I have never hid the fact that I am not a fan of so-called infant chiropractic. There is no condition for which there is a shred of plausibility that chiropractic manipulation or adjustment of the infant skull or spine would provide specific benefit. Not colic. Not reflux. Not ear infections. Not craniosynostosis. And not SIDS. Yes, you read that right.
So-called clinical placebo effects do, however, give parent and practitioner alike the false perception of benefit. This is unavoidable when the majority of infant conditions that are claimed to be amenable to chiropractic care are self-resolving over time. Or when a caregiver’s perception of the severity of subjective symptoms, such as fussiness, is altered by an expectation of improvement.
This hasn’t prevented chiropractors from marketing their services to caregivers of young children, a group that is particularly vulnerable in my opinion. Thousands of infants are brought to chiropractors every year, in large part because of credulous reporting like this, complete with the standard opening anecdote, token skepticism (it’s me), and false balance. And thousands of individual chiropractic websites that a curious parent might stumble upon present worthless testimonials and case reports as proof of effectiveness.
With all this in mind, as it almost always is, I recently read a paper published in the Journal of Manipulative and Physiological Therapeutics that claims to have used the power of “big data” to show that maybe there is something to this infant chiropractic thing after all. In fact, according to the authors, it is the “first large prospective study reporting parental report of effectiveness of routine chiropractic care for infants”. Manipulative indeed.
Maternal report of outcomes of chiropractic care for infants
I’m confident that you already know where this is heading. But before I get to just what this study involved, there were a few interesting comments in the introduction:
An important development in health care in the past decade has been the increasing interest in patient-centered care emphasizing the patient’s own perspective on their illness and treatment. Patient-reported outcome measures are designed to assess a patient’s subjective impressions of their health with respect to symptoms, function, and health-related constructs such as health-related quality of life. The WHO points out that involvement of the patient in their own health care is a social, economic, and technical necessity.
This is true. We have come a long way in the practice of real medicine in regards to leaving our paternalistic past behind, though we are far from perfect. An increased focus on patient autonomy, or parental authority in pediatrics, and in shared decision making is incredibly important. The authors, however, are setting up an implication that downplaying or appearing to outright disregard the perception of parents whose infants have undergone chiropractic care is poor form.
It gets worse:
Routine collection of patient-reported outcomes has been suggested as a method of collecting large amounts of data to capture information about what occurs in routine clinical practice. These large amounts of data may be used to measure quality of care and evaluate comparative effectiveness of treatments.
They then segue into collecting “big data” for infants undergoing chiropractic care, which they see as crucial for determining treatment effects, safety, and parental satisfaction. They even acknowledge that chiropractic has been criticized for a lack of “high-quality evidence to support the care of this population”, and position studies such as this as an improvement. While there is certainly some truth to this, that patient-reported outcomes are an important aspect of clinical research and care, they are leaving out some very important caveats.
There can be significant problems with focusing too much on certain patient-reported outcomes in certain clinical situations. In the case of conditions that are highly subjective, uncontrolled data can be very misleading if you are looking for evidence that a treatment truly works. The same can be said for developing an understanding of potential adverse outcomes. In regards to satisfaction, using this as a metric has been shown to actually result in lower quality and more expensive care with worse outcomes.
I’ll cut to the chase now. This was an observational study involving prospectively collected data using the United Kingdom Infant Questionnaire (UKIQ) from mothers who sought chiropractic care for their infant. So right off the bat, there is risk of selection bias. Mothers reported their child’s age, sex, complaint at presentation, type of birth, and referral source. The UKIQ also allowed the mothers to grade their child’s feeding, sleep, and crying on a scale from 0 to 10, with 10 representing a “serious problem”. Also included were questions on infant sleep position and general pain or discomfort as well as maternal anxiety and quality of life.
The treatment each child underwent was not standardized but left up to the individual chiropractor, of which there were many at several different clinics. Follow up questionnaires were filled out at the fourth or final visit, whichever came first. At that time, mother’s rated their overall satisfaction with their child’s care from 0 to 10, with 10 representing “very satisfied”, as well as their perception of “global impression of change” in their child’s condition. This response could range from “much worse” to “completely better, made a huge difference, like a different baby”.
A subset of mothers, although how this was determined was not explained, were asked about the development of adverse events or side effects. The authors state that a question about cost effectiveness “was added later when it was discovered that this type of research had particular utility in asking such a research question.” This sounds like one of those researcher degrees of freedom that Steve is always talking about.
2,001 mothers completed the intake questions. Unfortunately, only half gave follow-up information, with even some of those responses being incomplete. This is a pretty glaring problem and certainly calls their results into questions. And speaking of results, to absolutely nobody’s surprise they were all positive and statistically significant.
They found that the mothers who completed the second questionnaire reported significant improvements in infant feeding, sleeping, crying, supine sleeping, pain, neck movement and tummy time, as well as in maternal anxiety, depression, and satisfaction with motherhood. Virtually all responding mothers rated their children as being improved with over 80% rating them as much or completely better. Over 80% of responding mothers also rated their satisfaction with chiropractic care as a 9 or 10. It all sounds amazing, doesn’t it?
None of this really means anything when you take into account the limitations of the study and the extreme lack of plausibility that gently prodding a baby’s spine would play a role in any of these conditions. In the discussion, the authors admit as much. Well not the part about the lack of plausibility. These results tell us that studies designed to have positive results…have positive results.
The authors admit that this type of study can’t determine if chiropractic actually is effective. But they weasel out of this by claiming that proper studies might not be able to tell if a treatment is truly clinically effective anyway. For this, they ponder, perhaps asking the parents really is the best means of answering the question.
Perhaps the most grandiose claim made by the authors of the study involve postpartum depression. They claim that because there were apparent reductions in maternal anxiety and improvements in quality of life, chiropractic treatment of infants might prevent postpartum depression. They even slip in an apparent dig at conventional doctors who fail to help mothers when they merely reassure them that their child’s condition is not life-threatening rather than providing “specific contextualized help”. In other words, by telling a mother that their child is crying because of a fictional subluxation and then providing a bogus remedy.
Conclusion: Just another worthless collection of anecdotes
This study adds nothing of value to our understanding of infant medicine. It took a highly selective population of mothers likely already sold on the benefit of chiropractic care of infants and asked for subjective perceptions of subjective symptoms. All of the complaints involved in the questionnaires are likely to resolve over time without intervention, and are very likely to seem improved when non-specific placebo effects are involved. And I have to wonder how statistically significant would these results have been if the 1,000 mothers lost to follow-up had completed their surveys.
By questioning this data, I’m not attacking mothers or ignoring caregiver assessments of their children. I rely on parental perception every day to provide care for infants. It’s a team effort. But these reports have to be evaluated in context. I’ve lost count of how many times I’ve been told by a parent that their child is subjectively improving only to find objective evidence that shows the opposite. Medicine, and I mean real medicine rather than the make believe cargo cult pretend medicine that some chiropractors play at, is rarely as simple as they would have us believe.