Complications of chiropractic spinal manipulation

Article section 1 of 3.  Next

By Douglas J Lanska MD MS MSPH

Chiropractic spinal manipulation has been associated with multiple serious neurologic sequelae (Powell et al 1993; Lee et al 1995; Assendelft et al 1996; Hurwitz et al 1996), including brainstem and cerebellar stroke, Wallenberg syndrome, locked-in syndrome (Pratt-Thomas and Berger 1947; Schwartz et al 1956; Mueller and Sahs 1976; Beatty 1977; Krueger and Okazaki 1980; Schellhas et al 1980; Frumkin and Baloh 1990; Shafrir and Kaufman 1992; Powell et al 1993; Hurwitz et al 1996; Klougart et al 1996; Haldeman et al 1999), cervical spinal cord injury (Rinsky et al 1976; Davis 1985; Powell et al 1993; Padua et al 1996), thoracic intervertebral disc herniation with partial Brown-Sequard syndrome (Lanska et al 1987), cauda equina syndrome (Hooper 1973; Malmivaara and Pohjola 1982; Gallinaro and Cartesegna 1983; Powell et al 1993), radiculopathies (Gallinaro and Cartesegna 1983; Padua et al 1996), brachial plexopathies (Deshayes and Geffroy 1962), and phrenic nerve injury (Tolge et al 1993). Published reports of neurologic complications following chiropractic spinal manipulation have been predominantly individual case reports and small case series. Other reported complications include dislocation of the atlas, vertebral fracture, vertebral artery pseudoaneurysm, spinal epidural hematoma, tracheal rupture, cardiac arrest, and various soft tissue injuries (Patmas 1993; Hurwitz et al 1996).

Physicians and chiropractors need to be aware of the neurologic complications of chiropractic spinal manipulation, as limitation of morbidity and mortality may depend on prompt treatment (Lanska et al 1987).

In a survey of California neurologists who encountered patients with complications in the first 24 hours following chiropractic spinal manipulation, 102 patients with complications were reported over the prior 2 years, with 56 cases of stroke, 30 of radiculopathy, and 16 of myelopathy - most related to cervical manipulations (Lee et al 1995). At 3 months after onset of neurologic symptoms, 45 (44%) were left with moderate or severe clinical deficits, and 1 patient had died (of a chiropraxis-related stroke). Strokes were generally in the vertebral artery territory in association with vertebral artery dissections (95%) and only rarely (5%) in the carotid distribution. Approximately three fourths (73%) of radiculopathies occurred in the cervical region, and the remainder occurred in the lumbosacral region. More than four fifths (81%) of myelopathies occurred in the cervical region, with only 1 case in the thoracic region and 2 cases in the lumbosacral region (cauda equina syndrome). This exploratory study had a number of limitations that the authors acknowledged.

In an analysis of 141 complications of spinal manipulation therapy for relief of lumbar or cervical back pain, Powell and colleagues identified 6 "risk factors" associated with such complications (although this appears to have been based on simple tallies, rather than factors statistically associated with such complications in cases compared with controls): misdiagnosis, failure to recognize the onset or progression of neurologic signs or symptoms, improper technique, predisposing illness (eg, coagulation disorder or herniated nucleus pulposus), and manipulation of the cervical spine (Powell et al 1993). The authors concluded that the risk/benefit ratio of spinal manipulation therapy was "unacceptably high for patients with radicular symptoms or signs associated with prolapsed discs and neck pain."

In a systematic review of adverse events associated with pediatric spinal manipulation, the authors identified 9 cases with serious direct adverse events involving neurologic or musculoskeletal systems (eg, subarachnoid hemorrhage, paraplegia) and 20 cases of indirect adverse effects involving delayed diagnosis of serious conditions (eg, diabetes, neuroblastoma) and/or inappropriate use of spinal manipulation for serious medical conditions that would have no potential benefit and could delay needed therapy (eg, meningitis, rhabdomyosarcoma) (Vohra et al 2007).

The vast majority of strokes reported in the literature related to chiropractic spinal manipulation have been in the vertebral territory associated with vertebral artery dissections (Terrett 1987; 1990; Hurwitz et al 1996; Smith et al 2003; Williams and Biller 2003). The vertebral artery is most commonly injured during manipulation at the level of the atlantoaxial joint where the vertebral artery changes direction from a vertical to a horizontal orientation. Rotating and extending the neck stretches the extracranial vertebral arteries at the level of the atlantoaxial joint and produces a shearing force that can produce intimal tearing and in some cases dissection, thrombus formation, and artery-to-artery embolism (Lee et al 1995; Assendelft et al 1996; Rothwell et al 2001).

