Research Review by Dr. Michael Haneline©


Apr. 2008

Study Title:

Predictors of adverse events following chiropractic care for patients with neck pain


Rubinstein SM et al.

Authors’ Affiliations:
Institute for Research in Extramural Medicine (EMGO Institute), VU University Medical Center, Amsterdam, The Netherlands.

Publication Information:

Journal of Manipulative and Physiological Therapeutics 2008; 31(2): 94-103.


There is a fair amount of evidence that suggests manipulation is an effective therapy for neck pain, especially when combined with exercise. Adverse events have been reported following cervical spine manipulation (CSM), although their precise rate of occurrence and relative severity are not well documented. There are numerous reports of stroke occurring very rarely following CSM, but even that relationship is disputed as far as the true rate, and whether the association is causal.

The objective of this report was to identify clinical and/or socio-demographic factors that might be predictive of adverse events after chiropractic treatment of patients with neck pain.

This was a practice-based study involving patients recruited from the practices of volunteer chiropractors throughout the Netherlands. Members of the Netherlands Chiropractors’ Association were invited to take part in the study via mailed flyers and personal contacts. Although 189 chiropractors were invited to participate, only 79 of them actually accepted the invitation.

Patients were included if they had neck pain, defined as cervicothoracic and/or periscapular pain, of any duration. Patients were excluded if they had received chiropractic care or manual therapy in the prior 3 months; if they had a red flag condition (e.g., suspected infection, fracture, tumor, metastasis, or intravenous drug use); or if the had any other condition thought to be a contraindication for cervical spine manipulation.

Self-administered questionnaires were completed by the patients at baseline and prior to treatment on their second and fourth visits. Questionnaires administered at the second visit asked about changes that occurred following the first visit, while those administered at the fourth visit asked about changes after the second or third visit.

Chiropractors also completed questionnaires regarding demographic information on themselves and their practices, as well as details of the care that was delivered on the patients’ first and third visits.

Both chiropractor and patient predictor variables were considered as possible prognostic factors of adverse events following CSM.

Chiropractor variables included:
  • the practitioner’s age, gender and experience level
  • chiropractic educational institution attended
  • type of manipulative technique used
  • the use of multiple attempts to manipulate the neck and the number of adjustments given
  • the use of neck rotation during manipulation
  • area of the cervical or thoracic spine that was treated
Patient variables included demographic factors, as well as:
  • the presence of kinesiophobia
  • disability, via the Neck Disability Index (NDI)
  • eleven-point numerical rating scales (NRS) to assess the expected effectiveness of treatment, fear of or apprehension concerning the treatment, self-reported general health, neck pain in the preceding 24 hours, and symptoms at baseline that might be reported as adverse events
An adverse event was defined as either a new related complaint that was not present at baseline or a worsening of the presenting complaint by >30%. Potential adverse events included:
  1. increased pain/stiffness at the treated area
  2. increased pain/stiffness in another treatment-related area
  3. headache
  4. tiredness/fatigue
  5. radiating pain in the arm or hand
  6. dizziness or lightheadedness
  7. nausea
  8. ringing in the ears
  9. confusion or disorientation
  10. depression or fear
  11. any other not specified reaction
Adverse events that resulted in an intensity score >8 on the 11-point NRS were considered “intense”. The authors pointed out the difference between “intense” and “serious” adverse events. The “serious” variety of adverse events denotes events that cause death, life-threatening situations, the need for hospital admission, or disability.

The type of treatment provided was left to the discretion of each chiropractor.

Pertinent results of this study:
  • The 79 participating chiropractors were able to collect data on 529 patients.
  • 96% and 87% of the study population returned for a second and fourth visit, respectively.
  • At baseline, many of the patients complained of headache, tiredness or dizziness, while few of them complained of nausea, depression, confusion or ringing in the ears.
  • A manipulative technique (most commonly Diversified) was used at any of the first 3 treatments.
  • Multiple manipulative attempts were made at the same segment in approximately one-fifth of the treatments.
  • The chiropractors reported that they used rotation at either the first or third visit slightly more than half of the time.
  • 46% of the study population reported at least one adverse event following the first visit and 22% following the second or third visit, with only 14% and 15%, respectively, of these adverse events being high in intensity.
  • Adverse events were typically musculoskeletal or pain related, while nausea, ringing in the ears, or psychological symptoms were relatively uncommon, affecting less than < 8% of the patients.
  • An intense adverse event at any of the first 3 treatments was reported by 13% of the patients, although none of these patients were worse or much worse at follow-up 12 months later.
  • No serious neurologic complications were reported.
  • More than 80% of those who reported adverse events considered them to have no to minor influence on their activities of daily living.
  • Only the following 3 variables were considered to be predictive of adverse events: 1) the reported use of rotation by the chiropractor, 2) working status of the patient, and 3) a longer duration with neck pain in the preceding year.
  • Patients who had visited a general practitioner in the 6 months prior to their first chiropractic visit were less likely to have an adverse event. Thus, this variable was considered to be protective.
  • Adjusted odds ratios, 95% confidence intervals, and levels of significance for some of these variables were as follows: rotation - 2.07; 1.18-3.66; (.01)…working status – 2.88; 0.87-9.47; (.08)…visited a general practitioner - 0.59; 0.32-1.09; (.09)
  • Increased neck pain following the first visit could be predicted using the following explanatory variables: duration with neck pain, pattern of pain in the preceding year, headache, and neck disability at baseline.
  • Previous studies had reported chiropractor or treatment variables, such as years of experience or type of technique used, to be predictive of adverse events following CSM and the authors expected their findings to concur. Likewise, they thought patient variables, such as gender, expected treatment effectiveness, or neck disability would likely be predictive. However, none of them were.
Study limitations included:
  • potential misclassification of the outcome variables
  • imprecision of the effect estimates
  • lack of a control group
  • potential patient recall bias, which was not considered likely since questionnaires were completed shortly after treatment
Study strengths included:
  • the use of multilevel modeling, which is designed to deal with clustering of patients
  • a large sample size
  • a high follow-up rate
  • the availability of baseline data on event-like symptoms, which allowed comparisons with adverse events.

Conclusions & Practical Application:

The authors suggested that the “the use of rotation should perhaps be modified upon initiating treatment.” This may make sense during the first few sessions of neck manipulation, after which the practitioner’s usual adjustments that include rotational components could be applied, if necessary.

This advice would be particularly true in a patient who was disabled from work and/or had a longer duration with neck pain in the preceding year.