Defining Substantial Clinical Benefit Following Lumbar Spine Arthrodesis

Glassman, Steven D. MD; Copay, Anne G. PhD; Berven, Sigurd H. MD; Polly, David W. MD; Subach, Brian R. MD; Carreon, Leah Y. MD, MSc

Journal of Bone & Joint Surgery - American Volume:
doi: 10.2106/JBJS.G.01095
Scientific Articles
Abstract

Background: Validated health-related quality-of-life measures have become important standards in the evaluation of the outcomes of lumbar spine surgery. However, there are few well-defined criteria for clinical success based on these measures. The minimum clinically important difference is an important demarcation, but it could be considered a floor value rather than a goal in defining clinical success. Therefore, we sought to define thresholds of substantial clinical benefit for commonly used health-related quality-of-life measures following lumbar spine arthrodesis.

Methods: Prospectively collected preoperative and one-year postoperative health-related quality-of-life measures from 357 patients who were managed with lumbar spine arthrodesis for the treatment of degenerative conditions were identified. Candidate substantial clinical benefit thresholds for the Short Form-36 physical component score, Oswestry Disability Index, and back and leg pain numeric rating scales were identified with use of receiver operating characteristic curve analysis. Receiver operating characteristic curves were used to discriminate between patients who reported being “much better” or “about the same” with use of the validated Short Form-36 health transition item and between those who reported being “mostly satisfied” or “unsure” with use of a nonvalidated but more surgery-specific satisfaction-with-results survey. For each health-related quality-of-life measure, three response parameters were used: net change, percent change, and raw score at the time of the one-year follow-up.

Results: Substantial clinical benefit thresholds for the Short Form-36 physical component score were a 6.2-point net improvement, a 19.4% improvement, or a final raw score of ≥35.1 points. Substantial clinical benefit thresholds for the Oswestry Disability Index were an 18.8-point net improvement, a 36.8% improvement, or a final raw score of <31.3 points. Substantial clinical benefit thresholds for the back pain and leg pain numeric rating scales were a 2.5-point net improvement or a final raw score of <3.5 points. Substantial clinical benefit thresholds for percent change were 41.4% for the back pain numeric rating scale and 38.8% for the leg pain numeric rating scale.

Conclusions: We believe that thresholds of substantial clinical benefit for commonly used health-related quality-of-life measures following lumbar spine arthrodesis are important as they describe a magnitude of change that the patient recognizes as a major improvement.

Author Information

1Kenton D. Leatherman Spine Center, 210 East Gray Street, Suite 900, Louisville, KY 40202. E-mail address for S.D. Glassman: tallgeyer@spinemds.com

2The Spinal Research Foundation, 1831 Wiehle Avenue, Reston, VA 20190

3Department of Orthopaedic Surgery, University of California at San Francisco, Box 0728, 500 Parnassus Avenue, MU 320, San Francisco, CA 94143-0728

4Department of Orthopaedic Surgery, University of Minnesota, 2450 Riverside Avenue South, R 200, Minneapolis, MN 55454

Article Outline

Validated patient-based outcome measures such as the Medical Outcomes Study Short Form-36 (SF-36)1 and the Oswestry Disability Index2 have become standards for clinical assessment, providing the opportunity to measure the effect of lumbar spine interventions on health-related quality of life and to develop an evidence-based approach to care3-11. Simultaneously, the evaluation of larger study populations has created the paradox that small, clinically unimportant changes in outcome may become significant. This has led to the application of the minimum clinically important difference, an analysis that identifies a threshold of improvement that is clinically relevant as well as significant12-18.

However, as the minimum clinically important difference is defined as a limited rather than an optimal clinical benefit, it should probably be considered as a floor value, not a target, in defining clinical success. Furthermore, in the evaluation of a group or cohort, a mean change at the minimum clinically important difference level implies that some of the patients have exceeded the threshold level of improvement whereas others have not reached the threshold. Therefore, it is the proportion of patients reporting improvement above or below a minimum clinically important difference threshold that may be the better measure of the group's success.

Having established thresholds for the minimum clinically important difference in patients with lumbar spine abnormalities, it will also be valuable to identify a magnitude of health-related quality-of-life improvement that a patient recognizes as a substantial benefit. Hence, the purpose of the present study was to define thresholds of clinical improvement that represent a substantial clinical benefit for commonly used health-related quality-of-life measures following lumbar spine arthrodesis for the treatment of degenerative disorders.

