Standardised test protocol (Constant Score) for evaluation of functionality in patients with shoulder disorders

As treatment modalities for shoulder diseases evolve and new treatments are introduced, the need for assessment of functional outcome is increasing. The Constant Score (CS) [1] was among the first shoulder score systems developed and is considered the most commonly used scoring system for evaluation of various disorders of the shoulder [2]. The CS assesses subjective and objective shoulder function with respect to: A: pain; B: activities of daily living; C: range of motion and D: strength (Figure 1). The CS is often used to evaluate treatment progress and to compare results of clinical trials for several specific shoulder disorders [2-4]. The European Society for Surgery of the Shoulder and Elbow (ESSSE) and the Journal of Shoulder and Elbow Surgery recommend the CS for use in research on shoulder disorders [2, 5]. Nevertheless, it has been criticised for relying on imprecise terminology and for diffuse definitions of methodlogy in the original description of the score [2, 6, 7]. Thus, a standardised test protocol is not available. This has left room for different interpretations and use of the score between centres [7]. To address these weaknesses, Constant et al published a guideline paper in 2008 with recommendations for modifications of the original CS [2]. However, no standardised test protocol was included in these new recommendations and the validity and usefulness of the score therefore remains questionable. The need for a standardised test protocol was further emphasised in a 2010 systematic review of the psychometric properties of the CS [8]. Moreover, the CS revised in accordance with the newly modified guideline has not been translated and adapted into Danish.

The aims of the study were: 1) to develop an easy-to-use test protocol in English for the CS with a detailed accompanying methodology guideline and 2) to translate and cross-culturally adapt this new test protocol of the CS into Danish.

MATERIAL AND METHODS

An English test protocol of the CS that included all sub-elements of the score according to the original [1] and the modified guideline [2] were developed at two independent centres in Denmark (Figure 2 and Figure 3). Consensus was then achieved. The test protocol was kept simple, and a detailed methodology guideline was made as a supplement to the protocol.

Next, the CS was translated into Danish at both ​​​​​centres using the agreed English version (see Figure 4 for the Danish version). The translation at each centre followed the principles of the ISPOR Task Force for translation and cultural adaptation [9].

The translation process involved the following  ​seven steps:

1. Independent forward translation by three bilingual health professionals at each centre and one professional translator. The persons involved all had Danish as their native language and some of them had prior experience with translation of outcome measures.

2. Consensus on a preliminary forward translation was achieved.

3. Cognitive pretesting to determine the acceptability and comprehensibility of the translation was performed at each centre. Six patients (two men and four women, mean age 47 years) diagnosed with shoulder rotator cuff/impingement syndrome and undergoing subacromial decompression and two physiotherapists participated.

4. Evaluation and revision of the translation were performed based on the result of the cognitive debriefing from the two centres and feedback from the health professionals involved.

5. An independent back translation was performed at each centre by a bilingual health professional with English as a native language. The back translators were not familiar with the original modified guidelines. The back translations were harmonised in agreement between the two centres.

6. The back translation was forwarded to one of the original authors of the modified guideline (Dr. Jens Ole Søjbjerg) for review and verification. The reviewer was Danish and also reviewed the Danish version.

7. The main translators from each centre and the project manager compared the final back translation with the original guideline. Final adjustments were incorporated and consensus on the Danish and English versions of CS was achieved.

Trial registration: not relevant.

RESULTS

We encountered several difficulties in defining the properties of some of the sub-elements of the score when developing the standardised test protocol for CS. As an example, the modified guideline recommends using a visual analogue scale with a sliding curser to assess the subjective sub-elements of pain and activities of daily living [2]. This visual analogue scale was constructed specifically for the guideline paper, where it is depicted, but is not commercialised and therefore not available for use in clinical practice. Instead, we chose to include a 15-centimetre “paper” visual scale both for pain and for activities (Figure 2). This procedure can be directly implemented in all settings as a usual ruler is used for calculating points. Also the modified Constant et al guideline introduced a time-period (last 24-hours) to assess pain, but did not introduce a time-period for the assessment of activities of daily living. We therefore added a time-period (within the past week) to assist patients in their evaluation of their activities of daily living (Figure 2). The Constant guideline also recommends that some sub-elements of the objective range of motion (ROM) part are assessed with the subject seated to avoid spinal tilting, whereas there is no recommended positioning for other sub-elements of the ROM testing. As item C4 of the ROM testing is performed with the hand/arm behind the back, it is considered appropriate to perform this part standing. We therefore chose to deviate from the guideline and recommend the standing position for all ROM tests, as it reflects normal clinical assessment, and as we believe that it will reduce the risk of different handling during testing (Figure 3).

We found only minor inconsistencies between the two forward translations into Danish, and consensus was achieved in all cases without much difficulty (see Figure 4). Due to cultural and linguistic differences between the English and Danish languages, a few questions and words were rephrased to facilitate adaptation into Danish. Most of the linguistic differences were found in the subjective parts of the score.

