Subjective and Objective Measures of Symptoms, Function, and Outcome in Patients With Degenerative Spine Disease

Abstract

The management of patients with degenerative diseases of the spine requires reliable measures of symptoms, function, and outcome. Choice of conservative or surgical therapy in this cohort is complex and depends on many factors, including the history, quality, and severity of pain; functional limitations; and health-related quality of life (HRQOL) 1, 2. An accurate and thorough evaluation of both the subjective and objective condition is helpful to select adequate treatment for the individual patient 3. In recent years, subjective patient-reported outcome measures (PROMs) in the form of questionnaires have been established as the gold standard for the clinical evaluation of spine patients 4. General pain measures, including the visual analog scale (VAS) or the numeric rating scale (NRS) for low back pain (LBP) or irradiating extremity (arm/leg) pain, are among the most commonly used PROMs. However, they are generic and not specific tools for spinal diseases 5. The Oswestry Disability Index (ODI), which is included in this review, is a prime example of an in-depth validated PROM and represents one of the most established instruments for a variety of different diseases of the lumbar spine 6, 7. There is a broad variety of degenerative spine diseases, some of which are unspecific, whereas others (eg, lumbar spinal stenosis [LSS]) present with characteristic symptoms (eg, neurogenic claudication) that can be addressed by disease-specific tools 8. Cervical or thoracic degenerative spine disease may present with both radicular pain and/or myelopathy; outcome measures have to account for these different clinical manifestations 9. Apart from subjective PROMs, objective measures of function are gaining increasing attention in spine research and have found their way into clinical practice 8, 10. Measurements like the Timed Up-and-Go (TUG) test and the motorized treadmill test (MTT) assess a patient's objective functional impairment (OFI) and add a new dimension to the comprehensive patient evaluation 11. Because patients prefer objective functional tests rather than questionnaires and considering the continuous validation and standardization of objective outcome measures, we included the two most frequently applied tests of this relatively new field of outcome assessment in this review 8, 12. Degenerative disease of the spine encompasses a wide range of different pathologies and disease-specific symptoms. This leads to an even greater number of outcome measures that cannot all be included in the scope of this review. However, this article should provide the reader with a comprehensive summary of carefully selected instruments (Tables 1 and 2) 5, 8, 13. The SSM was developed in 1995 by Stucki et al as a short self-administered questionnaire to assess pain-related disability and health-related parameters in patients diagnosed with LSS 14. The SSM is also known as the Zurich Claudication Questionnaire, the Swiss Spinal Stenosis Questionnaire, or the Brigham Spinal Stenosis Questionnaire. The measure specifically addresses symptoms and functional deficits resulting from neurogenic claudication 14. It also includes an optional domain on patient satisfaction regarding the result of surgery. The SSM is one of the leading PROMs used by both spine surgeons and rheumatologists 5, 15. The SSM contains three scales, with seven questions on symptom severity (SSM symptoms), five questions on physical function (SSM function), and six questions on satisfaction (SSM satisfaction). The SSM symptom scale encompasses a pain subdomain (three items) and a neuroischemic subdomain (four items). Eighteen items. Each item is measured on a three- to five-point ordinal scale, ranging from the best scenario to the worst scenario. The scale relates to symptoms experienced over the past month. None. The items are reported in the original publication 14. A copy is provided as an appendix to this article (English version). The SSM is self-administered by patients using paper and pen. Administration by telephone is also possible. The symptom severity score is the unweighted mean of the first seven questions; the resulting score indicates greater symptom severity. The first six items are scored from 1 to 5, whereas the seventh item has only three possible responses corresponding to scores of 1, 3, or 5. The symptom severity score can be further divided into a subscale for pain (questions one through three) and neuroischemic symptoms (questions four through seven). The physical function score is calculated as the unweighted mean of the following five questions, all scored from 1 to 4, with higher scores representing greater disability. The third scale relates to patient satisfaction after treatment, with higher scores indicating greater satisfaction. The unweighted mean is only valid if no more than one item is missing in the SSM function and SSM satisfaction scores and no more than two items are missing in the SSM symptom score. Some authors prefer to depict the SSM as the sum score without taking