Original Article
Definition of the construct to be measured is a prerequisite for the assessment of validity. The Neck Disability Index as an example

https://doi.org/10.1016/j.jclinepi.2013.02.005Get rights and content

Abstract

Objective

To determine the content, structural, and construct validity of the Dutch version of the Neck Disability Index (NDI).

Study Design and Setting

To assess content validity, 11 neck pain experts and 10 patients commented on the construct, comprehensiveness, and relevance of the NDI. Structural validity was assessed by item factor analysis (FA) and item response theory modeling using the generalized partial credit model. Differential item functioning (DIF) analysis for gender was examined. Pearson correlation coefficient with the Disabilities of the Arm, Shoulder and Hand (DASH) questionnaire was calculated to assess construct validity.

Results

In addition to a suboptimal translation, we found a lack of consensus on the construct the NDI intends to measure. Experts and patients suggested that the NDI measures more than physical functioning. Unidimensionality of the NDI could not be confirmed. DIF analysis for gender showed DIF for the headache item. The goodness-of-fit statistics for FA with one factor were satisfactory when the item “concentration” was omitted. A correlation of 0.75 with the DASH was found supporting construct validity.

Conclusion

It is questionable whether in research the NDI should be the instrument of choice for use as a primary outcome measure. Definition of the construct to be measured is a prerequisite for the assessment of validity.

Introduction

A recent systematic review on the measurement properties of disease-specific questionnaires in patients with neck pain indicated that the Neck Disability Index (NDI) is the most frequently evaluated questionnaire [1]. Researchers and clinicians often use the NDI to measure the level of disability and to study the effect of an intervention on patients with neck pain [2], [3], [4], [5]. In developing the NDI, Vernon and Mior were inspired by Fairbank and Pynsent [6] who developed the Oswestry Disability Index (ODI) for measuring disability in patients with low back pain. The NDI was the first instrument designed to assess self-rated disability in patients with neck pain. Currently, there are more than 450 articles internationally that have cited the NDI. Numerous clinical guideline organizations, especially for whiplash management, have endorsed the NDI as the questionnaire of choice for patients with neck pain.

Vernon declared in an e-mail conversation (November 2011) that the construct of the NDI was “self-rated disability,” where disability was understood as the perceived effect of pain and impairment on the patient's performance and enjoyment of activities of daily living.

However, different opinions exist with regard to the meaning of the construct that the NDI aims to measure. Some researchers interpret the NDI as a measure of functional status [7], [8], whereas others have a broader interpretation and see it as a measure of pain and disability [9]. As a result, confusion might arise as to what the NDI aims to measure and how scores should be interpreted.

The lack of consensus in the construct that the NDI aims to measure might hinder an assessment of the validity of the instrument as validity is defined as the extent to which an instrument measures what it purports to measure [10].

The aim of this article was to evaluate the quality of the translation of the most commonly used Dutch version of the NDI (NDI-DV) and to determine its content, structural, and construct validity.

Section snippets

The NDI

The NDI is a patient-reported outcome (PRO) measure. It consists of 10 items. The 10 items have six response categories (range, 0–5; total score range, 0–50) [11] (Appendix A; available on the journal's website at www.jclinepi.com).

Translation

To check the Dutch translation of the NDI, the most commonly used NDI-DV [12] was translated back into English by two independently operating professional translators, blinded to the original English version. Discrepancies between the translated version, the versions

Quality of the Dutch translation

A number of problems were identified with the Dutch translation of the NDI. First, the patient instruction that precedes the original version of the NDI is not included in the NDI-DV. Second, for some items, the Dutch translation (or part of it) was considered inadequate. For example, the Dutch translation of the phrase “… but it causes extra neck pain” in the item “personal care” was back translated into English as “although this increases the pain.” Saying that personal care causes extra neck

Discussion

The most commonly used Dutch translation of the NDI was considered to be suboptimal. Poor translation implies that scores on the Dutch and the English versions of the NDI might not be comparable. It may also affect content validity if translated items are interpreted differently.

Many researchers interpret the questionnaire as a measure of physical functioning [7], [8]. However, our study shows that the NDI measures more than physical functioning. Review of the literature and personal

Conclusion

It is unclear what the NDI aims to measure. It measures more than physical functioning. However, for a broader construct, important items are missing. The content validity of the NDI is poor. An important lesson from this study is that a clear definition of the construct to be measured is a prerequisite for the assessment of validity. In line with current recommendations, for example, from the US Food and Drug Administration [33], we conclude that more attention should be given to the

References (33)

  • G. Bronfort et al.

