J Manipulative Physiol Ther 2000 (Mar); 23 (3): 175–179 ~ FULL TEXT
Licht PB, Christensen HW, Hoilund-Carlsen PF
Department of Clinical Physiology and Nuclear Medicine,
Odense University Hospital,
BACKGROUND: Spinal manipulative therapy is used millions of times every year to relieve symptoms from biomechanic dysfunction of the cervical spine. Concern about cerebrovascular accidents after cervical manipulative therapy is common but rarely reported. Premanipulative tests of the vertebral artery are presumed to identify patients at risk but controversy exists about their usefulness.
OBJECTIVE: The aim of this study was to examine vertebral artery blood flow in patients with a positive premanipulative test for contraindication to spinal manipulative therapy and to investigate if chiropractors would reconsider treating such patients if dynamic vascular Doppler examination was normal.
DESIGN AND SETTING: A prospective study at a university hospital vascular laboratory.
METHODS: Chiropractors in private practice from 3 Danish counties referred patients with a positive premanipulative test for an examination of vertebral artery blood flow. Premanipulative testing was performed by an experienced chiropractor. Flow velocities were measured in both vertebral arteries by color duplex sonography. In addition, chiropractors were asked if they would treat their patient despite a positive premanipulative test if the vascular ultrasound examination was normal.
RESULTS: A total of 20 consecutive patients with a positive premanipulative test were referred. Five were excluded because symptoms could not be reproduced during the vascular examination. In the remaining patients, no significant difference in peak flow velocity or time-averaged mean flow velocity with different head positions was found. Nineteen of 21 chiropractors would treat a patient with a positive premanipulative test if the vascular examination was normal. Eight of the patients with a positive manipulative test were treated without complications. Six are now symptom-free, and 2 have improved symptoms. The remaining 8 patients refused manipulation and continue to have the same symptoms.
CONCLUSION: It appears that a positive premanipulative test is not an absolute contraindication to manipulation of the cervical spine. If the test is able to identify patients at risk for cerebrovascular accidents, we suggest patients with a reproducible positive test should be referred for a duplex examination of the vertebral artery flow. If duplex flow is normal, the patient should be eligible for cervical manipulation despite the positive premanipulative test.
From the Full-Text Article:
Concern regarding cerebrovascular accidents after cervical manipulative therapy is common. According to a recent editorial, cervical manipulation arouses far more concern about safety than the use of lumbar manipulation.  Fortunately, serious complications are rare, whereas minor symptoms, such as local discomfort or headache, are quite frequent.  To identify subjects at risk, manual therapists examine their patients thoroughly before cervical manipulation and usually include functional testing of the vertebral arteries. Maitland is often credited with the introduction of such a test in 1968.  However, a test was described 6 years earlier by Smith and Eldridge.  A number of different tests have been proposed: Barré-Leiou's sign test, George's cerebrovascular craniocervical functional test, Maigne's test, Hautant's test, Underberg's test, Hallpike maneuver, and deKleyn's or Wallenberg's tests. [11, 12] All are variations of the same theme: extreme rotation and extension of the head designed to provoke cerebral ischemia during positional change of the cervical spine. It is believed that the contralateral vertebral artery stretches during rotation because it is relatively fixed at the transverse foramina and blood flow subsequently ceases.  A test is positive if it provokes signs or symptoms of vertebrobasilar insufficiency (eg, nystagmus or symptoms of vertigo, dizziness, tinnitus, visual blurring, nausea, or faintness).  A positive test is considered a contraindication to cervical manipulation. One of the most commonly used tests is deKleyn's test,  described previously.
Several clinical papers have questioned the sensitivity and specificity of these premanipulative tests. [9, 11, 14-19] Doppler ultrasound has been used to investigate vertebral artery blood flow during head rotation in a number of studies and with varying results.  Few have dealt with premanipulative testing. [16, 17, 26] They all used a calculated index of vascular impedence, which did not change with head position, instead of measuring flow velocities directly. It is generally believed that blood flow in the contralateral vertebral artery ceases completely during rotation,  in which case it makes no sense to calculate sensitivity or specificity in this study. There was no significant difference in either peak flow velocity or the time-averaged mean flow velocity in the various head positions. No patient had a complete cessation of flow. Of course, this does not eliminate a complete positional vertebral artery obstruction. A vertebral artery obstruction, which has been demonstrated in several angiographic case reports of patients with positional vertebrobasilar insufficiency, may cause a positive test. However, these results suggest that a positive test is seldom associated with changes in vertebral artery blood flow. Consequently, deKleyn's test is neither sensitive nor specific. Furthermore, in 5 of 20 patients referred because of positive test results, the symptoms could not be reproduced after 1 month.
These results seem to indicate that a positive deKleyn's test should not always be considered an absolute contraindication to cervical manipulation. The literature appears to indicate that a test may be negative in the prescence of vertebral artery occlusion [15, 18] and that vascular accidents may occur despite a negative test. [20, 21] In addition, considering that the incidence of cerebrovascular accidents with cervical manipulation is extremely low, why does this test have a role in identifying patients at risk? We suggest that a positive test should lead to reexamination after 1 or 2 months and, if reproducible, a duplex ultrasound evaluation of vertebral artery flow during premanipulative testing should be considered. The results of the questionnaire (Table 1) indicate that most chiropractors would treat patients, despite a positive premanipulative test, if only the vascular examination is normal. Eight of the patients with a positive deKleyn's test were treated without complications. Six are now symptom-free, and 2 have improved. The remaining 8 patients who refused manipulation still have the original symptoms.
The incidence of a positive premanipulative test is not known. The 21 chiropractors who answered our questionnaire indicated that they see approximately 2 such patients within a 1-year period. Chiropractors in private practice in Denmark perform an average of 6000 spinal manipulative treatments every year; approximately one third of these are cervical manipulations.  Assuming that premanipulative testing is performed before each treatment, a positive test rate of 1‰ can be estimated. Consequently, with approximately 250 million spinal manipulations performed annually in the United States alone,  approximately 83,000 patients do not receive cervical manipulation each year as a result of a positive premanipulative test.
Of course, other pathologic conditions could be responsible for the symptoms evoked by a premanipulative test. Neck proprioceptive afferent nerves, which are involved in postural adjustment, are closely related to vestibular function. The origins of these neck afferents are the joint receptors of the first 3 cervical vertebrae, especially the atlantoaxial joint and atlanto-occipital joints.  Previous research has suggested that a positive premanipulative test could arise from this proprioceptive dysfunction,  which could result in a disturbance of the upper cervical sensory nerves causing “cervical vertigo” by way of their projections to the vestibular nucleus. [28, 29] This presents a therapeutic dilemma—the very symptoms that give contraindication to therapy would most likely be relieved by the treatment they exclude. However, there are reports of a favorable outcome after cervical manipulation in these patients. [28, 30, 31] In contrast to the untreated patients, the treated patients in this study achieved relief, which seems to favor an explanation of this kind and suggests that a positive deKleyn's test should generally not be considered a contraindication to manipulative therapy.
It appears that a positive premanipulative test is not an absolute contraindication to manipulation of the cervical spine. If the premanipulative test is able to identify patients at risk for cerebrovascular accidents, we suggest that patients with a reproducible positive test should be referred for a duplex examination of the vertebral artery flow. If duplex flow is normal, the patient should be eligible for cervical manipulation, despite the positive premanipulative test.