Chiropractic' s Dirty Secret: Neck Manipulation and Strokes

Stephen Barrett

Stroke from chiropractic neck manipulation occurs when an artery to the brain ruptures or becomes blocked as a result of being stretched. The injury often results from extreme rotation in which the practitioner' s hands are placed on the patient's head in order to rotate the cervical spine by rotating the head. The vertebral artery, whch is shown in the picture to the right, is vulnerable because it winds around the topmost cervical vertebra (atlas) to enter the skull, so that any abrupt rotation may stretch the artery and tear its delicate lining. The anatomical problem is illustrated on page 7 of The Chiropractic Report, July 1999. A blood clot formed over the injured area may subsequently be dislodged and block a smaller artery that supplies the brain. Less frequently, the vessel may be blocked by blood that collects in the vessel wall at the site of the dissection.
Chiropractors would like you to believe that the incidence of stroke following neck manipulation is extremely small. Speculations exist that the odds of a serious complication due to neck manipulation are somewhere between one in 40,000 and one in 10 million manipulations. No one really knows, however, because (a) there has been little systematic study of its frequency; (b) the largest malpractice insurers won't reveal how many cases they know about; and (c) a large majority of cases that medical doctors see are not reported in scientific journals.

Published Reports
In 1992, researchers at the Stanford Stroke Center asked 486 California members of the American Academy of Neurology how many patients they had seen during the previous two years who had suffered a stroke within 24 hours of neck manipulation by a chiropractor. The survey was sponsored by the American Heart Association. A total of 177 neurologists reported treating 56 such patients, all of whom were between the ages of 21 and 60. One patient had died, and 48 were left with permanent neurologic deficits such as slurred speech, inability to arrange words properly, and vertigo (dizziness). The usual cause of the strokes was thought to be a tear between the inner and outer walls of the vertebral arteries, which caused the arterial walls to balloon and block the flow of blood to the brain. Three of the strokes involved tears of the carotid arteries. In 1991, according to circulation figures from Dynamic Chiropractic, California had about 19% of the chiropractors practicing in the United States, which suggests that about 147 cases of stroke each year were seen by neurologists nationwide. Of course, additional cases could have been seen by other doctors who did not respond to the survey.
A 1993 review concluded that potential complications and unknown benefits indicate that children should not undergo neck manipulation. Louis Sportelli, DC, NCMIC president and a former ACA board chairman contends that chiropractic neck manipulation is quite safe. In an 1994 interview reported by the Associated Press, he reacted to the American Heart Association study by saying, "I yawned at it. It's old news." He also said that other studies suggest that chiropractic neck manipulation results in a stroke somewhere between one in a million and one in three million cases.
The one-in-a-million figure could be correct if California's chiropractors had been averaging about 60 neck manipulations per week. Later that year, during a televised interview with "Inside Edition," Sportelli said the "worst-case scenario" was one in 500,000 but added: "When you weigh the procedure against any other procedure in the health-care industry, it is probably the lowest risk factor of anything." According to the program's narrator, Sportelli said that 90% of his patients receive neck manipulation.
In 1996, RAND issued a booklet that tabulated more than 100 published case reports and estimated that the number of strokes, cord compressions, fractures, and large blood clots was 1.46 per million neck manipulations. Even though this number appears small, it is significant because many of the manipulations chiropractors do should not be done. In addition, as the report itself noted, neither the number of manipulations performed nor the number of complications has been systematically studied. Since some people are more susceptible than others, it has also been argued that the incidence should be expressed as rate per patient rather than rate per adjustment. In 1996, the National Chiropractic Mutual Insurance Company (NCMIC), which is the largest American chiropractic malpractice insurer, published a report called "Vertebrobasilar Stroke Following Manipulation, " written by Allen G.J. Terrett, an Australian chiropractic educator/researcher . Terrett based his findings on 183 cases of vertebrobasilar strokes (VBS) reported between 1934 and 1994. He concluded that 105 of the manipulations had been administered by a chiropractor, 25 were done by a medical practitioner, 31 had been done by another type of practitioner, and that the practitioner type for the remaining 22 was not specified in the report. He concluded that VBS is "very rare," that current pretesting procedures are seldom able to predict susceptibility, and that in 25 cases serious injury might have been avoided if the practitioner had recognized that symptoms occurring after a manipulation indicated that further manipulations should not be done.
A 1999 review of 116 articles published between 1925 and 1997 found 177 cases of neck injury associated with neck manipulation, at least 60% of which was done by chiropractors.
