NDI Global Health CoursesWe invite you to join the NDI family by participating in a 9-day Global Health Course in the longest running sustainable naturopathic clinic in Latin America. NDI courses offer a unique field experience by combining a curriculum of study in global health that examines worldwide health issues with clinical immersion in an interprofessional, community empowered, team care model. NDI courses fill up quickly so apply well in advance.
The Neck Disability Index (NDI) was developed in 1989 by Howard Vernon. The Index was developed as a modification of the Oswestry Low Back Pain Disability Index with the permission of the original author (J. The objectives of this study were to translate and culturally adapt the Spanish version of the Neck Disability Index Questionnaire (NDI), and the Core Outcome Measure (COM), to validate its use in Spanish speaking patients with non-specific neck pain (NP), and to compare their psychometric characteristics with those of the Spanish version of the Northwick Pain Questionnaire (NPQ).
To apply, go to our! 2019 NDI Global Health Courses CIVIL UNREST CONTINUES IN NICARAGUA HOWEVER NDI HAS DECIDED TO RESUME GLOBAL HEALTH COURSES/BRIGADES. PLEASE CONSULT YOUR DEPARTMENT OF STATE TO EVALUATE THE RISK OF TRAVEL AND KNOW THAT YOU TRAVEL AT YOUR OWN RISK. #66 AUGUST 16th - 24th, Lead Doctor: Dr. Tabatha ParkerAll courses are 9 days total (including 2 travel days).
Most groups arrive on Saturday (Day 1) and fly out the following Sunday (Day 9). Spanish is not required, nor is being in a healthcare professional or student. More About NDI CoursesLocated on the beautiful UNESCO Biosphere Reserve of Ometepe Island, this unique course gives health professionals an opportunity to work in an integrative medicine clinic officially within a Ministry of Health system. Naturopathic physicians, integrative MDs, licensed acupuncturists, chiropractors, herbalists, medical students, allied health care professionals, or anyone interested in global health that wishes to volunteer is encouraged to attend one of our 9-day international global service learning trips. Course participants work with naturopathic doctors, local farmers, and collaborate on NDI community outreach projects with the opportunity to immerse themselves in the island culture of Ometepe. Courses are held 4-6 times per year.What Can I Expect?What is the “ Natural Medicine in Global Health” course like? The NDI program offers participants a one-of-a-kind service learning opportunity to experience the unique combination of Nicaragua, naturopathic medicine, and teaching methods developed by Paulo Freire.
The course combines classes in global economics, naturopathic medicine, global health, combining popular education with medical service learning. Health care practitioners get the opportunity to serve in a tropical medicine setting while while students receive an opportunity to precept. Non medical participants help in community projects and other clinic services. How We TeachThree forms of learning are utilized in our courses: experiential, observation in a medical setting, and didactic.
The mixture of preceptorship and community project work allow participants to observe patients and work side-by-side with farmers. Living with locals in their homes give participants a glimpse into the reality of life in a developing country like few experiences offer. Students receive hands-on training working with naturopathic and local doctors and attend classes on global health, economics, and history. How We ProcessIt is very important to us that our students not only have an incredible experience, but are given tools to process the experience. NDI group processing sessions allow participants to delve more deeply into issues and feelings that arise in global health situations. These sessions allow participants to more fully integrate their experience when they return home. Daily group check-in, reflection and insights provide additional learning opportunities for participants.
Check-ins create a space for participants to share feelings. Reflection sessions raise and discuss difficult questions. Insights focus on sharing sacred/creative/contemplative space where participants can share a moment of silence, prayer, thoughtful contemplation, poem or song.
All three exercises create space for participants to integrate and evaluate their physical, mental, emotional and spiritual experience.
Horizon estimation suggested the potential for 1 missed paper. The agreement between raters on quality assessments was high (kappa = 0.82). Half of the studies reached a quality level greater than 70%.
