448 - You Scratch My Neck, I´ll Scratch Your Back: Towards a Total Disabilit...

#448 You Scratch My Neck, I´ll Scratch Your Back: Towards a Total Disability Assessment Model

Cervical Therapies and Outcomes

Poster Presented by: D. Ryan

Author(s):

D. Ryan (1)
T. Protopsaltis (1)
C. Ames (2)
J. Smith (3)
R. Hart (4)
I. Mccarthy (5)
C. Li (6)
T. Errico (1)
V. Lafage (1)
International Spine Study Group

(1) NYU Hospital for Joint Diseases, Orthopaedic Surgery, New York, NY, United States
(2) University of California, San Francisco Medical Center, Neurosurgery, San Francisco, CA, United States
(3) University of Virginia Medical Center, Neurosurgery, Charlottesville, VA, United States
(4) Oregon Health and Science University, Orthopaedic Surgery, Portland, OR, United States
(5) Baylor Scoliosis Center, Orthopaedic Surgery, Plano, TX, United States
(6) University of California, Neurosurgery, San Francisco, CA, United States

Abstract

Introduction: The Neck Disability Index (NDI), a 10-item survey for assessing neck pain, is derived from the Oswestry Disability Index (ODI). Neck and back symptoms are often concurrent in spine patients. The purpose of this study was to assess the correlations between similar individual NDI and ODI questions, and to examine which questions drive total score among patients with cervical, lumbar and mixed complaints.

Methods: Retrospective chart review of consecutive outpatient spine subjects completing both NDI and ODI. Patient complaints were reviewed and categorized: isolated neck pain (NP), isolated back pain (BP) or both (NB). Correlations between individual survey items were assessed, and a stepwise regression was performed to predict total score of each survey from its components. Regression models were then tested on another center's ODI and NDI data using a different group of patients.

Results: There were a total of 74 subjects (mean age 50.3 years; 56% female) with 39 BP (52.7%), 10 NP (13.5%) and 25 NB (33.8%). For all patients, the mean NDI and ODI scores were 37.8 and 40.8 respectively. For the subgroups, NDI/ODI were NP 41.0/35.9, BP 34.2/41.2 and NB 42.3/42.2.

The highest NDI and ODI item correlations were between “Personal Care” (r=0.89, p< 0.01); “Lifting” (r=0.85, p< 0.01); NDI “Recreation”/ODI “Social Life” (r=0.75, p< 0.01); NDI “Recreation”/ODI “Traveling” (r=0.70, p< 0.01); and “Sleeping” (r=0.70, p< 0.01).

The stepwise regression for predicting NDI total score using 5 items (“Recreation,” “Concentration,” “Driving,” “Lifting” and “Work”) produced R2=0.96 and using 8 items yielded R2=0.99. The analysis for ODI total score generated a 5-item (“Traveling,” “Lifting,” “Sleeping,” “Standing” and Employment/Housemaking”) model with R2=0.97 and an 8-item model with R2=0.99.

The stepwise models were tested using another center's NDI and ODI data from 45 patients. The NDI 5-item/8-item regressions generated R2=0.89/0.96. The mean absolute difference from actual NDI was 5.1 (range 0.2-14.0) for the 5-item regression and 3.0 (range 0.01-8.9) for the 8-item. The ODI 5-item/8-item regressions produced R2 = 0.91/0.96, with a mean absolute difference of 4.5 (range 0.5-11.7) for the 5-item model and 2.7 (range 0.1-8.6) for the 8-item.

NDI total score was predicted for the other center's patients using the 8-item NDI model but substituting 3 ODI domains (“Personal Care,” “Lifting” and “Sleeping”) for their NDI equivalents. This analysis obtained an R2=0.92 with a mean absolute difference from the actual NDI score of 4.0 (range 0.1-10.9).

Conclusion: While neck and back pain can be concurrent, elevated NDI and ODI scores in subjects with isolated regional complaints suggest that disability in one region may affect clinical scores in the adjacent one. There are high single-item correlations between the NDI and ODI surveys, and total NDI and ODI score can be accurately predicted using fewer than 10 items from the questionnaires. This analysis suggests that it is feasible to develop a combined neck and back total disability questionnaire, reducing the burden to patients without sacrificing the quality of functional assessment.