Validation of a Clinical Prediction Rule to Identify Patients With Neck Pain Likely to Benefit From Thoracic Spine Thrust Mobilization: A Randomized Clinical Trial

Background: Recently a clinical prediction rule (CPR) has been developed that identifies patients with neck pain who are likely to respond rapidly and dramatically to thoracic spine thrust manipulation and an active range of motion exercise. Although the initial predictor variables identified during the development of a thoracic spine manipulation CPR seem to have adequate face validity, there is no guarantee that these factors will persist in a different group of patients, even ones with similar characteristics as those used in the initial exploratory study.

Purpose: The purpose of this follow-up study in which patients will be randomly assigned to receive thoracic spine thrust manipulation followed by therapeutic exercises or therapeutic exercise alone will be to investigate the validity of the previously developed CPR. If the CPR is indeed meaningful, patients who are positive on the CPR and receive thoracic spine thrust manipulation should experience improved outcomes compared to patients who are negative on the CPR and receive thoracic spine manipulation, and compared to patients who are positive on the CPR but receive the intervention believed to be effective for another subgroup of patients with neck pain.

Design: We will conduct a multi-center randomized clinical trial to assess the effectiveness of a previously-developed CPR for identifying patients with neck pain likely to respond to a thoracic spine thrust manipulation. Subjects between the ages of 18 and 60 with a primary report of neck pain will be randomized to receive manipulation plus an exercise program or to receive an exercise program only.

Methods: One hundred and forty patients with a primary complaint of neck pain who meet the inclusion/exclusion criteria and consent to participate will be enrolled in the study. We based sample size calculations on detecting a statistically significant difference between any of the 4 cells of the study which include the patients status on the rule (patients that meet the CPR and patients that do not meet the CPR) and treatment group (manipulation and exercise) by using the 1-week Neck Disability Index (NDI) score at an alpha-level of 0.05. To detect a 10 point change in NDI, with an expected standard deviation of 12 points at the 1 week follow up (effect size 0.80) with 90% power using a two-tailed hypothesis and assuming a 50% distribution of patients who do and do not meet the rule, 31 patients per cell are required. We will recruit 140 subjects into the study to control for drop-outs prior to the 1 and 4 week follow-up and possible distribution discrepancies between the classifications. The primary outcome point is four-weeks after randomization, with change in disability as measured by the NDI, serving as the primary outcome measure.

Data Analysis: We will examine the primary aim with a 3-way repeated-measures analysis of variance (ANOVA) with treatment group (manipulation vs. exercise) and status on the rule (positive or negative) as the between subjects independent variables and time (baseline, 1 week and 4 weeks) as the within-subjects independent variable. The dependent variable will be disability (NDI score). The hypothesis of interest is the 3-way group * CPR status * time interaction. We will perform planned pairwise comparisons at each follow-up period by using the Bonferroni inequality.