Failures to report clear psychometric objectives/hypotheses or to rationalize the sample size were the most common design flaws. Studies often focused on less clinically applicable properties, like construct validity or group reliability, than transferable data, like known group differences or absolute reliability (standard error of measurement SEM or minimum detectable change MDC). Most studies suggest that the NDI has acceptable reliability, although intraclass correlation coefficients (ICCs) range from 0.50 to 0.98. Longer test intervals and the definition of stable can influence reliability estimates.
A number of high-quality published (Korean, Dutch, Spanish, French, Brazilian Portuguese) and commercially supported translations are available. The NDI is considered a 1-dimensional measure that can be interpreted as an interval scale. Some studies question these assumptions.
The MDC is around 5/50 for uncomplicated neck pain and up to 10/50 for cervical radiculopathy. The reported clinically important difference (CID) is inconsistent across different studies ranging from 5/50 to 19/50. The NDI is strongly correlated (0.70) to a number of similar indices and moderately related to both physical and mental aspects of general health.
It is estimated that a third of all adults will experience neck pain during the course of 1 year, and 70% is the approximate lifetime prevalence., About 19% of the population may suffer from chronic neck pain at any given time, creating a substantial societal burden. Monitoring outcomes is a key component of monitoring the effects of evidence-based care, in justifying services, and in program evaluation. However, the episodic, fluctuating nature of neck pain and the lack of clear and consistent physiological findings complicate this process. A fundamental component of monitoring outcomes is having reliable and valid tools with known measurement properties.
In the case of neck disorders, self-reported pain and disability are usually the primary focus.Isolated studies on outcomes measures provide a context and method-specific view of the measure's ability to provide useful clinical information. It is only by synthesizing information from multiple studies that we can understand how a measurement tool performs across different contexts and applications.
The synthesis of larger pools of data is a mechanism to provide more stable estimates of measurement errors and benchmarks for change/outcomes. In fact, 2 systematic reviews have been conducted that address self-report measures for neck pain. Both compared different outcome measures using a semistructured process and concluded that the Neck Disability Index (NDI) was the most commonly used self-report instrument for evaluating status in neck pain clinical research., Both reviews alluded to the psychometric and clinical properties of the tools, but neither attempted to deal with specific properties or to synthesize this knowledge. The developer of the NDI published a summary paper in 2008 summarizing a 17-year history with the NDI. The author stated that “The Neck Disability Index has been used in over 300 publications, translated into 22 languagesand endorsed for use by a number of clinical practice guideline committeesmaking it the most widely used and most strongly validated instrument for assessing suffering disability in patients with neck pain.”While previous reviews have claimed to be systematic, none have included a key element of systematic review, critical appraisal of the quality of individual studies. This may reflect inherent difficulties in performing the critical appraisal due to lack of instrumentation. The lead author of this current review has published a scale and interpretation guide, for this purpose.
Psychometric studies are important to establish measurement properties like the relative difficulty of items, appropriate grouping of items into subscales, reliability, validity, and responsiveness. In addition to psychometric properties, clinicians are concerned with issues on feasibility, floor/ ceiling effects, availability of different language/cultural adaptations, and administration burden for themselves and their patients. Terms like “clinician/patient friendliness,” or “clinical utility/applicability,” or, as we prefer, “usefulness” have been used in reference to these practical considerations. When therapists try to integrate the NDI into their clinical practice, they are concerned with the psychometric issues but also need information on usefulness.
The purpose of this study was to conduct a systematic review that would summarize the quality and content of current research regarding the psychometric properties and usefulness of the NDI. Development of the NDIVernon and Mlor developed the NDI using the Oswestry Low Back Pain Index (OLBPI) as a template for identifying items and a scoring metric. They initially selected 6 items from the original scale: pain intensity, personal care, lifting, sleep, driving, and sex life, and submitted this to a consulting team. The consulting team added 4 items: headache, concentration, reading, and work, resulting in the 10-item scale. These 10 items were modified for clarity and relevance based on feedback from 5 individuals with a history of whiplash injury and the review team. In the end, the NDI contained 5 items from the OLBPI, 2 of which were modified, and 5 new items. The questions are measured on a 6-point scale from 0 (no disability) to 5 (full disability).
