Lateral cervical curve changes in patients receiving chiropractic care after a motor vehicle collision: a retrospective case series

Concern no. 1 

Harrison et al indicate that stretching is responsible for the improvement in our subjects.

Our study1 looked at 13 patients, who averaged 6.4° of increase in cervical lordosis following chiropractic care. Harrison et al in their letter indicated that they selected 2 subjects out of our group of 13. These were the only 2 subjects that received adjustments from the Activator Adjusting Instrument (AAI) but did not do stretching exercises. The stretching exercises that the other subjects performed were, as we stated in the article, mild in nature and did not involve cervical extension. They note that the total change for those 2 subjects was 4.5°, and they concluded from this that the Activator Adjusting Instrument caused 2° of change, which they indicate is within the average error of measurement, in essence that the adjustments made no change and that it is the stretching that accounts for the overall improvement in the lordosis of the group of 13.

By way of explanation, let us use their logic and pick 2 other subjects out of the group of 13 total subjects. We will choose 2 subjects that together had an increase in lordosis of 39.5°. We can then, using the logic of Harrison et al, say that since the stretching exercises are responsible for the increase in lordosis, then the average increase in lordosis due to stretching is slightly less than 20°. This would make mild stretching far more effective than what Harrison, Ferrantelli, and Oakley refer to, as “…the ‘gold standard’ of CBP cervical extension traction.” In one of the articles that was used as a reference to support this “gold standard” statement and which is one that we referenced in our article,1 they were only able to obtain 13.2° of improvement in cervical lordosis with 60 visits.2 Using the Harrison et al logic, the gold standard has now become mild stretching, which does not include cervical extension.

However, what would occur if we picked 2 other subjects from the group of 13. If we select 2 subjects whose total is a loss of 8° of cervical lordosis and if we invoke the Harrison et al theory that only 2° of change toward lordosis was caused by the Activator Adjusting Instrument but that we have discarded this 2° as it was within the error of measurement, then we would say that mild stretching that did not include cervical extensions caused a 4° decrease in lordosis in the cervical curve.

Obviously, you cannot select 2 subjects of your choosing and make claims regarding the entire group based on the 2 selected subjects. They have indicated that we did not comment on this issue, but we note that we did not divide out the amount of change that may be attributed to any single part of care. In fact, we indicated that we cannot account for the individual effects of adjustment, stretching, and natural progression of the condition.1 We discussed the limitations of our study in our article.

As to the question about exactly what head motions were involved in the stretching, they included 5 subjects who performed cervical flexion and cervical rotation stretches and 6 subjects who performed these movements and also performed lateral cervical flexion stretches. These stretches were mild in nature and, as can be seen, did not include cervical extension.

Concern no. 2 

Our critics indicate that we have inaccurately summarized the results of an article written by Dr Donald Harrison et al2 when we noted that no change was made by the use of Chiropractic Biophysics adjusting procedures.

In this article, Dr Donald Harrison et al2 state, “…no change in the control group and essentially no change in treatment group 2 was measured.” They further state, “In the treatment group 2, no statistically significant changes existed between the pre- and posttests except atlas angulation to horizontal…” Treatment group 2 received “…diversified spinal manipulation and drop table adjustments five times per week for 10-14 wk (12 wk ± 2).” And from the conclusion, “From this study we conclude that a treatment regimen of diversified spinal manipulation combined with drop table adjustive procedures showed no change in lordosis on post x-ray following 10-14 wk of treatment five times per week.”

In our article,1 we stated, “Harrison et al, in a retrospective pilot study showed no change in the magnitude of cervical lordosis with Chiropractic Biophysics spinal adjusting procedures…” It should be clear that the statement in our article was appropriate.

They go on to state, “Harrison et al found 2.9° increase in the C2-C7 lordotic angle in the control group and 1.2° in patients receiving PA drop table adjustments at mid neck (without traction).” He then indicates that this change in the treatment group is the same as what we observed in our study.

The statement that the Harrison et al2 study obtained the same improvement that we found in our study is obviously not correct, as we observed an average of 6.4° of improvement in our study.

Even though Dr Donald Harrison et al2 had referred to the cervical lordosis change in their study as not statistically significant, let us follow the logic of Dr Deed Harrison et al. If we accept that these changes in the Harrison et al study2 should have been noted in our article, then we would have to point out that the lordosis in the control group of the Harrison et al2 study improved more than the subjects who received the Chiropractic Biophysics adjusting procedures. In other words, Chiropractic Biophysics adjustment procedures interfered with the natural improvement in the cervical lordosis.

