Knee Surgery, Sports Traumatology, Arthroscopy

, Volume 21, Issue 6, pp 1356–1360

Cross cultural adaptation of the Achilles tendon Total Rupture Score with reliability, validity and responsiveness evaluation

Authors

    • The Princess Royal Hospital
  • Karin Grävare Silbernagel
    • Department of Physical TherapySamson College of Health Professions, University of the Sciences
  • Katarina Nilsson-Helander
    • Department of OrthopaedicsKungsbacka Hospital
  • Omer Mei-Dan
    • Department of Orthopaedic Sports MedicineThe University of Colorado Hospital
  • Jon Karlsson
    • Department of OrthopaedicsSahlgrenska University Hospital
  • Nicola Maffulli
    • Barts & the London School of Medicine and DentistryQueen Mary University of London
Ankle

DOI: 10.1007/s00167-012-2146-8

Cite this article as:
Carmont, M.R., Silbernagel, K.G., Nilsson-Helander, K. et al. Knee Surg Sports Traumatol Arthrosc (2013) 21: 1356. doi:10.1007/s00167-012-2146-8

Abstract

Purpose

The Achilles tendon Total Rupture Score (ATRS) was developed because of the need for a reliable, valid and sensitive instrument to evaluate symptoms and their effects on physical activity in patients following either conservative or surgical management of an Achilles tendon rupture. Prior to using the score in larger randomized trial in an English-speaking population, we decided to perform reliability, validity and responsiveness evaluations of the English version of the ATRS. Even though the score was published in English, the actual English version has not be validated and compared to the results of the Swedish version.

Methods

From 2009 to 2010, all patients who received treatment for Achilles tendon rupture were followed up using the English version of the ATRS. Patients were asked to complete the score at 3, 6 and 12 months following treatment for Achilles tendon rupture. The ATRS was completed on arrival in the outpatient clinic and again following consultation.

Results

The outcomes of 49 (13 female and 36 male) patients were assessed. The mean (SD) age was 49 (12) years, and 27 patients had treatment for a left-sided rupture, 22 the right. All patients received treatment for ruptured Achilles tendons: 38 acute percutaneous repair, 1 open repair, 5 an Achilles tendon reconstruction using a Peroneus Brevis tendon transfer for delayed presentation, 1 gracilis augmented repair for re-rupture and 4 non-operative treatment for mid-portion rupture. The English version of ATRS was shown to have overall excellent reliability (ICC = 0.986). There was no significant difference between the results with the English version and the Swedish version when compared at the 6-month- or 12-month (n.s.) follow-up appointments. The effect size was 0.93. The minimal detectable change was 6.75 points.

Conclusions

The ATRS was culturally adapted to English and shown to be a reliable, valid and responsive method of testing functional outcome following an Achilles tendon rupture.

Keywords

Cross culturalATRSAchillesScore

Introduction

Rupture of the Achilles tendon is increasingly common [13], and an incidence of 18 per 100,000 has been reported [11]. There has been considerable research into whether ruptures are best managed non-operatively or with surgical intervention followed by a structured physiotherapy programme [3, 8, 9, 15, 17, 18, 23, 24].

One of the difficulties with randomised controlled studies is the parameter used to determine the outcome of treatment. Re-rupture rate is a definite marker, as are the complications related to surgical intervention [9]. Strength deficits have been noted at two years following rupture [19], although several confounding factors influence this outcome measure.

Several scoring systems have been developed for the use in foot and ankle surgery. The AOFAS Ankle Hind-Foot Score is an established score [10] that has been used as an outcome measure in studies on patients with Achilles tendon rupture. However, the validity and responsiveness of this score have not been evaluated in this patient population. On the other hand, the Leppilahti [12], Hannover [22] and Merkel [14] scores are designed for evaluating the outcome in patients with Achilles tendon rupture, but these scores have in common that they combine both subjective and objective measures, which can pose difficulty when comparing different studies. Moreover, there are no studies related to the validity of the Leppilahti score.

The Achilles tendon Total Rupture Score (ATRS) was developed because of the need for a reliable, valid and sensitive instrument that could evaluate patient-reported symptoms and their effect on physical activity in patients with Achilles tendon rupture. The Score consists of ten items evaluating aspects of symptoms and function. Each item ranges between 0 and 10 on a Likert scale with a maximal score of 100 indicating no symptoms and full function [16]. This score was devised in a Swedish population and has been shown to have good reliability, validity and responsiveness for evaluating outcome related to symptoms and physical activity in patients with an Achilles tendon rupture [7, 16].

Even though the score was published in English, the actual English version has not be validated and compared to the results of the Swedish version. Prior to using the score in larger randomized trial in an English-speaking population, we felt it was important to perform reliability, validity and responsiveness evaluations of the English version of the ATRS.

Therefore, the purpose of this study was to evaluate the English version of ATRS and to compare to the Swedish version. Specifically, the purposes were to perform reliability, validity and responsiveness evaluations and calculate the minimal detectable change of the score.