Back to Top | Article Outline

Materials and Methods

A data set of prospectively collected preoperative and one-year postoperative patient-reported outcome measures for 357 patients who were managed with lumbar spine arthrodesis forms the basis for the present study. Health-related quality-of-life measures were administered to consecutive patients who were managed by one of five fellowship-trained spine surgeons (including one of the authors [S.D.G.]) at a tertiary spine center as a part of their standard clinical practice. The study population included 148 men (41.5%) and 209 women (58.5%) who had a mean age of 53.8 years (range, twenty-three to eighty-three years) at the time of the index procedure. Sixty-one patients (17.1%) were smokers, and the average body mass index was 29.8 kg/m2. The most frequent diagnoses were spinal stenosis (25.1%), lumbar disc degeneration (25.9%), and spondylolisthesis (11.4%). Other diagnoses included degenerative scoliosis, post-discectomy instability, adjacent-level degeneration, flexion-extension instability, and discogenic pain. Patients undergoing surgery for the treatment of infection, tumor, trauma, or deformity were excluded.

The patients self-reported their health status with use of four outcome instruments: the Medical Outcomes Study Short Form-36 (SF-36), the Oswestry Disability Index, and numerical rating scales for back and leg pain. Complete data were available for all patients, and the SF-36 physical component score was used for analysis. For analysis, we used the baseline score recorded at the time of enrollment (the time at which the patient was scheduled for surgery), one-year follow-up scores, and change scores (calculated as the difference between the score at the time of the one-year follow-up and the baseline score). Oswestry Disability Index change scores and pain numeric rating scale scores were multiplied by −1 so that a positive change score would reflect an improvement for all health-related quality-of-life measures. A table in the Appendix lists definitions of terms used in this study.

The primary global assessment rating anchor was the health transition item from the SF-3619,20. The health transition item is not used for scoring the scales or the summary measures of the SF-3621 and is therefore an independent assessment of patient satisfaction. The health transition item asks patients to compare their current health with their health one year ago. The available answers are “much better,” “somewhat better,” “about the same,” “somewhat worse,” or “much worse.” Although this anchor question is not specific to the surgical intervention, it is advantageous in that it is routinely available as part of the SF-36 survey. A second nonvalidated global assessment rating anchor was constructed from a set of five satisfaction statements that were administered concurrently with the one-year follow-up outcome surveys. These satisfaction statements were related specifically to the lumbar spine surgery (Table I). A numerical value (from 1 to 5) was attributed to patient answers (from “definitely true” to “definitely false”), and the values were summed for each of the five questions. The summed answers were divided into five equal intervals labeled “most satisfied,” “satisfied,” “unsure,” “dissatisfied,” and “most dissatisfied.” The psychometric properties of this scale are not known, but it was constructed to provide a direct assessment of the outcomes of the spine surgery.

Candidate substantial clinical benefit threshold values were developed with use of receiver operating characteristic curve analysis. Receiver operating characteristic curves assess each potential threshold value in order to optimize sensitivity and specificity in differentiating between cohorts. Sensitivity refers to the ability of the threshold to define all patients who self-report improvement on the global assessment rating. Specificity refers to the ability of the threshold to exclude patients who did not recognize an improvement on the global assessment rating. For each anchor, receiver operating characteristic curves were used to discriminate between two subgroups of patients: the “much better” and “about the same” patients on the SF-36 health transition item, and the “mostly satisfied” and “unsure” patients on the concurrent nonvalidated satisfaction survey.

Receiver operating characteristic curve accuracy is measured by the calculated area under the curve. A receiver operating characteristic curve with an area under the curve equal to 0.80 implies that, 80% of the time, a patient who reports being “much better” will have a higher health-related quality-of-life score than a patient who reports being “about the same.” The area under the curve was classified as excellent (0.90 to 1.00), good (0.80 to 0.90), fair (0.70 to 0.80), poor (0.60 to 0.70), and failed (0.50 to 0.60)22,23. For each health-related quality-of-life measure, three sets of scores were analyzed: the net change, the percent change, and the raw score at the time of the one-year follow-up. Statistical analyses were carried out with the statistical program for social sciences. The level of significance was set at p = 0.05.

Back to Top | Article Outline

Results

Preoperative and one-year postoperative mean health-related quality-of-life scores are reported in Table II.