In 2001, NCMIC published a second edition of Terrett's book, titled, "Current Concepts: Vertebrobasilar Complications following Spinal Manipulation, " which covered 255 cases published between 1934 and 1999. NCMIC's Web site claims that the book "includes an analysis of every known case related to this subject." That description is not true. It does not include many strokes that resulted in lawsuits against NCMIC policyholders but were not published in scientific journals. And it does not include the thoroughly documented case of Kristi Bedenbauer whose autopsy report I personally mailed to Terrett after speaking with him in 1995.
In 2001, Canadian researchers published a report about the relationships between chiropractic care and the incidence of vertebrovascular accidents (VBAs) due to vertebral artery dissection or blockage in Ontario, Canada, between 1993 and 1998. Using hospital records, each of 582 VBA cases was age- and sex-matched to four controls with no history of stroke. Health insurance billing records were used to document use of chiropractic services. The study found that VBA patients under age 45 were five times more likely than controls to (a) have visited a chiropractor within a week of the VBA and (b) to have had three or more visits with neck manipulations.
No relationship was found after age 45. The authors discuss possible shortcomings of the study and urge that further research be done. An accompanying editorial states that the data correspond to an incidence of 1.3 cases of vertebral artery dissection or blockage per 100,000 individuals receiving chiropractic neck manipulation, a number higher than most chiropractic estimates.
In 2001, British researchers reported on a survey in which all members of the Association of British Neurologists were asked to report cases referred to them of neurological complications occurring within 24 hours of neck manipulation over a 12-month period. The 35 reported cases included 7 strokes involving the vertebrobasilar artery and 2 strokes involving a carotid artery. None of the 35 cases were reported to medical journals. Edzard Ernst, professor of complementary medicine at the University of Exeter School of Sport and Health Sciences, believes that these results are very significant. In a recent commentary, he stated: One gets the impression that the risks of spinal manipulation are being played down, particularly by chiropractors. Perhaps the best indication that this is true are estimates of incidence rates based on assumptions, which are unproven at best and unrealistic at worse. One such assumption, for instance, is that 10% of actual complications will be reported. Our recent survey, however, demonstrated an underreporting rate of 100%. This extreme level of underreporting obviously renders estimates nonsensical. In 2002, researchers representing the Canadian Stroke Consortium reported on 98 cases in which external trauma ranging from "trivial" to "severe" was identified as the trigger of strokes caused by blood clots formed in arteries supplying the brain. Chiropractic- style neck manipulation was the apparent cause of 38 of the cases, 30 involving vertebral artery dissection and 8 involving carotid artery dissection. Other Canadian statistics indicate the incidence of ischemic strokes in people under 45 is about 750 a year. The researchers believe that their data indicate that 20% are due to neck manipulation, so there may be "gross underreporting" of chiropractic manipulation as a cause of stroke.
In 2003, another research team reviewed the records of 151 patients under age 60 with cervical arterial dissection and ischemic stroke or transient ischemic attack (TIA) from between 1995 and 2000 at two academic stroke centers. After an interview and a blinded chart review, 51 patients with dissection and 100 control patients were studied. Patients with dissection were more likely to have undergone spinal manipulation within 30 days (14% vs 3%). The authors concluded that spinal manipulation is associated with vertebral arterial dissection and that a significant increase in neck pain following neck manipulation warrants immediate medical evaluation.
In 2006, the Journal of Neurology published a German Vertebral Artery Dissection Study Group report about 36 patients who had experienced vertebral artery dissection associated with neck manipulation. Twenty-six patients developed their symptoms within 48 hours after a manipulation, including five patients who got symptoms at the time of manipulation and four who developed them within the next hour. In 27 patients, special imaging procedures confirmed that blood supply had decreased in the areas supplied by the vertebral arteries as suggested by the neurological examinations. In all but one of the 36 patients, the symptoms had not previously occurred and were clearly distinguishable from the complaints that led them to seek manipulative care. This report is highly significant but needs careful interpretation. Although it is titled "Vertebral dissections after chiropractic neck manipulation . . . " only four of the patients were actually manipulated by chiropractors. Half were treated by orthopedic surgeons, five by a physiotherapist, and the rest by a neurologist, general medical practitioner, or homeopath. It is possible-although unlikely-that the nonchiropractors used techniques that were more dangerous than chiropractors use in North America. The authors suggested that the orthopedists' treatment was safer, but there is no way to determine this from their data. Regardless, the study supports the assertion that neck manipulation can cause strokes-which many chiropractors deny.