the unweighted mean. An unweighted mean is calculated for all three subscales. The SSM symptom severity score ranges from 1 (no symptoms) to 5 (maximal symptoms), which is further divided into the pain subscales (scores between 1 and 5) and neuroischemic symptom subscale (scores between 1 and 5). The SSM functions score ranges from 1 (no disability) to 4 (maximum disability), and the SSM satisfaction score ranges from 1 (very satisfied) to 4 (very dissatisfied). When the SSM is declared as a sum score, the SSM symptom score ranges from 7 (no symptoms) to 35 (maximal symptoms), the SSM functions score ranges from 5 (no disability) to 20 (maximum disability), and SSM satisfaction score ranges from 6 (very satisfied) to 24 (very dissatisfied). The SSM is relatively short and simple to fill out. It can be completed by the patient in less than 5 minutes. The SSM requires little administrative burden. The time to score (by hand) is short. No training is required. Translations are available in German and English. There are published adaptations in Spanish, Chinese, French, Polish, Norwegian, Iranian, Korean, Slovenian, and other languages, which showed satisfactory to excellent reliability and validity 15-23. The authors are not aware of any studies reporting on floor or ceiling effects of the scales. The reliability in test-retest assessment consistently scores of more than 80% to 90% agreement 24-26. The SSM has been validated against the self-paced walking test, the ODI, the Oxford Spinal Stenosis Score, the Short Form 36 (SF-36), the Health Utilities Index, the Center for Epidemiologic Studies Depression Scale, and other measures with high correlations 25-27. Several studies have supported the construct validity of the SSM and psychometric properties and physical function in the LSS population 25, 27, 28. Standardized response means (SRMs) are reported to be 1.07 for the SSM function scale and 0.96 for the SSM symptom scale in the original study with 130 patients 28. In a subsequent independent study among 91 patients with LBP, the SRM of SSM function (1.00) and SSM symptom (1.43) were confirmed, and the composite SSM average score even increased in responsiveness (1.50) 29. Because of its disease specificity, the SSM has been shown to be more responsive than the Sickness Impact Profile and the Roland-Morris Disability Questionnaire (RMDQ) in assessing patients with LSS 28. Stucki et al reported a difference in the physical function scale and symptom severity scores between the unsatisfied and somewhat satisfied patients of 0.52 (13%) and 0.48 (9.6%) 14. The SSM is specific to symptoms and functional deficits characteristic of neurogenic claudication. For LSS, excellent generalizability has been proven by consistent results across multicultural studies 16-23. The SSM has been used in various recent studies and clinical trials 30-33. The SSM is a short disease-specific questionnaire for patients with LSS with three different subscales, namely, symptom severity, physical function, and satisfaction. The SSM proved to be the most precise in patients with LSS when compared with the ODI or the Oxford Claudication Score 25. A publication by Comer et al used Rasch analysis to evaluate the psychometric properties of the SSM 24. In their study, the SSM symptom scale, by measuring pain and neuroischemic symptoms as two separate constructs, failed to function as a unidimensional domain. Because of this, the SSM symptom scale was further subdivided into the two separate subscales. The SSM function proved valid for group comparison, although question 11 was not considered clinically meaningful 24. The SSM has established psychometric properties, and both administrative and respondent burdens are low. It is frequently used to assess and monitor outcome. As the current gold standard of outcome assessment in patients with LSS, it has been endorsed by the North American Spine Society and termed "best and most specific outcome measure for LSS" 34. The SSM is widely used in clinical research. The SSM symptom scale should be further subdivided into the SSM pain symptom and SSM neuroischemic symptom scales to adequately address these separate conditions. Ever since a multinational group of experts proposed the COMI as a standardized outcome assessment in LBP research in 1998, this short multidimensional outcome measure has been thoroughly validated. It is widely used, especially to monitor outcomes in patients with various spinal disorders who are undergoing surgery or any form of intervention 35-37. Initially proposed as a set of six questions, a seventh item (overall quality of life) was added in 2005 by Mannion et al to construct what is known as the COMI Back 38. A validated and reliable version designed for cervical spine diseases, the Core Outcome Measures Index Neck, is also available 39, 40. The COMI Back focuses on the patient's perspective 29. Following this premise, it covers the domains of pain intensity (LBP and leg/buttock pain), back-related function, symptom-specific well-being, general quality of life, and social/work disability. A further part covers patient satisfaction after surgical