Back to Top | Article Outline
SF-36 Physical Component Score
Net Change (Tables III and IV)

The mean net change in the SF-36 physical component score was 13.8 for patients who reported being “much better” and 2.4 for patients who were “about the same” postoperatively. With use of the satisfaction scale, the corresponding mean net changes were 14.0 for the “mostly satisfied” group and 2.9 for the “unsure” group. Receiver operating characteristic curve analysis based on the SF-36 health transition anchor identified a net change of 6.2 points (area under the curve, 0.846) as the threshold that best differentiated the “much better” and “about the same” groups (Fig. 1).

Back to Top | Article Outline
Percent Change (Tables III and V)

When the “much better” and “about the same” groups were compared on the basis of the percent change in the SF-36 physical component score, the mean percent change was 51.3% in the “much better” group and 10.8% in the “about the same” group. Receiver operating characteristic curve analysis based on the SF-36 health transition anchor identified an optimal threshold change of 19.4% (area under the curve, 0.820). Analysis based on the satisfaction score anchor identified a 23.4% change (area under the curve, 0.806) in the SF-36 physical component score as the optimal division between “mostly satisfied” and “unsure” groups.

Back to Top | Article Outline
One-Year Postoperative Score (Tables III and VI)

The mean SF-36 physical component score at one year was 43.9 for patients who reported being “much better” and 31.0 for patients who were “about the same” postoperatively. An optimal one-year postoperative raw SF-36 physical component score threshold was found to be ≥35.1 points (area under the curve, 0.854). Eighty-one percent of patients in the “much better” group had a final raw SF-36 physical component score at or above the 35.1-point threshold, compared with 18% of patients in the “about the same” group. The satisfaction scale anchor generated a one-year postoperative physical component score threshold of 34.7 points (area under the curve, 0.901).

Back to Top | Article Outline
Oswestry Disability Index
Net Change (Tables III and IV)

With use of the SF-36 health transition anchor, the mean change in the Oswestry Disability Index score was 29.5 for the “much better” group and 7.2 for the “about the same” group. The receiver operating characteristic threshold between the two groups was 18.8 points (area under the curve, 0.875). Similar values were observed with use of the satisfaction score anchor, with a mean change of 30.4 for the “mostly satisfied” group and 8.2 for the “unsure” group.

Back to Top | Article Outline
Percent Change (Tables III and V)

With use of the health transition anchor, evaluation of the percent change revealed an improvement of 61.8% for the “much better” group and 12.4% for the “about the same” group. The receiver operating characteristic threshold between the groups was 36.8% for the SF-36 health transition anchor (area under the curve, 0.891) and 37.5% for the combined satisfaction score anchor (area under the curve, 0.944).

Back to Top | Article Outline
One-Year Postoperative Score (Tables III and VI)

With use of the health transition anchor, the mean one-year postoperative raw Oswestry Disability Index was 18.4 for the “much better” group and 46.4 for the “about the same” group. Receiver operating characteristic thresholds between the optimal and fair groups for the one-year postoperative raw Oswestry Disability Index were 31.3 for the health transition anchor (area under the curve, 0.901) and 31.7 for the satisfaction score anchor (area under the curve, 0.948) (Fig. 2).

Back to Top | Article Outline
Numeric Back Pain Score
Net Change (Tables III and IV)

The health transition and satisfaction score anchors yielded almost identical values for the mean change in the back pain numeric rating scale: 4.9 points for the “much better” group and 1.5 points for the “about the same” group. The receiver operating characteristic curve threshold dividing the optimal and fair outcome groups was 2.5 points in both instances (area under the curve, 0.826 for the health transition anchor and 0.838 for the satisfaction score anchor).

Back to Top | Article Outline
Percent Change (Tables III and V)

Examination of percent change revealed a 61.6% improvement in the “much better” group as compared with a 10.1% improvement in the “about the same” group with use of the health transition anchor and a 59.3% improvement in the “mostly satisfied” group as compared with a 9.1% improvement in the “unsure” group with use of the satisfaction score anchor. Receiver operating characteristic curve analysis revealed a dividing threshold of 41.4% between optimal and fair outcome groups for both the health transition anchor (area under the curve, 0.865) and the satisfaction score anchor (area under the curve, 0.871).

Back to Top | Article Outline
One-Year Postoperative Score (Tables III and VI)

The mean one-year postoperative raw score for numeric back pain was 2.1 points for the “much better” group and 6.2 points for the “about the same” group with use of the health transition anchor. Receiver operating characteristic curve-generated thresholds were 3.5 points in both instances (area under the curve, 0.909 for the health transition anchor and 0.931 for the satisfaction score anchor).