Concern no. 3 

Harrison et al, in their letter, stated, “…Coleman et al stated that neutral resting posture is repeatable, and utilize this statement to conclude a 6.4° increase in lordosis is likely a treatment effect.”

We wish to point out again what we said in our article.1 “We were not able to determine the individual effects of adjustment, stretching, and natural progression of the condition.” It is Dr Deed Harrison et al who are making claims for our article. We reported what we found.

Next, we did use a reference for repeatability of posture.3 This reference was also used by Dr Deed Harrison et al4 in a previous article.

Dr Deed Harrison et al go on to state, “However, we find it necessary to point out that x-ray positioning is repeatable in non-whiplash injured control subjects; but not in whiplash patients.”

It appears that they are using a reference to an article by Dr Deed Harrison et al4 to show that nonwhiplash injured control subjects have repeatable radiograph positioning and 2 articles as references to show that whiplash patients cannot be positioned for a lateral cervical radiograph in a repeatable manner.5, 6

We do not feel that a comment is needed on the appropriateness of the use of the Dr Deed Harrison et al4 article as a reference because that section does not question our article.1 However, we will comment on the use of the other 2 articles.5, 6

Harrison et al cite an article by Loudon et al5 to support their position that radiograph positioning is not repeatable in whiplash patients. This reference looks at patients in a seated posture. We note that Dr Deed Harrison et al4 in their article “Repeatability over time of posture, radiograph positioning, and radiograph line drawing: An analysis of six control groups” point out that their patients are standing, and in a paper that we reference in our article and which is one of the articles that Dr Deed Harrison et al cite to prove that CBP possesses the gold standard method, the patients are also standing.2 The patients in our article1 are indeed standing and we included a reference for the repeatability of standing posture in our article.3 The reference to seated posture that Harrison et al use is therefore not applicable.5

However, we would like to explore further the statement by Dr Deed Harrison et al, “…that x-ray positioning is repeatable in non-whiplash injured control subjects; but not in whiplash patients” in conjunction with his use of the Loudon et al reference.5 As a whiplash injury is a sprain/strain type injury to the neck, then if we follow Dr Deed Harrison's argument, are we to assume that all patients who have a sprain/strain problem in the neck cannot be accurately radiographed, as their posture is not repeatable? In at least 1 of the articles referenced by Dr Deed Harrison et al to show that CBP was the gold standard,2 the patients were not rejected if they had suffered a sprain/strain injury. As CBP adjustments are based on posture and radiograph, does this mean that patients who have strain/sprain injuries of the neck should avoid being treated by someone using CBP technique?

The next article referenced by Harrison et al is also not applicable to this discussion. This article by Ruben et al6 is not strictly about whiplash. This study is about a group of subjects who were dizzy following mild head injury or whiplash injury. The 2 types of injury are “lumped” together to form 1 group, ie, the whiplash cases were not divided from the mild head injury patients. Also, this article does not measure head posture or the posture of the head in relation to the thorax or the cervical lordosis. This article used a force platform and measured the body's “center of pressure” (COP) as the patient stood on the platform. For these reasons alone, the Ruben et al6 article is not applicable; however, there are additional reasons as to why this article is not helpful to Dr Harrison's claim.

Ruben et al6 note, “The trends in the somatosensory assessment data indicate that when accurate vision was combined with orientationally inaccurate somatosensory inputs, there was little disruption in the balance of the group with head injury.” They also state, “These patients exhibited higher COP sway patterns when deprived of accurate visual cues during feet-together and feet-apart stances.” In our study, our patients were positioned in the natural standing position with their eyes open. However, we did note1 that the CBP method of positioning the patients for lateral cervical radiographs involved the patients closing their eyes and flexing and extending their heads prior to assuming a neutral position.2 It becomes apparent that the Rubin et al6 article is not applicable to our article.1

However, if Harrison et al still feel that the Rubin et al6 article is an appropriate criticism, we would like to point out that it would invalidate a portion of CBP nonprofit research and bring into question the CBP technique.

Harrison et al then go on to discuss another CBP nonprofit paper.7 They indicate that extending the head 14° or less only changed the cervical lordosis 7°. Then they note that Hellsing found that extending the skull 20° only caused a 10° change in cervical lordosis. They indicate that we did not provide head extension measurements and that it was likely that our change of 6.4° in cervical lordosis was “…due to slight positional differences on the two sets of radiographs.”

First, we do not feel that 14° or 20° of head extension is a “slight” positional difference and CBP authors have long argued that posture is repeatable. Next, we note again, that posture is repeatable and we referenced Sandham3 in our article.1 Either of these 2 points is enough to reject the reasoning by Harrison et al on this point.