treatment. Seven items. (There are an additional four questions at follow-up that cover patient satisfaction and treatment results.) The first two questions regarding pain intensity are measured on a 0 to 10 NRS. The other five items are measured on a five-point Likert scale ranging from "best" to "worst." All items refer to the last week (except for disability, which refers to the last 4 weeks). None. A copy in different languages (including English, German, French, and Spanish) can be downloaded from the EUROSPINE Spine Tango registry website (https://www.euros​pine.org/forms.htm). The COMI Back is self-administered by patients using paper and pen or online forms. The higher of the two pain scores (scores of 0-10) is taken as the pain intensity score. The other six items on the five-point Likert scales correspond with scores of 0, 2.5, 5.0, 7.5, and 10, respectively. The average of the two disability items form the disability score. The five domain scores for pain intensity, back-related function, symptom-specific well-being, general quality of life, and disability are then averaged to give a COMI Back score that ranges from 0 to 10. Scores range from 0 (best) to 10 (worst). The COMI Back is relatively short and simple to fill out. It can be completed by the patient in less than 5 minutes. The COMI Back requires little administration burden. The time to score (by hand) is short. No training is necessary. The COMI Back is available and cross-culturally validated in many languages, including English 38, German 38, Korean 41, Italian 42, French 43, Turkish 44, and Japanese 45 among others. In one of the original articles, floor or ceiling effects in the range of 20% to 50% were observed for some items of the COMI Back before surgery (function and symptom-specific well-being) and after surgery (disability and function) 40. Several research groups examined the psychometric properties of the COMI Back in patients with various pathologies presenting with LBP. The test-retest reliability consistently scored highly, with an intraclass correlation coefficient (ICC) of 0.8 to 0.9 or more. High internal consistency for the core item index was shown for patients with chronic LBP (Cronbach's α = 0.90 or more); however, less was shown in patients with acute osteoporotic fractures (Cronbach's α = 0.64 or more) 38, 40, 46. Good construct validity of the score was demonstrated by a moderate to high correlation with reference questionnaires, including the RMDQ, the SF-36, and the ODI (r = 0.60-0.79) 38, 46. The item symptom-specific well-being, however, showed little correlation to other measures (r = 0.25-0.31), which might indicate that this item delivers unique information that may be of importance to the multidimensional nature of the overall index 38. Internal responsiveness corresponding with the SRM at the 12-month follow-up in a cohort of 91 patients undergoing surgery for LSS showed great responsiveness (SRM 1.44), which was similar to or better than the SSM average (1.50), the RMDQ (1.13), and the NRS pain (1.28) 29. In the same cohort, external responsiveness—meaning the strength of correlation between its change in scoring and the change in other outcome instruments—showed moderate correlation with the SSM (r = 0.62) and RMDQ (r = 0.43). In an area under the receiver operating characteristics (AUROC) curve analysis, good discriminative ability between good and bad outcomes was demonstrated (area under the curve of more than 0.83) 30. One of the original studies examining the COMI Back in 277 patients with LBP reported a minimum clinically important difference for improvement (MCIDimp) as a two- to three-point decrease, depending on the anchor used to indicate treatment success 38. In a group of 3056 patients undergoing spine surgery for a variety of indications, the groups' mean MCIDimp at 12 months was −2.6 points, and the minimum clinically important difference for deterioration (MCIDdet) was 1.2 points. AUROC curves of 0.88 for the MCIDimp and 0.89 for the MCIDdet indicated good discriminative ability. The cutoffs for individual improvement and deterioration were −2.2 points or less (sensitivity 81%, specificity 83%) and 0.3 points or more (sensitivity 83%, specificity 88%), respectively 36. Psychometric properties and sensitivity to change of the COMI Back are strong in patients with LBP from a variety of causes 38, 40, 46, 47. Thus, it can be applied to a broad range of patients with LBP. The COMI Back has been used in multiple clinical trials 48-50. The COMI Back is a short but independently and repeatedly validated outcome measure for a variety of different pathologies presenting with LBP and leg pain. It is easily accessible and is associated with a low burden for both patients and physicians. As in other back-specific outcome measures 51, 52, the COMI Back's values for the MCIDdet are lower than those for the MCIDimp, indicating that the COMI is less responsive to deterioration than to improvement 36. The COMI Back has shown great psychometric properties and is brief enough to be practical for routine clinical use and quality management. Based on these values, it has