The numeric response for each item is summed for a score varying from 0 to 50. Some evaluators choose to provide a score out of 100%, particularly as a strategy to deal with questions left unanswered ( APPENDIX A). Literature Search and Study IdentificationA database search was performed using Medline, PsychInfo, CINAHL, and Embase, which included papers written between January 1966 and September 2008 ( ). The following keywords were used to search all databases for eligible studies: (NDI OR Neck Disability Index) AND (reliability OR validity OR responsiveness OR clinically important difference OR MCID OR Rasch OR factor analysis OR translation OR validation). In addition, we conducted Google searches and hand searches of retrieved study references lists. The number of studies retrieved and the net results of abstract/title and full review are noted in the. We conducted a Horizon estimation to evaluate the potential that any articles were missed and the efficiency of our search using a procedure developed by Foster and Goldsmith for SAS software (full details available from author Charlie Goldsmith).
The first stage of study identification was of title/abstracts, which were independently reviewed by at least 2 of the authors. An article was accepted if it met the following inclusion criteria: reported on at least 1 psychometric property of the NDI in patients with neck pain and was written in French or English (2 with English abstracts and non-English full text 1 Dutch, 1 Spanish were included for abstract data extraction but not full critical appraisal). In total, 37 primary studies, 3 structured reviews, and 1 unpublished in press manuscript proceeded to a full review (, ).Evaluation criteria: 1. Thorough literature review to define the research question; 2. Specific inclusion/exclusion criteria; 3.
Specific hypotheses; 4. Appropriate scope of psychometric properties; 5. Sample size; 6. Follow-up; 7. The authors referenced specific procedures for administration, scoring, and interpretation of procedures; 8. Measurement techniques were standardized; 9. Data were presented for each hypothesis; 10.
Appropriate statistics-point estimates; 11. Appropriate statistical error estimates; 12. Valid conclusions and clinical recommendations. †Papers where NDI evaluations were performed while evaluating a different measure as the primary purpose. Quality scores were rated in content for the NDI-related methods. Three structured reviews underwent data extraction but no critical appraisal., Two papers had data extraction from the English abstract and study tables but no critical appraisal, because they were in non-English text., One paper had item comparison but no NDI data, so no full critical appraisal.Each paper's score was converted into a percentage because 1 item was based on follow-up and some psychometric studies are cross-sectional, leaving unequal denominators for different studies.
We rank ordered studies on quality and considered this ranking when making conclusions and recommendations, although there was no formal mechanism to weight conclusions, based on the quality of the associated source document. The horizon analysis indicated a possibility that our search strategy missed 1 article (95% confidence interval, 0–2; 97% yield). An optimal search strategy would have been Embase first, then CINAHL, and then Medline. PsychInfo did not contribute any information after these 3 databases.In total, 41 studies were identified that addressed at least 1 psychometric property of the NDI.
Two neck disability instrument reviews were identified, although neither included formal critical appraisal., Similarly, a recent comprehensive review of the NDI by the developer did not include formal critical appraisal of individual studies. The primary studies crossed different populations, interventions, and time intervals, and addressed different psychometric properties.
Quality of the individual studies was variable, ranging from 21% to 96%, with 37% of papers reaching or exceeding a score of 75% on the quality rating. The most common flaws observed in the psychometric articles were (1) not reporting specific psychometric hypothesis/objectives, (2) inadequate sample size calculations/justification, and (3) absence of error estimates such as confidence intervals or standard error of measurement (SEM). A descriptive synthesis of the findings for psychometric properties across all identified studies is summarized in through. Due to the heterogeneity of study populations and properties evaluated, no meta-analyses were performed. Most studies addressed a spectrum of psychometric properties, but few were comprehensive. Few studies provided specific conclusions or numbers that could readily be integrated into clinical practice but, rather, tended to rely on generalities when making conclusions. The type of data collected during NDI validation studies was typically comprised of less clinically useful data, like correlations indicating construct convergent validity, versus more useful information, like known group differences that could be used as comparative data for clinical comparisons.