Back to Top | Article Outline
Numeric Leg Pain Score
Net Change (Tables III and IV)

The mean leg pain numeric rating scale was similar for the two global assessment rating anchors. The mean score was 4.7 points for the “much better” group and 1.2 points for the “about the same” group. The receiver operating characteristic curve-generated threshold between the two groups was 2.5 points in both instances (area under the curve, 0.763 for the health transition anchor and 0.815 for the satisfaction score anchor).

Back to Top | Article Outline
Percent Change (Tables III and V)

The mean percent change was 56.8% for the “much better” group and −8.7% for the “about the same” group, as compared with 65.7% for the “mostly satisfied” group and −6.8% for the “unsure” group. Receiver operating characteristic curve analysis revealed a dividing threshold of 38.8% (area under the curve, 0.761) (Fig. 3) and 47.2% (area under the curve, 0.841) between the optimal and fair outcome groups for the health transition and satisfaction score anchors, respectively.

Back to Top | Article Outline
One-Year Postoperative Score (Tables III and VI)

The mean one-year postoperative final raw score for the leg pain numeric rating scale was 2.0 points for the “much better” group and 5.6 points for the “about the same” group with use of the health transition anchor. The receiver operating characteristic curve-generated threshold for the health transition anchor was 3.5 points (area under the curve, 0.804).

Back to Top | Article Outline

Discussion

Health-related quality-of-life measures, particularly the Oswestry Disability Index and SF-36 physical component score, have become important standards for the evaluation of the outcomes of lumbar spine surgery3-8,11,17. Despite their widespread use, there are few well-defined criteria for clinical success based on these measures. One key threshold, the minimum clinically important difference, has been defined and is being utilized with increasing frequency. Although the minimum clinically important difference represents a meaningful demarcation, it is inherently more of a floor value than a goal in terms of defining clinical success. Therefore, we sought to identify values that are indicative of substantial clinical benefit for health-related quality-of-life measures that are commonly used to evaluate the results of lumbar spine arthrodesis (Table VII).

The present study compared outcome scores and global assessment rating anchor responses for four health-related quality-of-life instruments: the SF-36 physical component score, the Oswestry Disability Index, and numeric back and leg pain scales. We sought to differentiate between patients who identified themselves as “much better” or “about the same” on the basis of either the SF-36 health transition item or a satisfaction-with-results survey. Potential thresholds of substantial clinical benefit for each of the health-related quality-of-life measures were investigated on the basis of three relevant parameters (net change in outcome score, percent change in outcome score, and one-year postoperative raw score) as there are conceptual advantages to each assessment tool. Net change seems to be the most versatile benchmark for demonstrating a substantial clinical benefit. Net change is simple to calculate, is simple to understand, and can be readily used to interpret the benefit of a given intervention. It is important to recognize that all of the health-related quality-of-life scales used are nonlinear in that a 10-point change at the low end does not equal a 10-point change at the higher end, so there may be limitations of net change at points away from the mean.

Percent change gives weight to preoperative health status as patients with a lower preoperative health status require less absolute change to result in a larger percent change. In a study of patients with back pain that was treated nonoperatively, Lauridsen et al.24 found a minimum clinically important difference based on percent change to be more independent of the patients' baseline severity. The use of a final raw score threshold is intriguing in that a preoperative score might not be required for assessment. The ability to identify a threshold with use of postoperative scores alone suggests that some absolute threshold of pain or disability must be reached in order for patients to be satisfied with the end point of care. The disadvantage of the final raw score threshold is that it does not directly convey information on effect size or the performance of a given intervention.

The weaknesses of the present study include the fact that the relationship between the change in health-related quality-of-life scores and perceived benefit was studied at only a single time-interval, one year postoperatively. It is unknown whether this relationship is time-sensitive or if the thresholds might differ with either shorter or longer-term follow-up. Furthermore, the impact of baseline severity on substantial clinical benefit is not clearly elucidated by this analysis. Previous studies have emphasized this parameter, but only a minimal effect of baseline severity was observed in a previous study of minimum clinically important difference thresholds in our patient population24,25.