been incorporated into the European EUROSPINE registry as an outcome questionnaire of choice 53. The availability and ease of administration is meant to encourage clinicians and researchers to collaborate in registries and research projects on a national and international level 54. However, on an international level, outcomes measures like the ODI and RMDQ are still more established 5. Outcome research in cervical spine patients is historically less developed compared with lumbar spine patients 5, 55. The NDI is one of the best established and most commonly used PROMs for chronic neck pain 56, 57. Originally published in 1991, it is also known as the Vernon-Mior Disability Index or the Neck Pain Disability Index 58, 59. The NDI measures pain and disability in patients with neck pain. The NDI covers the domains of pain intensity, personal care, lifting, work, headache, concentration, sleeping, driving, reading, and recreation. Ten items. Each of the 10 items are scored on a six-point Likert scale ranging from 0 (no pain/disability) to 5 points (maximal pain/disability). The questions refer to the current clinical condition and pain intensity. The exact time frame is not defined. None. Copies in English and other languages can be found in published sources or from the EUROSPINE Spine Tango registry website (https://www.euros​pine.org/forms.htm) 56, 58, 59. The NDI is self-administered by patients using paper and pen. All 10 items are scored from 0 to 5. Numeric responses are summed up to a total score ranging from 0 (best) to 50 (worst). The developer and others recommend scoring the NDI out of 50 points. If three or more items are missing, the questionnaire is not valid. If two or fewer items are missing, the score should be normalized to 50 58, 59. Besides presenting raw test results, many groups have expressed the score as a percentage, ranging from 0% to 100% neck disability. No consensus exists regarding score interpretation. The original developers suggested the following: scores between 0 and 4 represent no disability, scores between 5 and 14 represent mild disability, scores between 15 and 24 represent moderate disability, scores between 25 and 34 represent severe disability, and scores greater than 35 represent complete disability 56. The NDI is relatively short and simple to fill out. It can be completed by the patient in 3 to 8 minutes 56. The NDI requires no administration burden and takes less than 3 minutes to score by hand. No training necessary. The NDI was originally developed in English but has been culturally adapted and translated into several languages including German 60, 61, Spanish 62, Arabic 63, Chinese 64, Turkish 65, Japanese 66, Polish 67, Finnish 68, Greek 69, Portuguese, and other languages 70. Patients with baseline score in the lower 10th and the upper 90th percentile are subject to significant floor or ceiling effects, respectively. This is why caution is required when using the NDI to monitor outcome in these high- or low-performing patients 56, 71, 72. A systematic review published in 2009 analyzed 41 studies that examined at least one aspect of the psychometric properties of the NDI. A high test-retest reliability was demonstrated in populations with both acute and chronic neck pain, with a reliability coefficient of more than 0.90 in most studies 56. The NDI score correlates strongly (r of more than 0.70) with other neck disability measures, including the Neck Pain and Disability Scale, the Cervical Spine Outcome Questionnaire, the Disability Rating Index, and the VAS pain scale. Furthermore, high Cronbach's α scores of 0.70 to 0.96 show good internal consistency 56, 57, 60, 70. The NDI is, however, only moderately correlated with both physical and mental aspects of general health as assessed by the SF-36 56. The NDI has a good ability to detect changes over time. The SRM ranges from to in a systematic review 56. range from 5 of 50 to of 50 points 56. Neck pain of to show a lower of more than 5 compared with pain 56, It is important to that is disease specific but on average a change of 5 to 7 points on the NDI can be considered a clinically meaningful change in pain and disability. Because of its strong psychometric properties and its in several languages, the NDI can be used in a variety of different patient populations with neck pain. The NDI has been used in multiple clinical trials The NDI is a and responsive measure in various This includes patients with both acute and chronic as as patients with neck pain associated with and cervical 56, The across study populations and which the standardized of NDI results over time in clinical practice and research 56. The NDI not and although these are in patients with chronic neck pain 56. recommend scoring the NDI out of 50 points, as originally The NDI is one of the most PROMs for neck pain and is used by spine and 56, 57. studies reporting for specific pathologies and patient populations are required 56. The good psychometric properties using the NDI in research on cervical spine or neck like to refer to the article by et al for a comprehensive of the we provide only a brief and on the The ODI