Osteopathy in the NICU: False Claims and False Dichotomies

I would like to preface this post by stating that I have worked with many DOs (Doctors of Osteopathy), and I have helped train many pediatric residents with DO degrees. I have found no difference in the overall quality of the training these students have received, and some of the very best clinicians I have ever worked with have been DOs. I would never prejudice my assessment or opinion of a physician based on whether they have an MD or a DO after their name.

Now, on to the discussion at hand.

I recently stumbled upon an article entitled, “Effect of osteopathic manipulative treatment on length of stay in a population of preterm infants: a randomized controlled trial”. There is nothing particularly exciting or interesting about this study, as there have been many published on the use of osteopathic manipulative therapy (OMT) in children. There aren’t that many RCTs, however, and this particular one, although published in the open-access BioMed Central Pediatrics (impact factor 1.98), was chosen to be included in AAP Grand Rounds. AAP Grand Rounds is a publication put out by the American Academy of Pediatrics (AAP) to help pediatricians “Stay current and save time with monthly critical, evidence-based summaries of clinical content from nearly 100 journals.” Because the AAP found this important enough for mention in this widely read publication, with a distribution of 19,000 (source: AAP, 2014), I thought it would be interesting to take a closer look at it. I am also interested in the very odd existence of the two, distinct paths to becoming a physician in this country, osteopathic and traditional medical school training. The distinction between the two is rarely discussed, even within the halls of academia or in our health care centers. That’s not to say that the topic isn’t discussed at all (in fact it was highlighted very recently right here on SBM), it has just remained a somewhat politically incorrect subject, sliding mostly under the radar. Having worked with and trained pediatricians with osteopathic degrees, I can tell you that discussions about this are considered taboo. This is primarily because osteopathic physicians have become mainstreamed over time (see below), and discussing the validity of the existence of their “specialness” is an awkward proposition. After taking a look at the paper in question, I’ll address this issue some more as I think it deserves additional attention.

Effect of osteopathic manipulative treatment on length of stay in a population of preterm infants: a randomized controlled trial.

This was a single-blinded RCT conducted at Santo Spirito Hospital in Pescara, Italy to explore whether OMT could shorten the length of stay among premature infants in their neonatal ICU (NICU). Secondary outcomes studied were the differences in daily weight gain and total cost of the NICU stay.

I’ll discuss the methods in a moment, but first let’s review the results.

Results

Infants who received OMT in addition to an osteopathic evaluation were found to have a significantly shorter NICU length of stay (LOS) than those in the control group who received only an osteopathic evaluation, without OMT. The mean average LOS was 26.1±16.4 days for the study group and 31.3±20.2 days for the control group (p<0.03). This was found to be true even after accounting for likely confounders using linear regression analysis. In terms of the secondary outcomes measures, there was no significant difference in daily weight gain between the two groups, but the authors found an almost 3 000€ (~$4 100) savings in hospital costs for the study group, related to the shortened LOS.

Methods

Now let’s look at how this study was conducted. A total of 101 preterm infants were included in the study. Infants were randomly assigned to one of two groups:

1. Osteopathic evaluation only
2. Osteopathic evaluation and OMT

Here’s where it gets silly, and I’d like to briefly step away from the article to review the fundamentals of osteopathy before moving on.

Osteopathy is a pseudoscientific belief system developed in 1874 by the physician Andrew Still. The fundamental premise of osteopathy is the inter-relatedness of mind, body, and spirit. Still believed that this tripartite system had intrinsic, self-healing and self-correcting capabilities. He believed that disease states were caused by misalignments or obstructions of various structural elements of the body (bones, muscles, fascia, nerves, blood vessels, lymphatics), and that specific corrective manipulations could facilitate the body’s innate ability to heal. Like many wacky medical belief systems of this era, osteopathy was not any worse than much of the prevailing thought and practice of the day. Today, the practice of osteopathy varies considerably within the profession, as does the degree of adherence to the fundamentals of the original dogma preached by Dr. Still. I will revisit this at the end of the post and discuss how it pertains to the current training and practice of osteopathic physicians in this country. Now, let’s return to the study.

