Research Review By Dr. Shawn Thistle©

Date Posted:

April 2009

Study Title:

Development of a clinical prediction rule to identify patients with neck pain likely to benefit from cervical traction and exercise


Raney NH et al.

Author's Affiliations:

Physical Therapy Departments – Wilford Hall Medical Center, Brook Army Medical Center (both in San Antonio, TX), Army Physical Fitness Research Institute (El Paso, TX), Naval Special Warfare Center (Coronado, CA)

Publication Information:

European Spine Journal 2009; 18: 382-391.

Background Information:

Neck pain is not as common as low back pain, but still has a substantial lifetime incidence that has been estimated as high as 54%. About one third of neck pain patients experience persistent pain longer than three months, representing a significant health care concern. Manual therapists of all backgrounds see neck pain patients frequently, so possessing a working knowledge of the literature and latest assessment tools is crucial in order to stay competitive and provide high level care.

Clinical Prediction Rules (CPR) are tools designed to assist clinical decision making – using historical and physical examination findings, and combinations thereof, to direct patient management strategies. The aim of this study was to develop a CPR to properly identify patients with neck pain that are most likely to respond well to cervical traction. Study methods will be discussed next to facilitate better understanding of the results.

Study Methods:

This study included a prospective cohort of neck pain patients who were 18 years of age or older. They had a primary complaint of neck pain with or without arm symptoms, and a baseline Neck Disability Index score of at least 20. Patients were excluded if they:
  • had any red flags suggestive of serious pathology of non-MSK origin
  • were pregnant
  • had evidence of vascular compromise
  • had CNS involvement or multiple level neurological impairments
After providing demographic information and signing consent forms, all patients completed a variety of self-report measures followed by a standardized physical examination performed by one of five physical therapists. The self-report measures utilized included:
  1. a body pain diagram – measured location
  2. Numeric Pain Rating Scale (NPRS) – measured pain severity
  3. Neck Disability Index (NDI) – measure patients’ perceived disability
  4. Fear Avoidance Beliefs Questionnaire (FABQ) – measured patient beliefs on the influence of work and activity on their neck pain
The history and physical examination was standard, and included all pertinent components readers will be familiar with.

All patients received the same treatment regardless of the results of the history/physical examination. The treatment included a combination of mechanical neck traction and active exercise 2-3 times per week for a total of 6 treatments in a three week period. Patients were advised to maintain normal activity.

Cervical Traction:
  • Units used: the Chattanooga Trion Traction Table and Saunders 3D Active Trac Table (both were calibrated before the study)
  • each session was 15 minutes in length and performed with the patient supine – the angle of pull was set to 24° of flexion if the patient had full ROM, otherwise it was set to 15°
  • duty cycle was set to 60s of pull force (initially set to 10-12 pounds) with 20s rest (during this time pull force was 50% of duty cycle force)
  • force was increased with patient tolerance with the instruction that it should fee “moderately strong to strong” to a maximum of 40 pounds of pull
Exercise Intervention:
Patients were instructed on two exercises following their traction treatments
  1. Seated Posture Exercise: with the lumbar spine in neutral position, patients retracted their scapulae and gently elongated their cervical spines – they held this position for at least 10 seconds and were to perform it at least 2x/hour
  2. Supine DNF (Deep Neck Flexor) Strengthening Exercise: patients were supine, and instructed to perform a controlled craniocervical flexion without recruiting the larger superficial muscles – this was held for 10 seconds and performed 10x, twice daily
Treatment success was measured with a Global Rate of Change (GROC) scale ranging from -7 (a very great deal worse) to +7 (a very great deal better) – this method was selected over a region specific questionnaire such as the NDI to capture the patients’ perceived recovery. The GROC is a validated method for detecting clinically important changes.

Pertinent Results:

  • 68 patients (38 female) were included in the analysis
  • 44% of patients were categorized as having a successful treatment outcome (greater than or equal to a +6 on the GROC – “A great deal better’)
  • this successful outcome group experienced significantly greater (p < 0.001) reductions in pain (NPRS change of 2.2 average), disability (NDI score reductions of 12.5 average) compared to the non-successful group
5 variables were retained from a possible 15 in the final regression analysis:
  1. patient reported peripheralization with lower cervical mobility testing (C4-C7)
  2. positive shoulder abduction test
  3. age ? 55
  4. positive upper limb tension test A (ULTT A)
  5. positive neck distraction test
  • pre-test probability of successful outcome was 44% (30/68 patients)
  • having at least 3/5 of the above variables resulted in a positive likelihood ratio (LR) of 4.81 for successful outcome with traction and exercise (raising the probability of success from 44 to 78%)
  • if at least 4 variables were present, the positive LR was 23.1 resulting in an increased probability of success of 94.8% (no patients fulfilling at least 4/5 criteria had a poor outcome)
  • based on the 2 points above, the authors suggest that having 3/5 criteria seems to be the optimal threshold for decision making

Clinical Application & Conclusions:

Initial proposals for CPRs are often judged on the biological plausibility of their variables. The model that emerged in this study satisfies this criterion, as 4/5 variables appear to have a relationship to nerve root compression. This supports the notion that patients presenting with signs of nerve root compression may represent a subgroup of patients more likely to respond to cervical traction. Peripheralization of pain with mobility testing and a positive ULTT A may indicate adverse mechanical compression on neural tissue, while relief of symptoms with neck distraction and shoulder abduction tests may indicate relief of neural compression.

This is the first step in the development of a CPR to help us identify neck pain patients who are likely to benefit from cervical traction. Further studies must replicate these findings in larger and more varied patient samples, and perhaps fine tune the variables included in the rule. Further, the authors appropriately note that the exact dose and duration of traction treatments for a subgroup of neck pain patients still needs to be further studied. For the time being, prudent clinicians should bear in mind the five variables that emerged in this trial and apply them on a trial basis with neck pain patients to guide clinical decision making.

Study Strengths / Weaknesses:

One glaring limitation of this study is that no control group was utilized. Therefore, it cannot be stated with certainty that patients who improved did so due to the interventions rather than passage of time. The authors note that the variables that emerged in this study are not normally associated with a positive prognosis, so the influence of this factor could be considered minimal. The authors used a high threshold for a successful outcome, which further supports their notion that the positive outcomes were not attributable to natural history.