pain-related disability in patients with LBP associated with a wide range of It is especially in patients with disability Initially published in by it has been several 6, All of the ODI in however, not all have been subject to the same systematic validation as the The ODI version developed by et al is for general use and has been adapted by various spine 6, 7. The ODI contains one item on pain intensity and items on of care, lifting, sleeping, life, life, and Ten items. No is and the questionnaire is The ODI can be found in the original article 7. The ODI has been used in multiple clinical trials The ODI is one of the most commonly used clinical measures of pain-related disability in a variety of pathologies of the lower The ODI has established psychometric properties, a high reliability and as as low respondent and administrative burden The ODI is one of the most frequently used outcome questionnaires for assessing and outcomes in both patients undergoing surgical or conservative 5, It is frequently used as a to other outcome to the ODI, we like to refer to the article by et al for a more comprehensive of the Originally designed in the RMDQ physical disability to LBP It has been used for and chronic LBP as as in various clinical its first several have been proposed to for specific patients (eg, for those with However, changes only in and The use of the original RMDQ is The RMDQ the of and functions sleeping, It not items to physical disability but general impairment and health as items. No is and the questionnaire is in different languages is available on the RMDQ website A copy is in the original publication 7. The RMDQ has been used in multiple clinical trials it is to any spine the RMDQ is one of the most established and thoroughly measures for patients with LBP. It has psychometric properties ranging from to However, scores and changes be with caution some items are and the overall score not have Psychometric are for use in research. Rasch analysis and of score is as part of analysis 1 and The is the most frequently applied objective functional outcome measure in patients with lumbar degenerative diseases 8. It includes several that are to for patients with or disability resulting from lumbar walking and 8, The patient is to and back on an with on the the patient should up and as as possible to a on the floor at 3 from the this the should back to the and as as possible. The the time between up and None. is possible with any A is available in multiple languages and measures and scores the test on population reference It can be downloaded of in the and the The test not any for a a and 3 of walking It is that an takes the patient's however, patient have been found to be reliable for similar objective functional tests are if required. are expressed as raw test Because raw test it is to test results in a standardized for and Standardized scores a patient's from the population mean and can be calculated using the to as a standardized score by the high of the and and can disability on an scale test results can be to severity into no test results are less than the upper of the population range mild test results are between the and the percentile of the moderate test results are between the and percentile of the and severe test results are greater than the percentile of the The requires low administrative burden. takes less than 1 The administrative burden is lower than that of most PROMs, and the test was over a set of PROM by to of patients 12. No is for the It can be demonstrated to patients if For with and patients with spine the was set as the percentile of the test results 11. In subsequent that examined surgical with lumbar a of scored the population range (no Thus, the standardized of test values is to have a floor There is no upper time for the test, and the of standardized scores for the accurate of disability without a ceiling The showed an excellent = and reliability = with an of to 11. validity with PROMs, including VAS back (r = and VAS leg pain (r = (r = and ODI (r = as as the Short Form 12 physical summary (r = and Questionnaire (r = be demonstrated 11. Several studies demonstrated that the test is to the change in function of a patient with The SRM is in a of patients with lumbar who were examined before and 6 after lumbar spine surgery not In patients with lumbar a change in raw test time of at least is considered a clinically meaningful change in function The test has been used in a broad variety of pathologies of the lumbar spine 8. a studies have on the psychometric properties of the for a specific disease (eg, The has been used in multiple clinical trials The test is to and by patients 12. It is the most thoroughly validated objective outcome measure to assess in patients with degenerative disease of the lumbar spine 8. test result is independent from the of the and index and mental health but it takes disability resulting from lower extremity deficits into account The test cannot be applied to patients who are to walking of severe pain or There may be a significant of (eg, or which of test results in these patients to and without the for

Topics & Concepts