The present analysis of health-related quality-of-life measures in a large population of patients managed with lumbar spine surgery suggests that appropriate thresholds of substantial clinical benefit for the SF-36 physical component score are a net improvement of 6.2 points, a percent improvement of 19.4%, or a final raw score of ≥35.1 points. Substantial clinical benefit thresholds for the Oswestry Disability Index are a net improvement of 18.8 points, a percent improvement of 36.8%, or a final raw score of ≤31.3 points. Substantial clinical benefit thresholds for numeric rating scale back pain and leg pain are a net improvement of 2.5 points or a final raw score of ≤3.5 points. Substantial clinical benefit thresholds for percent change are 41.4% for the back pain numeric rating scale and 38.8% for the leg pain numeric rating scale. Comparable published values for the minimum clinically important difference are noted in Table VII.

One critical issue in utilizing either the minimum clinically important difference or the substantial clinical benefit threshold is to recognize that these are individual measures. In the evaluation of any given cohort, the appropriate measure would be the percentage of patients reaching the minimum clinically important difference or substantial clinical benefit threshold. Also, as no single clinical or radiographic measure can adequately reflect the value of a surgical procedure, these thresholds are not intended as benchmarks to define a satisfactory intervention.

However, as the utilization of patient-reported outcomes is a primary tenet of an evidence-based approach to care, a framework for understanding health-related quality-of-life scores is important in the surgeon's interpretation of evidence-based literature. As such, these thresholds may be useful as tools for discussion with patients and for shared decision-making. Substantial clinical benefit is critical because, when surgery is being considered, it is the probability of substantial benefit, not a minimal benefit, that is likely to direct treatment selection for both the patient and the physician.

Back to Top | Article Outline

Appendix Cited Here...

A table showing definitions of terms used in this report is available with the electronic versions of this article, on our web site at jbjs.org (go to the article citation and click on “Supplementary Material”) and on our quarterly CD/DVD (call our subscription department, at 781-449-9780, to order the CD or DVD).

Disclosure: In support of their research for or preparation of this work, one or more of the authors received, in any one year, outside funding or grants of less than $10,000 from Medtronic Sofamor Danek. In addition, one or more of the authors or a member of his or her immediate family received, in any one year, payments or other benefits in excess of $10,000 or a commitment or agreement to provide such benefits from a commercial entity (Medtronic Sofamor Danek). Also, a commercial entity (Medtronic Sofamor Danek) paid or directed in any one year, or agreed to pay or direct, benefits in excess of $10,000 to a research fund, foundation, division, center, clinical practice, or other charitable or nonprofit organization with which one or more of the authors, or a member of his or her immediate family, is affiliated or associated.

A commentary is available with the electronic versions of this article, on our web site (www.jbjs.org) and on our quarterly CD-ROM/DVD (call our subscription department, at 781-449-9780, to order the CD-ROM or DVD).

Investigation performed at the Kenton D. Leatherman Spine Center, Louisville, Kentucky