Pre-manipulation assessment of vertebrobasilar stroke risk is not feasible or likely to be useful for most patients undergoing chiropractic spinal manipulation, but practitioners performing such manipulation should at least be familiar with signs and symptoms of vertebrobasilar ischemia; unfortunately a RAND review headed by a chiropractor found that "in many cases, the manipulator failed to cease treatment even after signs and symptoms of vertebrobasilar ischemia (eg, dizziness, vertigo, nausea, loss of consciousness)" (Hurwitz et al 1996, p. 1753).

Because vertebrobasilar strokes following chiropractic spinal manipulation are rare, they are difficult to study with epidemiological methods and certainly prospective cohort studies would not be feasible; therefore, most reports have necessarily been case reports and small case series. The most realistic approach to obtaining better data on this problem is a case-control study utilizing existing administrative medical records. With that in mind, Rothwell and colleagues conducted a population-based case-control study using hospitalization records to identify vertebrobasilar territory strokes in the period from 1993 to 1998 in Ontario, Canada (Rothwell et al 2001). Cases were matched by age and gender to 4 controls with no history of stroke at the date of the case event. Public health insurance billing records were used to establish the use of chiropractic services before the event date. The authors found that for those aged less than 45 years, vertebrobasilar stroke cases were 5 times more likely than controls to have undergone chiropractic treatment in the week prior to stroke onset. In addition, for those aged less than 45 years, cases were 5 times more likely to have had at least 3 chiropractic visits with a cervical diagnosis in the month prior to stroke onset. No significant associations were identified in those at least 45 years of age. Still, even using administrative records, the number of identified cases with cervical manipulation was very small: of the nearly 600 cases of vertebrobasilar stroke, only 9 had cervical manipulation in the week prior to their stroke, and limiting the analysis to young cases (less than 45 years of age) reduced the number of cases of vertebrobasilar stroke by more than 80%, leaving only 6 with cervical manipulation in the week prior to their stroke.

In an attempt at synthesizing the outcome of vertebrobasilar strokes related to chiropractic spinal manipulation reported in the literature, Assendelft and colleagues reviewed published case reports through 1993 and found 165 cases of vertebrobasilar complications attributed to chiropractic spinal manipulation, of which 27% recovered, 52% had significant residual deficits, and 18% died (Assendelft et al 1996). Similarly, Hurwitz and colleagues reported 21 deaths among 118 cases (18%) in the English-language literature, and 52 (44%) survived with "serious neurologic deficit, paralysis, or other permanent functional impairment" (Hurwitz et al 1996). Terrett also reported 28 deaths of 113 cases (25%) in the English- and non-English-language literature from 1934 to 1987 (Terrett 1990).

Smith and colleagues used a nested case-control design to determine whether spinal manipulation is an independent risk factor for cervical artery dissection (Smith et al 2003). The authors reviewed all patients under age 60 with cervical artery dissection and ischemic stroke or transient ischemic attack in a prospective stroke registry at 2 academic medical centers over a 6-year period, ultimately identifying 51 cases with dissection and 100 controls matched by age and sex. In multivariate models, vertebral artery dissections were independently associated with spinal manipulation within the preceding 30 days (odds ratio 6.6) and with neck or head pain preceding stroke or transient ischemic attack (odds ratio 3.8).

Compared with injury to the vertebral artery, the extracranial carotid artery is less commonly damaged but can nevertheless be compressed against the transverse processes of the upper cervical vertebrae when the neck is rotated, resulting in similar complications as occur more commonly with vertebral artery injuries.

The torsional and extensional stress applied with chiropractic spinal manipulation may exceed the limits of extensibility of the annular disc fibers (especially with pre-existing disc degeneration) and bypass the normal neural reflex guards, resulting in symptomatic disc herniation and resultant radiculopathy, myelopathy, or cauda equina syndrome, depending on the level and location of disc herniation (Lanska et al 1987).

Most cases of neurologic injury related to chiropractic spinal manipulation have involved apparently normal patients (Hurwitz et al 1996), but some have involved patients with pre-existing disease that likely increased their risk of injury with chiropractic spinal manipulation. For example, Shafrir and Kaufman reported an infant with congenital torticollis caused by a spinal cord astrocytoma who developed quadriplegia after chiropractic manipulation (Shafrir and Kaufman 1992). Rinsky and colleagues reported a 44-year-old man with ankylosing spondylitis who developed a complete C4 quadriplegia after cervical chiropractic spinal manipulation (Rinsky et al 1976). Other reports also indicate that pre-existing vertebral pathology, including spinal canal stenosis and vertebral instability, predisposed patients to complications of cervical chiropractic spinal manipulation, including cervical radiculopathy and myelopathy (Schmidley and Koch 1984; Padua et al 1996). In patients with spinal canal stenosis and vertebral instability, forced rotary and extension maneuvers during cervical chiropractic spinal manipulation can result in direct concussion and ischemia of the cervical spinal cord (Schmidley and Koch 1984).