Are Complications Predictable?
Although some chiropractors advocate "screening tests" with the hope of detecting individuals prone to stroke due to neck manipulation. These tests, which include holding the head and neck in positions of rotation to see whether the patient gets dizzy, are not reliable, partly because manipulation can rotate the neck further than can be done with the tests. Listening over the neck arteries with a stethoscope to detect a murmur, for example, has not been proven reliable, though patients that have one should be referred to a physician. Vascular function tests in which the patient's head is briefly held in the positions used during cervical manipulation are also not reliable as a screen for high-risk patients because a thrust that further stretches the vertebral artery could still damage the vessel wall." In a chapter in the leading chiropractic textbook, Terrett and a conclude have stated:
Even after performing the relevant case history, physical examination, and vertebrobasilar function tests, accidents may still occur. There is no conclusive, foolproof screening procedure to eliminate patients at risk. Most victims are young, without [bony] or vascular pathology, and do not present with vertebrobasilar symptoms. The screening procedures described cannot detect those patients in whom [manipulation] may cause an injury. They give a false sense of security to the practitioner. Several medical reports have described chiropractic patients who, after neck manipulation, complained of dizziness and other symptoms of transient loss of blood supply to the brain but were manipulated again and had a full-blown stroke. During a workshop I attended at the 1995 Chiropractic Centennial Celebration, Terrett said such symptoms are ominous and that chiropractors should abandon rotational manipulations that overstretch the vertebral arteries. But, as far as I know, his remarks have not been published and have had no impact on his professional colleagues.
The lack of predictability has been supported by data published by Scott Haldeman, DC, MD, PhD, a chiropractor who has served as an expert witness (usually for the defense) in many court cases involving chiropractic injury. In 1995, he published an abstract summarizing his review of 53 cases that had not been previously reported in medical or chiropractic journals. His report stated: These cases represent approximately a 45% increase in the number of such cases reported in the English language literature over the past 100 years. . . . No clear cut risk factors can be elicited from the data. Previously proposed risk factors such as migraine headaches, hypertension, diabetes, history of cardiovascular disease, oral contraceptives, recent head or neck trauma, or abnormalities on x-rays do not appear to be significantly greater in patients who have cerebrovascular complications of manipulation than that noted in the general population. Haldeman's main point was he could not identify any factor that could predict that a particular patient was prone to cerebrovascular injury from neck manipulation. This report was published in the proceedings of 1995 Chiropractic Centennial Celebration and was not cited in either the RAND or NCMIC reports.
In 2001, Haldeman and two colleagues published a more detailed analysis that covered 64 cases involving malpractice claims filed between 1978 and 1994. They reported that 59 (92%) came to treatment with a history of head or neck symptoms. However, the report provides insufficient information to judge whether manipulation could have been useful for treating their condition. Of course, malpractice claims don't present the full story, because most victims of professional negligence do not take legal action. Even when serious injury results, some are simply not inclined toward suing, some don't blame the practitioner, some have an aversion to lawyers, and some can't find an attorney willing to represent them.

What Should Be Done?
Chiropractors cannot agree among themselves whether the problem is significant enough to inform patients that vertebrobasilar stroke is a possible complication of manipulation. In 1993, the Canadian Chiropractic Association published a consent form which stated, in part: Doctors of chiropractic, medical doctors, and physical therapists using manual therapy treatments for patients with neck problems such as yours are required to explain that there have been rare cases of injury to a vertebral artery as a result of treatment. Such an injury has been known to cause stroke, sometimes with serious neurological injury. The chances of this happening are extremely remote, approximately 1 per 1 million treatments.
Appropriate tests will be performed on you to help identify if you may be susceptible to that kind of injury. . . .
This notice is a step in the right direction but does not go far enough. A proper consent should disclose that (a) the risk is unknown; (b) alternative treatments may be available; (c) in many cases, neck symptoms will go away without treatment; (d) certain types of neck manipulation carry a higher risk than others; and (e) claims that spinal manipulation can remedy systemic diseases, boost immunity, improve general health, or prolong life have neither scientific justification nor a plausible rationale.
In 2003, a coroner's jury concluded that Lana Dale Lewis of Toronto, Canada, was killed in 1996 by a chiropractic neck manipulation. Among other things, the jury recommended that all patients for whom neck manipulation is recommended be informed that risk exists and that the Ontarion Ministry of Health establish a database for chiropractors and other health professionals to report on neck adjustments.

The Bottom Line
As far as I know, most chiropractors do not warn their patients that neck manipulation entails risks. I believe they should and that the profession should implement a reporting system that would enable this matter to be appropriately studied. This might be achieved if (a) state licensing boards required that all such cases be reported, and (b) chiropractic malpractice insurance companies, which now keep their data secret, were required to disclose them to an independently operated database that has input from both medical doctors and chiropractors.
Meanwhile, since stroke is such a devastating event, every effort should be made to stop chiropractors from manipulating necks without adequate reason. Many believe that all types of headaches might be amenable to spinal manipulation even though no scientific evidence supports such a belief. Many include neck manipulation as part of "preventative maintenance" that involves unnecessarily treating people who have no symptoms. Even worse, some chiropractors- often referred to as "upper cervical specialists" -claim that most human ailments are the result of misalignment of the topmost vertebrae (atlas and axis) and that every patient they see needs neck manipulation. Neck manipulation of children under age 12 should be outlawed.