    A randomized clinical trial of exercise and spinal manipulation for patients with chronic neck pain

    Spine

    (2001)
  • J.C. Fairbank et al.

    The Oswestry Disability Index

    Spine

    (2000)
  • D.L. Riddle et al.

    Use of generic versus region-specific functional status measures on patients with cervical spine disorders

    Phys Ther

    (1998)
  • J.J. Pool et al.

    Is a behavioral graded activity program more effective than manual therapy in patients with subacute neck pain? Results from a randomized clinical trial

    Spine

    (2010)
  • L.J. Cronbach et al.

    Construct validity in psychological tests

    Psychol Bull

    (1955)
  • H. Vernon et al.

    The Neck Disability Index: a study of reliability and validity

    J Manipulative Physiol Ther

    (1991)
  • Cited by (39)

    • Development and internal validation of prognostic models for recovery in patients with non-specific neck pain presenting in primary care

      2021, Physiotherapy (United Kingdom)
      Citation Excerpt :

      Is seems plausible that short-term prediction is more accurate compared to long-term prediction. The post-treatment disability model performed best, possibly because the outcome was measured with the NDI, which is an instrument that covers various health constructs [52]. The NRS is a single-item questionnaire which measures a narrower domain and may also have larger measurement error that can influence the performance of the models [53].

    • Excursion of the median nerve during a contra-lateral cervical lateral glide movement in people with and without cervical radiculopathy

      2021, Musculoskeletal Science and Practice
      Citation Excerpt :

      The NDI has long been advocated as a premier patient reported outcome measure (PROM) to establish the level of perceived disability in patients with neck pain (MacDermid et al., 2009). Recently however, it has been questioned whether in clinical research the NDI should be the instrument of choice for use as a primary outcome measure (Ailliet et al., 2013). As patients with CR generally suffer more from arm and periscapular pain than neck pain (Kuijper et al., 2009a; (Thoomes et al., 2012); Wainner et al., 2003a), the NDI might not be the most suitable PROM for patients with CR (Ailliet et al., 2013; Gartner et al., 2019).

    • Evaluation of Studies on the Measurement Properties of Self-Reported Instruments

      2020, Asian Nursing Research
      Citation Excerpt :

      In this study, conceptual definitions were clearly reported for only two [26,29] of the five newly developed instruments (Tables 1 and 2). The systematic review of a disability index by Ailliet et al. [35] revealed that the instrument was unclear regarding what it aims to measure and recommended developing a new neck-specific instrument with a clear definition. The structural validity, consistency, and cross-cultural validity/measurement invariance were measured to determine the internal structure of each self-reported instrument.

    • A user's survey of the clinical application and content validity of the DASH (Disabilities of the Arm, Shoulder and Hand) outcome measure

      2017, Journal of Hand Therapy
      Citation Excerpt :

      In most instances (70%-82%), the DASH and QuickDASH were being used across the entire UE and in the neck 10% of the time. This finding corresponds with the literature where measurement testing28,53-58 and application59-62 of the DASH and QuickDASH in people with neck disorders is evident. The measures were being used for a broad range of UE disorders, including MSK and arthritic disorders, for which it was intended.

    • Responsiveness and Minimal Important Change of the Pain Self-Efficacy Questionnaire and Short Forms in Patients with Chronic Low Back Pain

      2016, Journal of Pain
      Citation Excerpt :

      Content validity refers to the extent to which the most relevant and comprehensive aspects of a construct are adequately reflected in the content of a measurement instrument.39,46 The assessment of this measurement property for the PSEQ could be accomplished by conducting cognitive interviews in patients with chronic pain5 or by asking experts to judge the content of the items.2 The MIC value equal to 5.5 for the PSEQ found in this study differs from the MIC of 9 points found in a previous study.41

    View all citing articles on Scopus

    The authors declare that they have no competing interest. Luc Ailliet, the corresponding author, is as a PhD student financially supported by The European Chiropractic Union, the Belgian Chiropractic Union, The Netherlands Chiropractic Association (NCA), and l’Institut Franco-Européen de Chiropratique (IFEC, Paris, France). The views expressed in this manuscript are those of the authors, and no official endorsements by supporting agencies is intended or should be inferred.

    View full text