Recruitment was by convenience sampling via referrals from local physical therapists and flyers posted on a university campus. Participants were assigned to either a transverse vertebral pressures mobilization group or a control group. Inclusion criteria for the mobilization group were 1) ages 18 to 60, and 2) mechanical neck pain defined as pain exacerbated by movement. Among the exclusion criteria were unexplained night pain, unexplained weight loss, numbness and tingling in the arms and/or legs, balance or coordination problems, morning stiffness lasting greater than one hour, a history of whiplash injury within six weeks of treatment, diagnosis of osteoporosis, symptoms suggestive of nerve root compression, prior cervical or thoracic spine surgery, or a pending legal action. Inclusion criteria for the control group were 1) ages 18 to 60, and 2) asymptomatic neck and thoracic regions over the previous six years by subject report. The latter group was labeled "control" in the sense that it allowed blinding of the rater rather than in the sense of a sham treatment group. The project was approved by the University of Puget Sound Institutional Review Board, and written informed consent was obtained from all the subjects.
Outcome Measures Outcomes of interest were active neck motion, pressure pain threshold (PPT) and report of pain. For active range of motion, a single bubble inclinometer (MIE Inclinometer, Medical Research Ltd, London, UK) was used to measure flexion, extension, and side-flexion with the subjects seated. Since measurement of rotation by inclinometer would introduce a change from the sitting position, a universal goniometer was chosen to measure rotation. De Koning et al conducted a systematic review of seated inclinometry for quantifying active cervical ROM. Interrater intraclass correlation coefficients (ICCs) for flexion, extension and lateral flexion varied from 0.81 to 0.94 for healthy subjects and from 0.68 to 0.86 for those with neck pain. Among patients with mechanical neck pain, ICCs for cervical rotation using the universal goniometer in sitting have been reported as 0.77 to the right and 0.57 to the left.12 In order to consider a motion to have changed, we required statistical significance at ≤ 0.01 and a difference in degrees of ≥5. Before commencing the study, one researcher participated in a training session followed by intra-rater reliability testing for all ROM measurements.
PPT, the minimum pressure that induces pain, was measured at the right lateral epicondyle using a hand-held digital algometer (Model FPX 25, Wagner Instruments, Greenwich CT.) Three measurements were taken separated by 10 second intervals, and the mean of these trials was used for analysis. Good to excellent reproducibility, intra-rater and inter-rater reliabilities for PPT have been reported.13-15 However, Prushansky et al reported a significant difference between raters which prompted the recommendation that PPT measurement should be performed by the same rater when utilized as an outcome measure. We followed this recommendation.
An 11-point numerical pain rating scale (NPRS) was used to assess pain before and after the intervention. Cleland et al reported the psychometric properties of the NPRS specific to neck pain in a cohort of 137 patients. Test-retest reliability was moderate (ICC=.76; CI, .51-.87). The minimum clinically important difference for neck pain in his study was a change of 1.3 points. In our study, we selected a requirement of a 1.5-point change for clinical importance.
Study Protocol After determining eligibility, baseline ROM and PPT scores were taken by one blinded rater, and subjects completed an NPRS. The mobilization group was treated with non-thrust transverse vertebral pressures as described by Maitland. The patient lay prone with arms to the side and in a "forehead rest position". Pressures were applied to spinal levels T1 through T4. The spinous process of T1 was identified by first locating C6 using the cervical extension method17 and then counting caudally. The researcher stood at the level of the vertebra to be mobilized on one side of the subject. The pad of the researcher's non-dominant thumb was placed in contact with the lateral aspect of the spinous process of T1, while the dominant thumb was placed on the dorsal side of the other thumb. Pressure was applied to the spinous process to produce small amplitude, low velocity oscillation into resistance to the end-range of the vertebra (Grades IV to IV+). This procedure was performed for 30 seconds, then sequentially applied to the next caudal level through T4. The same pattern of application was used on the subject's contralateral side. The entire procedure was repeated once again for a total of 8 minutes. This was followed by a repeat of the outcome measurements. The control group received no treatment. Each control subject assumed the same prone position for 8 minutes as was used with the mobilization group and outcome measurements were repeated. The researcher performing the outcome measurements was blinded to group assignment. All subjects were clothed such that no localized residual evidence of thoracic manual contact was visible.
Statistical Analyses Data were analyzed with SPSS, Version 14.0 (SPSS, Inc, Chicago IL). We compared pre-intervention data between the transverse mobilization group and the control group utilizing independent t-tests for ROM and PPT, and the Mann-Whitney test for NPRS scores. We compared pre and post session measurements within each group utilizing paired t-tests for ROM and PPT, and the Wilcoxon signed-ranks test for NPRS scores.