Materials and methods

From 2009 to 2010, all patients who received treatment for Achilles tendon rupture were followed up using the English version of the ATRS (Fig. 1) at The Princess Royal Hospital. All patients gave informed consent for inclusion in the study and the use of their responses for analysis.
https://static-content.springer.com/image/art%3A10.1007%2Fs00167-012-2146-8/MediaObjects/167_2012_2146_Fig1_HTML.gif
Fig. 1

Cross cultural Achilles tendon Total Rupture Score

Several terms were altered from the original translation from Swedish to English. This made the explanation of the literal translation more understandable for the study group. These were as follows: question two tiredness replaces fatigue, question six ground replaces surfaces and, finally, heavy physical work replaces hard physical labour.

Patients were asked to complete the score sheet by the nursing staff as they arrived in the outpatient clinic. Sheets were collected at the start of the consultation, and a second identical score sheet was given to the patient following the consultation.

The functional scores reported allowed test–retest analysis at each visit, together with comparison between visits. Responses were compared for age, gender and treatment. These results were then compared with those obtained from the original Swedish cohort [21].

Statistical analysis

Statistical analysis was performed using SPSS (Statistical Package for Social Sciences). Data are reported both as median, mean (SD and interquartile range), mirroring how data were reported in the original article. Nonparametric statistics were used since the score presents ordinal data. Test–retest data were analysed with intra-class correlation coefficient (ICC), two-way random using average measures and Wilcoxon-paired test. As a measure of responsiveness, the effect size was calculated on the results from the 3- and 6-month evaluations. To evaluate the change with treatment, we used data from those patients who were evaluated on both these occasions. The independent samples Mann–Whitney U test was used to compare the English and Swedish versions. The effect size was calculated as the mean score difference divided by the standard deviation from the initial measurement according to Kazis et al. [6]. An effect size of >0.80 is considered high. Standard error of measurements (SEM) was calculated as follows: SEM = s √1 − r, where s is the standard deviation of the baseline measurements and r is the ICC [4]. The minimal detectable change (MDC) was calculated as follows: MDC = 2.77 × SEM [5, 20]. The level of significance was set at p < 0.05.

Results

Altogether, the outcomes of 49 patients were compared. Of these, 36 were male and 13 were female. The mean (SD) age was 49 (12) years, and 27 patients had treatment for a left-sided rupture, 22 the right (Table 1).
Table 1

Demographics

 

English version (n = 49)

Swedish version (n = 78)

Age mean (SD)

49 (12)

42 (9)

Sex (n)

n = 49

n = 78

 Males

36

65

 Females

13

13

Type of treatment (n)

n = 49

n = 78

 Surgical

45

43

 Of these:

  

  Percutaneous repair

38

 

  Open

1

 

  Open augmented repair

1

 

  Reconstruction

5

35

 Non-surgical

4

 

Time of evaluation (n)

n = 49

n = 78

 3-months

39

0

 6-months

21

78

 12-months

15

78

Most patients (38) elected to have their Achilles rupture repaired using a percutaneous technique [1]; one had an open repair performed at another institution, 5 patients had Achilles reconstruction using a peroneus brevis tendon transfer for delayed presentation [2], and one had a gracilis augmented repair due to re-rupture. Finally, four patients were managed non-operatively for mid-portion ruptures (Table 1). All patients received a structured physiotherapy programme.

Reliability

The English version of the ATRS was shown to represent overall excellent reliability (ICC = 0.986) with no significant differences between the testing occasions (Table 2). The reliability remained the same at all the different testing occasions (ICC = 0.960–0.998) (Table 2), with no significant differences between the testing occasions.
Table 2

Test–retest reliability—English version of ATRS

 

ATRS 1

ATRS 2

Wilcoxon-paired test (p value)

ICC

All n = 81

    

 Mean (median)

57 (55)

56 (53)

0.1

0.986

 SD (interquartile range)

24.7 (40)

25.7 (42)

3-month n = 39

    

 Mean (median)

46 (42)

44 (43)

0.06

0.960

 SD (interquartile range)

20.6 (36)

20.9 (34)

6-month n = 21

    

 Mean (median)

63 (65)

61 (62)

0.70

0.996

 SD (interquartile range)

22.7 (41)

24.8 (45)

12-month n = 15

    

 Mean (median)

78 (90)

79 (92)

0.29

0.998

 SD (interquartile range)

21.3 (34)

21.6 (35)

Validity

There were no significant differences between the results with the English version and the Swedish version when compared in a population of patients with an Achilles tendon rupture at the 6-month- (n.s.) or 12-month (n.s.) evaluations (Table 3).
Table 3

ATRS results (based on score 1 for the English version)

ATRS score

English version

Swedish version

3-month results

n = 39

N/A

 Mean (median)

46 (42)

 

 SD (interquartile range)

20.6 (36)

 

6-month results

n = 21

n = 78a

 Mean (median)

63 (65)

72 (77)

 SD (interquartile range)

22.7 (41)

16.8 (21)

12-month results

n = 15

n = 78b

 Mean (median)

78 (90)

88 (93)

 SD (interquartile range)

21.3 (34)

15.1 (11)

aNo significant difference between English version and Swedish version results at 6 months

bNo significant difference between English version and Swedish version results at 12 months

Responsiveness

There was a significant improvement (p = 0.001) in the English ATRS score from the 3-month evaluation to the 6-month evaluation (n = 17). The effect size was 0.93.