The testing of the preliminary version on patients revealed a few interpretation difficulties with some of the “subjective” questions. The more difficult problem had to do with the interpretation of “… your normal work …” (Item B2) in patients who were not employed (Figure 2). We rephrased the sentence to “… your normal daily work …” thereby implying that “daily work” does not have to be related to employment. A few unemployed patients who participated in the validation found this rewording relevant. The physiotherapists had minor difficulties interpreting the score system, and some methodology instructions were rephrased. Overall, both patients and physiotherapists found the test protocol relevant and easy to complete.

Dr. Jens Ole Søjbjerg approved the Danish as well as the back-translated English test protocol.

DISCUSSION

The CS, which was originally developed as a scoring system for the evaluation of functional outcome in patients with general shoulder disorders, is widely used internationally and recommended by the ESSSE, but at the same time often criticised for relying on imprecise terminology and for lack of a standardised, internationally accepted test protocol.

Based on the newly modified guideline for the CS [2], the present study provides clinicians with a standardised English test protocol in addition to a version for use in Denmark. A previous reliability review study of the CS [7] revealed considerable differences in seven test protocols of the CS used by various European shoulder units, and none of the protocols were sufficiently standardised. Notably, in a study published in 2012, Blonna et al [10] demonstrated that both inter- and intraobserver reliability improved significantly by using a standardised CS protocol, and the importance of standardisation has been further emphasized in other studies [2, 6]. The paper by Blonna et al did actually provide a standardised test protocol, but unfortunately they did not follow the new and modified guidelines provided by Constant et al in 2008, although reference was made to the study in the paper.

Thus, we are unaware of any official standardised test protocol for the CS including a version that follows the new recommendations [2]. To our knowledge, the test protocol presented in this paper (Figure 2 and Figure 3) is therefore the first of its kind that follows the new recommendations, and we recommend that it be used internationally as a standardised version as it includes precise definitions and descriptions of the scoring system as well as the method of assessment.

Still, we did identify several potential limits of the CS while developing the standardised test protocol. First of all, full standardisation of all items seems impossible, e.g. the standardisation of “painless” movements required when testing ROM and strength (Figure 3) seems difficult as pain differs in terms of type and location, and as some patients are “painless” owing to analgesics. Our test protocol does not solve this problem, but we emphasize that the interpretation of such ambiguous terms in the target population should clearly be reported when results are presented in future studies. Secondly, there are differences in the systems used for allocation of points to sub-elements of the same score item, e.g. the visual scales used to assess the subjective elements of pain and activities related to daily living (Figure 2). Thus, the scale used to assess pain (item A) uses 0 and 15 as anchor points, as depictured in the guideline [2]. This means that the “length” on the scale which corresponds to a score of 0 and 15 points (from 0 to 0.49 and from 14.5 to 15), respectively, is but half the length allocated to the remaining points (1-14) on the scale. In comparison, the visual scale used to assess activities of daily living (items B2 & B3) ranges from only 0 to 4 points and has no anchor points, making the five scoring possibilities (0-4 points) equally represented by the same “length” (from point to point) on the scale. Also, points are given by movements successfully performed, ​ whereas others are given as the movement progresses for different items of the ROM. We did not change these inconsistencies in the original guidelines for allocation of these points, but we tried to make clear recommendations in our test protocol and the accompanying methodology guideline.

The translation of the modified guideline for the CS [2] into Danish, which followed international recommendations [9], was successfully conducted. The process revealed only minor discrepancies concerning the wording and understanding of the CS, and pretesting revealed that both patients and health professionals found the translation acceptable and relevant and the test easy to complete. Still, the patients used for cognitive debriefing represent a small and specific group of all shoulder disorders and therefore do not fully match the target population, as the CS targets shoulder disorders in general.

Some potential limitations of the CS were acknowledged in the conclusion of the paper on the modified guideline [2], but at the same time it was pointed out that the CS is widely accepted and should therefore continue to be used until a more valid and reliable alternative has been developed. Thus, the CS is still recommended for assessment of shoulder disorders in general despite the obvious fact that the wide range of shoulder disorders represents slightly different problems. Whether the CS is valid for all shoulder disorders is therefore questionable as some items might be irrelevant and there might be ceiling effect for some subgroups. We therefore recommend that our standardised test protocol be validated on all subgroups of shoulder disorders before being used clinically.

CONCLUSION

A simple standardised English test protocol of the modified Constant Score was developed. With precise terminology and definitions, this test protocol is the first of its kind; and we suggest it be used internationally for standardised assessment of the CS. A translation of the CS into Danish was successfully conducted. We suggest this Danish version of the modified CS for use in Danish patients with shoulder diseases, as it is – to our knowledge – the first translation into Danish that follows international guidelines. However, testing of validity, reliability and responsiveness of these new English and Danish versions needs to be done before there use may be recommended for daily clinical use.

Correspondence: Ilija Ban, Ortopædkirurgisk Afdeling, Hvidovre Hospital, 2750 Hvidovre, Denmark. E-mail: ilija.ban@gmail.com

Accepted: 13 February 2013

Conflicts of interest: None. Disclosure forms provided by the authors are available with the full text of this article at www.danmedj.dk

Acknowledgement: We would like to thank Derek Curtis, Morten Pilegaard and David John Glasscock for their assistance with forward and back translations.