Comparing manual and mechanically assisted manipulations of the thoracic spine in neck pain patients: A pilot study

Study design

A pilot study was conducted to compare the effects of two different methods of manipulation (manual manipulation vs. electromechanical manipulation) on the thoracic spine for patients with neck pain. A complete study protocol for a randomized controlled trial was published in May 2015³⁰. The study received ethical approval from the Ethics Commission of the Canton of Zurich (2012-0248) and was registered at Current Controlled Trials (ISRCTN88585962). A completed CONSORT checklist can be found in Supplementary File 1.

Participants

Every participant received all relevant patient information in advance. When a potential participant had read the information and was willing to participate, he or she signed the informed consent form. Participants were recruited from, May 2013 until August 2015, in private physical therapy practices in Zurich, Switzerland. Inclusion criteria were the presence of acute or chronic neck pain grade I or II³¹, age 18 years or older, able to speak and read German or English, no previous manual therapy of the thoracic spine, and interest in participating in the study.

Patients were excluded if they had severe disorders of the cervical spine, such as disk prolapse, spinal stenosis, postoperative conditions in neck and shoulder areas, a history of severe trauma, spasmodic torticollis, frequent migraine headaches, peripheral nerve entrapment, fibromyalgia, shoulder diseases (causing reduced mobility of the joint, e.g., fractures, adhesive capsulitis), inflammatory rheumatic diseases, osteoporosis, or cancer or if they were currently undergoing legal procedures related to their neck pain.

Additional evaluations (history taking and physical examination) of the participant were added after baseline measurements to confirm eligibility of the participant to be included in the study population³⁰.

Randomization

A clinician not involved in the study randomized the participants independently into one of the two groups. Block randomization (20 blocks of 4) was carried out using a computer-based randomization program. The original list was stored in an opaque and sealed envelope, not accessible to the therapist delivering the manipulations. To reduce intervention bias, 80 individually numbered opaque and sealed envelopes were prepared, and the therapist received the assigned envelope immediately before each participant was to receive the manipulation. Additionally, after the participants completed the questionnaires, they were put into an opaque envelope that was individually and immediately sealed by the patient. The therapist was unaware of the outcome. The anonymous data were entered into the database and were untraceable to the participating patient.

Intervention

Each patient underwent three treatment sessions, with 4 days interval between each session. Additionally, during the first session, all participants were instructed to participate in an exercise program, which was to be completed as soon as possible. To evaluate compliance of the participants, they received a journal and were asked to mark the days that they had completed the required exercises. Follow-up assessments were conducted at 6 weeks, 3 months, and 6 months after the initial treatment. Patients received the questionnaires by mail, including prepaid envelopes, to facilitate return of the completed forms. An extensive explanation of the treatments can be found elsewhere³⁰.

Manual manipulations

For the manually performed manipulations, the patient was supine. The therapist’s hand was placed under the thoracic spine using a pistol grip (the fingers positioned with the index finger straightened and fingers 3 to 5 flexed). The patient’s forearms were crossed in front of their chest. The force was applied against the therapist’s flexed fingers and the thenar eminence and slightly cranial to the transverse process of the caudal vertebrae of the segment to be treated³².

Electromechanical manipulations

For the electromechanical manipulations, the Impulse iQ® (Neuromechanical Innovations, Chandler, AZ, USA) was used. The patient was prone on a treatment table, with arms parallel to the body, in a relaxed position. Before the treatment began, the patient was instructed that they would hear a rattling sound that indicated the treatment thrusts conducted by the device and a beep to signify the end of treatment. The Impulse iQ was placed onto the vertebrae previously identified as the painful segment. A double stylus and middle force setting (peak force = 200 N), as recommended by the manufacturer for the treatment of the thoracic spine, was used³³. The device recorded and analyzed the spinal acceleration response each time a thrust was delivered by using the in-built firmware. To begin, the machine produced a series of repetitive thrusts while monitoring the acceleration response. If the response improved, the treatment thrusts continued up to 3 more seconds. If the acceleration response was negative (flat line or decrease), the thrusts delivered by the instrument ceased.