The minimal detectable change

The minimal detectable change of the ATRS score was 6.8 points.

Discussion

The key message of this study is that the English version of the Achilles tendon Total Rupture Score is reliable, valid and sensitive to changes over time. The results from using the English version of ATRS were also comparable to the results when using the Swedish version of ATRS. This injury-specific score can therefore be recommended to be used in both research and every day clinical work in an English-speaking population.

We acknowledge that some minor changes were made to the wording of the original score. This included the exchange of tiredness for fatigue, ground for surfaces and heavy physical work for hard physical labour. We consider these terms are better understood by the general population, permitting independent score completion rather than requiring assistance. In addition, since the score was a limitation score, responses were collected so that a more severe limitation would be reflected by a lower score. This was so that patients completed a similar Likert scale to others commonly used in our clinics.

The ATRS reproduced in English has now been tested on an English-speaking population. The score can now be used to report patient-reported outcome measures in patients following Achilles rupture in those who primarily speak English. This study does have some limitations. The first was that responses were collected in an outpatient clinic rather than a neutral setting. This may have influenced the patients’ responses. Patients also received a consent form for data collection explaining that this scoring process was part of their routine follow-up assessment and would be undertaken on each visit. This form and the first score sheet were given to the patients on arrival in clinic and collected at the start of the consultation to minimise bias. The second sheet was then completed at the end of the consultation.

The test–retest distribution of score sheets was separated by only the waiting time and the time of the consultation, typically lasting only 15 min. We acknowledge that a short time period between test and retest may have influenced the results, but point out that having patients wait in clinic would be unnecessary, require explanation and bias encouraging them to reflect on their responses. The short time period, however, has the advantage of producing scores, which may not be susceptible to day-to-day variation. The distraction of the consultation may have been adequate for test–retest, but the consultation may have also influenced the responses. For all responses, there was no significant difference between the first and second scores for each time period. By comparison, in the original ATRS article [16], a significantly higher score was reported on the second test day compared with first test day when testing was performed twice within two weeks [16].

Another limitation is the surprise factor of the second retest. Continual analysis of the same patient will be subject to bias. Upon second and third scoring visits, the patient will be expecting both the test and the retest, and so may put more thought into their responses. This would suggest that test–retest scores should demonstrate that scores should be more similar with increasing assessment following injury. This was found not to be the case.

The validity of the responses of both the English and Swedish populations was found to be similar. This suggests that the ATRS can be used as an outcome measure for randomized studies between populations from different countries.

The effect size in this study was found to be high at 0.93, indicating that the English version of ATRS is sensitive to detect changes over time. In the original paper, the effect size was between 0.87 and 2.21, where the higher value was found between the evaluation at 3 and 6 months. Since these studies have used different treatments, it is not possible to ascertain whether the differences in the effect size are dependent on the treatments or any differences in the score. Of main importance is that the score is sensitive to change over time, and further analysis needs to be undertaken to evaluate the minimal clinically important difference. For the ATRS, the minimal detectable change that exceeds the error expected for the measuring instrument used was found to be 6.8 points. Therefore, for patient groups assessed using the ATRS unless the difference between the two groups was greater than 6.8 points, it is unlikely that the difference was due to the different treatment technique. This must be borne in mind when using the English version of the ATRS as an outcome measure in future studies and for calculating power of future studies.

Still another limitation observed for the score is that it is a subjective report on functional limitation. We are aware that the expected activity demands vary between patients, and thus their sense of limitation may also vary: for example, a young gymnastic instructor commencing rehabilitation following injury may score highly and yet have what the general population would consider good function for his stage. It is just that he is limited in his daily activities. By comparison, an older retired individual would rarely perform heavy physical work, jumping or running, and so is not actually particularly limited and therefore has a low score. For this reason, since the ATRS is a subjective functional score, we recommend it to be used in conjunction with objective functional measures (such as strength and range of motion) when evaluating the outcome following an Achilles tendon rupture.

As a tool for outcome, the ATRS is valuable in patient follow-up. The score allows the physician an immediate representation of the patient’s function at a glance. This can be used to re-assure patients at 3-month follow-up visit.

We were uncertain how much running, jumping and heavy physical work our patients actually undertook, but the patients’ perception of these activities is unlikely to change for each patient. Some sub-scores must be interpreted with caution, particularly those related to running and jumping. Thus, we recommend cautious interpretation of the score according to individual rehabilitation protocols.

Conclusion

The ATRS was adapted to English and shown to be a reliable, valid and responsive method of testing functional outcome of patients following Achilles tendon rupture.

Copyright information

© Springer-Verlag 2012