Similarly, group reliability, such as intraclass correlation coefficients (ICCs), was reported more often than more useful indicators of absolute measurement error (like SEMs, mean retest differences, or minimal detectable change MDC). Summary of Previous Structured ReviewsThree papers were identified where the authors performed a comprehensive structured review with use of a formal search for articles, but without use of multiple raters or formal critical appraisal. The first review is important because it established the relative strengths and weaknesses of the NDI when compared to other scales.
This review was based on a formal search of 2 databases, MEDLINE and CINAHL, citation tracking using the citation index, hand searching of relevant journals, and correspondence with experts to find additional papers (up to year 2000). No formal quality evaluation was performed.
Five standardized neck pain scales were identified and compared qualitatively in terms of content and measurement properties. These authors concluded that 3 of the scales were similar in terms of structure and psychometric properties: the NDI, the Copenhagen Neck Functional Disability Scale, and the North-wick Park Scale. It was suggested that the NDI had the strength of being most studied amongst these 3 and was at that time the only instrument revalidated in different study populations. The authors of this review suggested other scales had important limitations. These included that the Neck Pain and Disability Scale must be read to the patient and the Patient-Specific Functional Scale was considered “very sensitive to functional changes in individual patients, but comparisons between patients are virtually impossible.”The results of a subsequent systematic review of studies on outcome measures for the cervical spine published up until 2004 suggested that instrument development is ongoing as it identified an increased number of neck pain scales (a total of 11 scales). Again, while databases were used to identify papers, other elements of systematic review were not performed, including use of multiple raters/data extractors, quality ratings for individual studies, or a formal process to synthesize results. Eight English-language (specific-to-neck) scales and 3 (nonspecific validated for the neck) scales were reviewed.
The instruments (and year published) are listed in. The psychometric properties of each scale were reviewed in a qualitative manner, while the content of instruments was compared more quantitatively in a summary table. This item analysis indicated that the most common content items on neck disability scales were self-care/activities of daily living (ADL), work, standing/running, pain intensity, and driving. Summary of Primary StudiesReadability/Language and Cultural Translation A number of papers have addressed issues around readability, usually in the context of language and cultural translations. Wlodyka-Demaille et al reported that the concepts of “leisure” and “social activities” had different meanings in French and American cultures. For that reason, they provided examples in these areas to make the questionnaire more understandable for this population. Through a series of translating and back translating, Lee et al concluded that their translation process retained the sound measurement properties of the original English version in the Korean version.
Cook et al ensured good readability of the Brazilian/ Portuguese version of the NDI through a committee who reviewed the translator reports. They reached consensus on discrepancies in 4 areas, including semantic equivalence, experiential equivalence, idiomatic equivalence, and conceptual equivalence. In the Swedish version, Ackelman and Lindgren changed each item of the tool to specify that only disability due to neck pain was of interest. In the Spanish version, 16% of patients had comprehension difficulties, but these did not appear to be related to educational level and reliability was excellent. Others identified that the driving item had a high rate of nonresponse in an Iranian population.
Overall readability in the English and all translated versions is deemed to be acceptable, although some patients require support to understand the items, and nonresponse is more common for task items like driving and, to a lesser extent, reading.Not all translations have been published in peer-reviewed literature. The developer reported that he worked with an independent organization to produce additional translations through standardized translation methodology (although psychometric testing was not performed). These translations included English (Australian/UK/US), Danish, Finnish, French (Canadian/Switzerland/ Germany), Italian, Norwegian, Polish, Spanish (Spain/US).Administration Burden Few authors specifically address or state how they measured the time taken to complete the NDI. However, all agree that only a short amount of time is required. Stratford et al reported that the time for patients to complete the NDI was about 3 minutes, while Wlodyka-Demaille et al reported that it took a mean (SD) of 7.4 (6.8) minutes (range, 1–60 minutes).