1. Ware JE Jr, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care. 1992;30:473-83.
2. Fairbank JC, Pynsent PB. The Oswestry Disability Index. Spine. 2000;25:2940-52.
3. Weinstein JN, Lurie JD, Tosteson TD, Skinner JS, Hanscom B, Tosteson AN, Herkowitz H, Fischgrund J, Cammisa FP, Albert T, Deyo RA. Surgical vs nonoperative treatment for lumbar disk herniation: the Spine Patient Outcomes Research Trial (SPORT) observational cohort. JAMA. 2006;296:2451-9.
4. Weinstein JN, Tosteson TD, Lurie JD, Tosteson AN, Hanscom B, Skinner JS, Abdu WA, Hilibrand AS, Boden SD, Deyo RA. Surgical vs nonoperative treatment for lumbar disk herniation: the Spine Patient Outcomes Research Trial (SPORT): a randomized trial. JAMA. 2006;296:2441-50.
5. Brox JI, Sørensen R, Friis A, Nygaard Ø, Indahl A, Keller A, Ingebrigtsen T, Eriksen HR, Holm I, Koller AK, Riise R, Reikerås O. Randomized clinical trial of lumbar instrumented fusion and cognitive intervention and exercises in patients with chronic low back pain and disc degeneration. Spine. 2003;28:1913-21.
6. Fairbank J, Frost H, Wilson-MacDonald J, Yu LM, Barker K, Collins R; Spine Stabilisation Trial Group. Randomised controlled trial to compare surgical stabilisation of the lumbar spine with an intensive rehabilitation programme for patients with chronic low back pain: the MRC spine stabilisation trial. BMJ. 2005;330:1233. Erratum in: BMJ. 2005;330:1485.
7. Fritzell P, Hägg O, Wessberg P, Nordwall A; Swedish Lumbar Spine Study Group. Lumbar fusion versus nonsurgical treatment for chronic low back pain: a multicenter randomized controlled trial from the Swedish Lumbar Spine Study Group. Spine. 2001;26:2521-34.
8. Malmivaara A, Slätis P, Heliövaara M, Sainio P, Kinnunen H, Kankare J, Dalin-Hirvonen N, Seitsalo S, Herno A, Kortekangas P, Niinimäki T, Rönty H, Tallroth K, Turunen V, Knekt P, Härkänen T, Hurri H; Finnish Lumbar Spinal Research Group. Surgical or nonoperative treatment for lumbar spinal stenosis? A randomized controlled trial. Spine. 2007;32:1-8.
9. Burkus JK, Sandhu HS, Gornet MF, Longley MC. Use of rhBMP-2 in combination with structural cortical allografts: clinical and radiographic outcomes in anterior lumbar spinal surgery. J Bone Joint Surg Am. 2005;87:1205-12.
10. Dimar JR, Glassman SD, Burkus KJ, Carreon LY. Clinical outcomes and fusion success at 2 years of single-level instrumented posterolateral fusions with recombinant human bone morphogenetic protein-2/compression resistant matrix versus iliac crest bone graft. Spine. 2006;31:2534-40.
11. Glassman S, Gornet MF, Branch C, Polly D Jr, Peloza J, Schwender JD, Carreon L. MOS short form 36 and Oswestry Disability Index outcomes in lumbar fusion: a multicenter experience. Spine J. 2006;6:21-6.
12. Ware JE Jr, Kosinski M, Keller SD. SF-36 physical and mental health summary scales: a user's manual. Boston: The Health Institute, New England Medical Center; 1994.
13. Hägg O, Fritzell P, Nordwall A; Swedish Lumbar Spine Study Group. The clinical importance of changes in outcome scores after treatment for chronic low back pain. Eur Spine J. 2003;12:12-20.
14. Crosby RD, Kolotkin RL, Williams GR. Defining clinically meaningful change in health-related quality of life. J Clin Epidemiol. 2003;56:395-407.
15. Copay AG, Subach BR, Glassman SD, Polly DW Jr, Schuler TC. Understanding the minimum clinically important difference: a review of concepts and methods. Spine J. 2007;7:541-6.
16. Osoba D, Rodrigues G, Myles J, Zee B, Pater J. Interpreting the significance of changes in health-related quality-of-life scores. J Clin Oncol. 1998;16:139-44.
17. Taylor SJ, Taylor AE, Foy MA, Fogg AJ. Responsiveness of common outcome measures for patients with low back pain. Spine. 1999;24:1805-12.
18. Grotle M, Brox JI, Vøllestad NK. Concurrent comparison of responsiveness in pain and functional status measurements used for patients with low back pain. Spine. 2004;29:E492-501.
19. Walters SJ, Brazier JE. Comparison of the minimally important difference for two health state utility measures: EQ-5D and SF-6D. Qual Life Res. 2005;14:1523-32.
20. Walters SJ, Brazier JE. What is the relationship between the minimally important difference and health state utility values? The case of the SF-6D. Health Qual Life Outcomes. 2003;1:4.
21. Ware JE Jr. SF-36 health survey update. Spine. 2000;25:3130-9.
22. Hosmer DW, Lemeshow S. Applied logistic regression. 2nd ed. Hoboken, NJ: John Wiley and Sons; 2000.
23. Tape TG. Interpreting diagnostic tests. http://gim.unmc.edu/dxtests/Default.htm. Accessed 2006 Dec 18.
24. Lauridsen HH, Hartvigsen J, Manniche C, Korsholm L, Grunnet-Nilsson N. Responsiveness and minimal clinically important difference for pain and disability instruments in low back pain patients. BMC Musculoskelet Disord. 2006;7:82.
25. Copay A, Glassman S, Subach B, Berven S, Schuler T, Carreon L. The minimum clinically important difference in lumbar spine surgery patients: a choice of methods using the Oswestry Disability Index, MOS Short Form-36, and Pain Scales. Spine J. 2007;7(5 Suppl):60S.
Copyright 2008 by The Journal of Bone and Joint Surgery, Incorporated