference both favored surgery. This finding may be particularly true for older female patients, who thus might benefit from surgery. Nonetheless, it is our clinical impression that, in practice, these differences in heel-raise strength and calf circumference are not clinically important for most patients. The good ATRS results seen in both surgically and nonsurgically managed patients support this contention.
The retrospective cohort study design provided no proof of causality, and it is open to bias, but it gives more information about specific subsets of the cohort. In particular, in the present study, it was possible to evaluate outcomes in elderly patients and female patients, subgroups that previously have been less well investigated. However, the fact that older patients at Hospital 1 were more likely to undergo nonsurgical treatment creates a possible selection bias. These patients could have had concomitant systemic diseases and therefore subsequently could have experienced a higher rate of complications or inferior ATRS or heel-raise test scores. Accordingly, at Hospital 1, nonsurgically treated patients may have had worse results and surgically treated patients may have had more positive results because of this bias.
When evaluating the ATRS and heel-raise results for female patients and their correlation with age, we also separately analyzed nonsurgically managed female patients at Hospital 2 only (where almost all female patients were managed nonsurgically and no obvious selection bias existed) and found results that were similar to those for all nonsurgically managed female patients (that is, worsening results with increasing age). This finding makes us speculate that it was more likely the treatment rather than patient selection that made the difference in the results. There was a tendency to select older female patients for nonsurgical treatment at Hospital 1. If age per se is a predictor of inferior results, this would constitute a bias.
The number of patients participating in the clinical examination and answering the ATRS questionnaire differed between Hospital 1 and Hospital 2 (eighty compared with fifty-five). However, both of these groups of patients were similar to the total group of 487 patients with respect to age, sex, the percentage undergoing surgery, and the time between the rupture and completion of the ATRS questionnaire.
The decision to invite only 360 of the eligible 398 patients was based on calculations of personnel resources because they were selected by lottery. It is unlikely that the remaining thirty-eight patients who were not invited differed from the rest of the patients to a degree that would substantially impact the results.
The results presented in Table I regarding complications were documented in the medical records and were confirmed with the questionnaire responses. Because of the low number of patients who were lost to follow-up (thirty-eight), it is not likely that patients who failed to answer the ATRS questionnaire or who had moved to another location had complications that would have changed the results.
The number of infections in the present study (three) was low. The infections were not considered deep, and they caused no sequelae. This could have been due to the fact that prophylactic antibiotics were used intraoperatively. The relatively low frequency of re-rupture and other complications in the nonsurgically managed patients at Hospital 2 suggests the use of a functional brace and a standardized rehabilitation program with physiotherapy is advantageous and that such a program may also be a good addition to further improve results in surgically managed patients.
The findings of the present study are in agreement with those of a recent study that showed improved results in association with nonsurgical treatment with use of a more functional rehabilitation program11. A screening program as demonstrated by Kotnis et al.17, in which only persistent separation between the torn tendons ends with the foot in plantar flexion is an indication for surgery, may facilitate the selection of patients who might benefit surgical treatment. The need for additional study comparing surgical and nonsurgical treatment in females (and maybe also males) over the age of fifty years appears obvious; however, to reach statistical power, it probably needs to be carried out as a multicenter study. The effect of the specific immobilization method (functional brace versus cast) and its effect on the ATRS also need to be further addressed.
In conclusion, the good ATRS scores in the nonsurgically treated group, together with the relatively low rate of re-rupture and other complications, make this treatment a preferable option for most patients. However, the tendency for a lower re-rupture rate and better performance in terms of heel raises following surgical treatment suggests that surgery may be beneficial for some patients.