There was only a single report of therapist burden stating that it takes 8–10 minutes to complete and 5 minutes to analyze the (Dutch) NDI. Stratford et al commented that the questionnaire could be completed in the waiting room and thus did not add any additional time to the patient's visit. Interpretability/Subgrouping of Differential OutcomesNDI scores vary from 0 to 50, where 0 is considered “no activity limitation” and 50 is considered “complete disability.” Originally, no explicit recommendations were made by the developers on the handling of missing items, or minimum number of items required for validity. More recently, it has been suggested that if 3 or more items are missing, the score may not be valid.
The majority of authors report the NDI out of a total of 50; however, some authors, provide a percentage score as a strategy to account for questions that are left unanswered., In the study by Ackelman and Lindgren, if more than 2 items were left unscored, the subjects were not included.Vernon and Mior offered the following interpretation of NDI scores: 0 to 4, no disability; 5 to 14, mild disability; 15 to 24, moderate disability; 25 to 34, severe disability; and greater than 35, complete disability. However, no process was described for how these ratings were derived and no validation of these categories was performed. One group of authors set the “normal limit” of the NDI between 0 and 20 points. Again, the methodology behind setting this benchmark was not described, and this cutoff has not been validated. Sterling et al used data from clinical studies to define patients who had recovered as having NDI scores of less than 4 (8%), those with mild disability as having scores of 5 to 14 (10%–28%), and those with moderate to severe disability as having scores of greater than 15 (30%). Conversely, Nederland32 found that patients defined as recovered at 24 weeks had a score of 14 or less, and subsequently defined a score of less than 15 as a recovery cutoff. In the same study, patients defined as having persistent pain had a score of 28 or greater.
Miettinen used a recovery cutoff of 20/50 at 3 years.Floor-Ceiling Effect Floor and ceiling effects have practical clinical relevance, as they represent patients for whom pain and disability estimates may be invalid and for whom changes may not be measurable. Few studies have specifically addressed this issue, and none have specifically analyzed how the MDC varies over the spectrum of possible NDI scores.
TABLE 4 Summary of Validity Properties ValidityData ExtractedContent (including analyses of question, appropriateness, missing items, ceiling/floor effects).Comparison of items across NDI, NPQ, and the Copenhagen Neck Functional Disability Index; items on all 3: sleeping and reading; items on 2 scales: pain, personal care, lifting, concentration, work, driving, recreation.Authors used a patient interview to elicit problems called the problem elicitation technique to identify functional problems in patients with chronic nontraumatic neck pain. It was determined that ≥75% of patients identify problems with sleep, mobility, role activity, emotion, and symptoms.
Sleep disturbance had the highest prevalence. More than half of subjects identified difficulties with frustration, driving, and lifting. Less common but mentioned by more than 30% of patients were looking into cupboards, gardening, headaches, housework, working overhead, and general exercise. The highest mean severity scores were found in depression, cooking, and sitting upright. Individual problems ranked most important by subjects were driving, sleep disturbance, and frustration. Abbreviations: AUC, operating characteristic curve; CID, clinically important difference; COM, core outcome measure; CWOM, core whiplash outcome measure; ES, effect size; GPE, global perceived effect; GRC, global rating of change; NDI, neck disability index; NPDI, Neck Pain Disability Index; NPQ, Northwick Park Neck Pain Questionnaire; PSFS, Pain-Specific Functional Scale; ROC, receiver operating characteristic; Rx, treatment; SF-36 physical, Short-Form 36 physical component scale; SRM, standardized response mean; VAS, visual analog scale.
This study synthesized current research in 37 studies addressing the psychometric properties of the NDI and was able to provide some clinical recommendations regarding its use, within the limits prescribed by the available evidence. Overall, there is moderate to strong evidence for a spectrum of psychometric properties supporting use of the NDI in patients with acute or chronic neck pain with symptoms of musculoskeletal or neurogenic origin. The relative importance of different psychometric properties will vary according to purpose.