The authors of the study define osteopathic practice much the way AT Still described it in the 1800s:

Osteopathic practitioners use a wide variety of therapeutic manual techniques to improve physiological function and restore homeostasis that has been altered by somatic (body framework) dysfunction…Osteopathic practitioners use their understanding of the relationship between structure and function to optimize the body’s self-regulation and self-healing capabilities. This holistic approach to patient care and healing is based on the concept that a human being is a dynamic functional unit, in which all parts are interrelated and possesses its own self-regulatory and self-healing mechanisms.

This definition is so rife with fantasy and meaninglessness that it is difficult to discuss rationally or with a straight face. But we shall try. First, it assumes there is some specific physiologic function that needs to be restored. Now, we can go along with this since pretty much every pathologic state, by definition, has some physiologic basis. But osteopathy then assumes that a detectable structural dysfunction lies at the center of this physiological “imbalance”, much the way many chiropractors claim that fictional spinal “subluxations” are the cause of myriad disorders. We know this is whole-cloth fantasy, since not one shred of evidence supports this claim. However, the entire concept of osteopathy is predicated on this fantasy. But the definition gets even stranger. It then claims that osteopaths use their understanding of this relationship between a specific structural disequilibrium/obstruction/misalignment/lack of homeostasis (pick and choose) and a specific, corresponding physiologic dysfunction to design an effective manipulative fix. These manipulations then allow the body to resume its innate, self-healing magic to restore the body to health. I still can’t figure out why, if the body has innate self-healing abilities, these manipulations are required, and what they are supposed to be doing.

Now, recall the two groups of infants in the study, Group A (the control group) and Group B (the study group). All infants received 10-minute osteopathic evaluations by one of four Group A osteopaths. The study group infants additionally received a 10-minute evaluation and a 10-minute OMT session from one of four Group B osteopaths who visited the infants at a different time. After the group A osteopaths performed their evaluations, they remained standing at the bedside for an additional 10 minutes, supposedly as a form of “blinding”. Let me repeat this in table form for clarification:

The authors explained blinding as follows:

NICU staff was blinded to patients allocations, since all infants were at least touched by osteopaths from group A and B and osteopaths spent the same amount of time in front of incubators and/or beds. Moreover osteopaths were unaware of study design and outcomes.

I read the methods over and over again and this explanation is either untrue or there is an error somewhere. Infants were not at least touched by osteopaths from each group. The control subjects were touched only by group A osteopaths, who then stood by the bedside (doing what, we don’t know) for an additional 10 minutes. Study subjects had this same evaluation and bizarre “standing” gimmick, and then at another time had an additional evaluation by group B osteopaths followed by an OMT session. I don’t see how group A osteopaths standing by the bedside for ten additional minutes after their evaluation accomplishes anything, since the study infants will again be visited by a second osteopath. NICU staff would certainly see that two different osteopaths were visiting the study infants. There is no blinding here as far as I can tell. The authors describe further “blinding” as follows:

Osteopaths from group A and B entered to the NICU in different hours of the schedule days, to provide blinding and to avoid possible confounding.

Even if there were different NICU staff working at these times, they could easily discuss their observations with one another. Again, this is not blinding. Of course, the treating osteopaths could not have been blinded.

So we know the methods are flawed and prone to producing biased and confounded results, but let’s discuss the intervention, which itself defies all Bayesian rules of logic. For those of you not familiar with the precepts of osteopathy, the fantasy is rich:

The aim of the structural examination is to locate the somatic disfunction [sic]. In newborns the structural exam is usually performed with the infant lying down on the table. Diagnostic criteria for somatic disfunction [sic] are focused on tissue texture abnormalities and tone. Areas of asymmetry and misalignment of bony landmarks are evaluated. The quality of motion, its balance and organization are noted.