However, we would like to follow this line of argument. We suggest that if a patient entered our office with their chin stuck on the chest due to a long-present loss of cervical lordosis, that we might consider it an appropriate clinical goal, for our care, to attempt to restore cervical lordosis and to allow this person to stand with their head in a much more upright position. If we then obtained our goal and a comparative radiograph showed that the patient had an increased lordosis, then the logic of Harrison et al would say that the patient did not have an increase in lordosis but merely elevated their head. We would like to arbitrarily choose some other methods of determining radiograph measurement. What could happen if we chose shoulder level instead of bite line for an indicator? We might take a case in which a person entered an office with a large lateral spinal curvature which caused him to stand with his spine laterally flexed far to the right and his shoulders not level with the horizon. If after a period of care the patient improved and no longer stood with the spine in lateral flexion and the shoulders were now level, the Harrison et al logic would say that the alignment of the spine had not improved but that the shoulders had merely been leveled. Perhaps we could use eye levels and take a patient with toricollis whose cervical spine was laterally flexed and his eyes not level with the horizon. If after a period of care the patient stood without a torticollis and the spine showed normal alignment, the Harrison et al logic would say that spinal alignment had not improved but the eyes had merely been leveled with the horizon.

Harrison et al have chosen an arbitrary method (giving credence to leveling the bite line) and an arbitrary selection of alignment points (they could have used the sight plane or some other points) and indicated that others should conform to their self-imposed method of analysis. We have reported on what we found. In a further note, Harrison et al7 in their “slight” head extension article state, “The results show that slight extension of the head does not change a reversed cervical curve into a cervical lordosis as measured on the lateral cervical radiographs.” In addition they state, “Therefore when slightly flexed head positions are visualized on initial radiographs and not on post-treatment radiographs, surgical and conservative kyphotic alignment changes verified post-radiographically are likely due to treatment interventions.” We note that 2 of our subjects1 began care with kyphotic spines and both of them regained lordosis. One of them improved 8.5°, while the other improved 24°. Obviously, this is not the result of “slight” head extension, according to the Harrison et al7 article. Finally, in this section, we note that Harrison et al7 in their “slight” head extension article used 40 neck pain patients.

The patients received lateral cervical radiographs, which were described as follows, “For the neutral lateral cervical radiograph, taken at 182.9 cm (72 in), subjects assumed a comfortable resting position, which has been shown to be reproducible.”7 In another article, while describing their method for obtaining a lateral cervical radiograph, Dr Deed Harrison et al4 state, “Note that if head flexion is present in the neutral resting posture, it is not artificially changed by leveling the bite line.” The radiographs for our article1 were taken in the natural standing posture, and we did not use the artificial method of leveling the bite line.

We stand by our article.

In the next section, Dr Deed Harrison et al continue to discuss CBP nonprofit studies. This section seems to be a listing of Dr Deed Harrison's and CBP publication accomplishments rather than a question, so we have no comment for this section.

In their last section, Harrison et al state, regarding our article, “At best, their data suggest that stretching might be responsible for increased lordosis but that adjusting with the AAI is not.” In reply, we note that we reported what we observed. The patients in our study improved an average of 6.4° following chiropractic care, which consisted of Activator Instrument Adjustments and mild stretching without extension. “We were not able to determine the individual effects of adjustment, stretching, and natural progression of the condition.”1

In a parting comment, Harrison et al indicate that CBP cervical extension traction is the gold standard. We feel that no technique has demonstrated that it is the gold standard.

We have carefully considered the comments of Drs Harrison and Farrentalli and Mr Oakley. They have not been persuasive and we stand by our article as it was published. But we do thank them for their interest.

In closing, we feel that our article speaks for itself.1 It is a study of 13 patients, which, as the title indicates, looks at “Lateral cervical curve changes in patients receiving chiropractic care after a motor vehicle collision: A retrospective case series.” “It is a descriptive study of the patients who met the criteria.”1 We reported on what we observed. Our present hope is as it was when we wrote this article, “We hope this case series will spark interest among researchers so that prospective studies utilizing many types and combinations of chiropractic care will be forthcoming.”1 In case further elaboration of prospective is needed, we offer the following: A claim of efficacy, effectiveness, or safety should be based on experimental designs (eg, randomized, blinded trial) that ideally control for the natural history of the disorder (2 groups). Any claim of efficacy for 1 clinical approach versus another would require a well-designed and executed randomized study of the 2 different interventions combined with a natural history control (3 groups).