Exercise program

After the first treatment session, all patients were introduced to the home exercise program and asked to perform this program once a day for 6 weeks. Compliance with the home exercise program was assessed using a training journal as follows: (1) retraining the craniocervical muscles (e.g., training the holding capacity of the deep neck flexors and head lift), (2) retraining the scapular muscles, and (3) eye/head coordination^34,35. For a complete overview of the study`s course see Figure 1.

Figure 1. Flowchart of the trial, describing the process of randomization, evaluation and treatment.

Outcome measures

Baseline measures included age, weight, height, and the European Quality of Life 5 Dimensions 5 Levels (EQ-5D-5L). Furthermore, each patient’s expectation of treatment outcome (the treatment he or she expected to be more beneficial) was assessed at the beginning of the study. The primary outcome measure was the Visual Analogue Pain rating scale (VAS) that assesses pain intensity level^36–39. The VAS was completed before and after every treatment session and at all follow-up points. Secondary outcomes were the Neck Disability Index (NDI) and the Patients Global Impression of Change Scale (PGIC)^40–44.

The NDI is a self-reported questionnaire to evaluate disability. It consists of 10 items that score from zero to five. The higher the total score, the higher is the disability level of the patient⁴¹. The NDI’s minimal clinical important difference (MCID) in patients with neck pain ranges from 3.5 to 9.5, which mainly depends on the patient characteristics^45–47. Pereira et al. (2015) state that in patients with chronic neck pain, the score has to be above 5.5 points, otherwise it is clinically irrelevant^45,48,49. Before the second, third treatments and all follow-up assessment, the patients were asked about their global impression of change, using the PGIC⁵⁰. The PGIC is a valid outcome measure that is based on a seven-point Likert scale⁵¹. Its purpose is to obtain a patient’s report of improvement over time during treatment. The scale ranges from much better, better, somewhat better, no change, somewhat worse, worse, and much worse. Much better is rated as 7 and much worse as 1⁴⁴. It can be used to dichotomize participants into two groups: improved (ratings of 6 and 7) and not improved (ratings 1 to 5).

Statistics

The data were stored and analyzed using the IBM SPSS 21 statistical software package (SPSS, Inc., Chicago, IL). Descriptive statistics were used to describe patient characteristics. Collected data were checked for normal distribution using the Shapiro-Wilk Test. The test identified that the scores for VAS and NDI were normally distributed. Because of outliers, the Friedman test was used as a nonparametric statistical test to retain all data points. The goal was to determine whether there were any statistically significant differences between the distributions of three or more related groups. Furthermore, pairwise comparisons were performed (SPSS Statistics, 2012) with a Bonferroni correction for multiple comparisons.

Primary outcome

Both groups showed a significant improvement in pain scores (Friedman test, VAS Χ² (5) = 14.161, p = 0.015). There were no statistically significantly differences between the two treatment groups: electromechanical X²(5) = 7.701, p = 0.174 / manual X²(5) = 7.774, p = 0.169. Bonferroni correction post hoc analysis revealed statistical differences in the rating from follow-up measurements at 3 months (median = 6.00) from the first appointment (median = 26.00) (p = 0.008). VAS scores at the baseline reduced by 11% at second treatment, 27% at third treatment, 54% after 6 weeks, 77% after 3 months, and 42% after 6 months. The % change of pain is shown in Table 2.

Secondary outcome

Because of outliers, the Friedman test was used for analyzing the NDI. There was no significant change (X² (5) = 10.562, p = 0.061). No significant change in NDI was observed between the two groups (electromechanical X² (5) = 5.234, p = 0.388 / manual group X² (5) = 5.546, p = 0.232). The change over time (median) is shown in Table 3.

PGIC

After the first treatment, four out of nine patients rated themselves as improved, five patients after the second treatment, three patients after 6 weeks, two patients after 3 months, and three patients after 6 months. All other participants stated that they had no improvement, but the pain did not get worse. In total, there were 10 ratings of improvement in the electromechanical manipulation group and 7 in the manual manipulation group (see Table 4 for a detailed rating).

Additional interventions

Overall, four participants had additional treatment during the 6 months of follow-up. After 3 months of treatment, two participants visited a physical therapist and one participant received acupuncture once a week. At the 6 months survey, another participant received chiropractic treatment (three times) and one continued to visit a physical therapist, if the neck pain was too severe.