For example, when using the NDI to evaluate clinical change in individual patients, the absolute measurement error and responsiveness should be considered most relevant. Conversely, when using the NDI to differentiate different levels of disability, known group validity, a form of discriminative validation that tests differences between known subgroups, would be more important.There is adequate evidence that the NDI is stable over short test-retest time intervals (0–3 days), although it is less likely that patients defined as stable will have stable scores in longer test-retest intervals. In short-term test-retest intervals researchers often assume that most untreated patients remain relatively stable over 1 to 3 days. In longer test-retest intervals it is more necessary to establish the patient has remained stable. This is often based on a −3 to +3 score on global rating of change instrument. Thus, by definition, patients with small to moderate self-reported changes are defined as stable in this analysis.
Conversely, ICCs reported when the test-retest was based on a single occasion, with the NDI administered in different languages or different question order, have been exceptionally high (0.90). This suggests that lower test-retest reliability across longer test intervals may reflect changes in scores due to the episodic nature of neck pain, early recovery, and how “stable” is defined within a given time window.
Some studies have used as much as a 5-week interval in patients defined as stable to determine test-retest reliability. While it is important to understand short-term and long-term stability of clinical measurements, it should be appreciated that an increasing number of factors may affect scores as test-retest intervals are extended. Even if a patient's severity of pain or task difficulty remains consistent over the course of several weeks, the experience of pain and disability will be mediated by psychosocial influences and calibrated against multiple internal and external references. Thus, changes in factors like coping, self-efficacy, or social support may contribute to alterations in perceived disability over a 5-week period.
Recalibration of the disability experience in the absence of changes in pain intensity is a plausible explanation for lower estimates of reliability in studies with larger test-retest intervals. However, it is important to consider that these variables do contribute to measurement error in clinical studies. Therefore, reliability studies with longer test-retest intervals may be important to consider when designing clinical research studies and reliability studies, whereas shorter intervals may be appropriate for consideration when estimating error across shorter clinical re-evaluations. The reliability data currently published in the literature provide estimates that reflect different definitions of stable, in different clinical subgroups, with different test-retest intervals allowing users to select estimates most similar to their situation and use.Only a few studies presented more clinically relevant indicators of measurement error, the SEM or the MDC, both being expressed in unit of the original score and based on the reliability coefficient and the sample variability. The developer of the NDI suggests that MDC is 5 points. The highest estimate comes from Cleland et al, who reported an MDC of 10.2 points for their patient population with cervical radiculopathy, and the lowest was from Voss et al, who reported an MDC of 1.66 for their study, which included stable patients with recurrent neck pain. Despite a range of 2 to 10, the more common estimate for MDC is around 5/50 (10%).
The reasons for this large range of MDC values are largely unclear, although either larger standard deviations or lower reliability coefficients can increase the value of the MDC, so it is not surprising that the larger MDC estimates come from studies with lower ICCs. Across the different studies, MDC variations reflect the effects of test-retest interval and the definition of stability on reliability estimates. Despite variability, the value of 5/50 or 10% commonly used in clinical situations appears to be appropriate for most clinical comparisons that tend to occur over 2 weeks or less.While the NDI has been shown to be responsive, estimates are highly variable among studies, suggesting that a number of fully powered studies that vary on factors such as the type of intervention, length of follow-up, comorbidity, and the nature of the neck condition will be required to provide more precise estimates in different clinical circumstances. This review did not attempt to evaluate different instruments and, therefore, was not designed to comment on whether the NDI is the most responsive neck scale. However, none of the studies that performed head-to-head comparisons of different neck disability scales indicated that another instrument was superior to the NDI in terms of responsiveness.There was agreement across studies that the NDI is easy to read and understand, in both its original English format, as well as in subsequent translations.
In general, published translations used at least some of the recommended procedures for valid translation and demonstrated equivalence. Due to the brief nature of the questionnaire, the NDI has minimal administrative burden, although scoring variations can create potential for confusion.