If this sounds to you like hocus-pocus, you are not alone. Structural, anatomic assessment is obviously important for detecting a wide array of newborn problems and is, of course, performed every time an infant is examined. But the osteopathic evaluations described above purport to detect subtle findings reflecting complex, systemic, physiologic processes. They claim to detect vague alterations of texture, tone, and alignment indicative of the overall physiologic state of these premature infants – a tidy package of physiologic dysfunction that is somehow contributing to their need to remain hospitalized. As with other forms of CAM, there is a drastic oversimplification at play here, of phenomena that are very complex. To claim that a premature infant’s hospital stay may be prolonged by physiological “imbalances” that can be predictably detected by an osteopath’s trained, hands-on assessment, and which can then be treated with simple manipulations, is not just fantastical and wishful thinking, it is a massive simplification of a very complex physiological reality. A reality over which no “traditional” practitioner would ever claim such mastery and control. Because these telltale stigmata do not actually exist, inter-rater reliability for osteopathic evaluations would be expected to be poor or non-existent, as has been shown in the few studies that have looked at it (Giles, 2011, Wirth-Pittullo et al., 1994, Pattyn et al., 2013). The “dysfunctions” identified by the evaluators in this study are not stated but are presumed to be related to significant and complex issues keeping these infants in the NICU, since the manipulations concocted to relieve them resulted in an improved LOS. Several hypotheses for how OMT may be working here are put forth in the discussion section of the paper. These include possible anti-inflammatory effects, and bringing “balance to the sympathetic and para-sympathetic inputs, creating an improvement of newborns clinical condition.” I will not be discussing the evidence for the benefits of touch and massage on newborns in the NICU (there is some interesting data on this), but this has nothing to do with the imaginary structure-function claims of osteopathy. Even if the hands-on interventions of the osteopaths in this study did truly improve the LOS for these infants (and this cannot be assumed from the evidence in this study), it in no way validates the mechanistic claims of osteopathy. If there is benefit to hands-on touch and massage in neonates, that’s one thing, and that is an interesting and potentially valuable thing to study. But it says nothing about and does not require the existence of osteopathy, which is based on non-existent relationships and false assumptions about anatomy and physiology.

Osteopathy today

There are large regional and individual differences in the ways in which osteopathic providers practice OMT. Different countries have their own regulations and even definitions of osteopathy. Though individuals trained as osteopaths in the 19^th and first half of the 20^th century practiced in a fashion closely aligned with Still’s original dogma, osteopathic training and practice in the US over the last half century has changed dramatically, coming to resemble traditional modern medical training and practice in nearly every respect. In the United States, there are two types of medical schools that confer degrees bestowing equivalent privileges to practice medicine and surgery. The DO curriculum in the US is now indistinguishable from that of the MD curriculum of traditional medical schools, with one glaring distinction. A vestigial, non-trivial remnant of osteopathic teaching has been retained and remains, at least for the sake of appearances, at the heart of the mission of these schools. While osteopathic schools also claim distinction by focusing on a “holistic” approach to health and on disease prevention, this is a false distinction as traditional medical schools place this approach at the core of their missions as well. Despite the one true distinction between traditional and osteopathic medical school (the focus on osteopathic manipulative therapy), a minority of DOs continues to practice OMT once they finish their training (Johnson et al., 2001). My experience working with many DO students and graduates is that most do not choose this path to a medical degree because of the DO component of the training, but because of the less competitive nature of the admissions requirements. The traditional DO student body has tended to be older and have lower GPA and MCAT scores than that of traditional medical schools. Probably a plus, it has also tended to be comprised of students with more diverse, non-traditional backgrounds. These distinctions, however, have grown smaller as traditional medical schools have broadened their perspective about what characterizes a good candidate. More and more, medical schools are seeking individuals with broader, more well-rounded backgrounds, often looking toward the liberal arts for their prospective students. So other than a lower admissions standard and the focus on a pseudoscientific belief system, there is no difference between the two educational systems. So why should this two-tiered system even exist? It is time to rid ourselves of this pre-scientific, dual system of medical education. It is time to acknowledge that there is no such thing as osteopathy, and focus our efforts on improving our science-based approach to medicine.