Ceiling/floor effects have been suggested as a potential problem, although there is no consistent definition of what constitutes a ceiling or floor. If one assumes that the MDC is usually at 5 points (and up to 10 points), then scores of 40 or higher and 10 or lower might be considered problematic for detecting worsening or improvement, respectively.
Evidence suggests that smaller changes are relevant at these ends of the scale. We would also suggest that within these ranges, clinicians should consider supplementing the NDI with other instruments, like the Patient-Specific Functional Scale, which is able to sample items that are of high or low difficulty, thus, making the scale more amenable to change for patients with this type of clinical presentation.One limitation of our review stems from the lack of agreed upon quality criteria for synthesis process for psychometric studies. Neither previous systematic review incorporated critical appraisal. Although the first author of this review has addressed the critical appraisal of individual studies by developing a tool for this purpose, there is no clear method to synthesize the extracted psychometric evidence.
In some systematic reviews only high-quality studies are synthesized. However, when evaluating an outcome measure, it is important to see how the instrument performs across different contexts and purposes. Furthermore, there are no levels of evidence that create clear categories for study quality. Therefore, we rank ordered studies by quality to allow the reader and ourselves a mechanism to place greater emphasis on the findings from high-quality studies.
We summarized the information on psychometrics and usefulness by adapting and expanding a framework used by others. While we tried to make the process as objective as possible, there are inherently subjective elements, as study results must be taken within the context of the study population, interventions, and purpose, making it difficult to synthesize results from individual studies into global recommendations.A second limitation in our review is that the scope of our search retrieved full-text papers written in only English or French. We don't expect this limitation to have a substantial impact on our results, as the majority of translation and validation articles were printed in English, and we were able to extract data from English abstracts in non-English text.Overall, the NDI has a number of features that suggest that it has good clinical utility. These include its brevity and the fact that it has been translated into a number of languages and its responsiveness to detection of clinical change. It is important to have outcome measures that can be applied across different cultural or language subgroups. Although this cross-validation is well under way, future studies may focus on whether the pain/disability experience in neck pain varies across the subgroups and how these variations are reflected on the NDI responses.It is certainly true that no other instrument has undergone sufficient development or validation to replace the NDI as the instrument of choice in routine evaluation of neck disability. The instrument is suitable for both clinical research and practice, although evidence and measurement principles suggest that patients who score at the extremes may benefit from supplemental scales, like the Patient-Specific Functional Scale.
It is an important consideration that introducing new instruments into clinical practice and research requires tremendous efforts in translation/cultural adaptation and knowledge translation. For these reasons, any suggestions of change to different instruments or in the NDI itself should be balanced with this consideration. The most recent suggestion is that an 8-item version of the NDI performs as well as the original NDI and can be considered as interval data. An advantage of this suggestion is that it is easier to change to a reduced-item instrument than a different one.
Finally, a practical issue for clinicians is the difficulty of obtaining official translations, as these are not usually published with validation studies. Because these are not published with the validation manuscript, it is the responsibility of the clinician to seek out these tools by contacting developers. However, moves to improve the accessibility of translations would benefit clinical practice.While the NDI may be considered the “gold standard” among outcome measures, this review suggests further investigation is needed. Because the NDI was not developed using a clinimetric process and original pilot testing was conducted on a very small sample, it is not clear whether the NDI captures all of the important concepts for patients with neck pain or weighs pain and disability according to their relative priority. This leaves open the possibility that the addition of items might enhance performance. Others have suggested that removal of items might improve performance. Currently, variations in scoring exist in the literature; thus, caution should be exercised when presenting or comparing scores as to whether scores are out of 50 points or 100%.Perhaps, most importantly, there are gaps in defining clinically useful comparative data and benchmarks.
For example, the work on MDC and CID is sparse and inconsistent. Defining detectable change and important difference for different clinical subgroups and the impact of different prognostic variables on these would allow clinicians to provide more accurate prognosis and outcome evaluation.
These would be important in assisting clinicians in using the NDI to set short- and long- term goals, and to communicate more effectively with payers.