“Osteopathy has no evidence base”

Evidence in medicine

Among ways of caring for health, there are two poles: working with a condition or diagnosis, and working with an individual person. Osteopathy is closer to the second, but we will begin with medicine.

For modern evidence-based medicine, the first approach is central, although there is a move towards personalisation. Personalised treatment choices are still based on findings such as “in women of reproductive age with migraine with aura, drug X reduces attack frequency by Y%”. Without this, medicine would remain in the Middle Ages. Such studies follow fixed rules:
1) A large group of people with the same condition or diagnosis.
2) The greatest possible homogeneity within the group.
3) A double-blind study. When testing a drug, neither patient nor doctor knows who receives the drug and who receives a placebo.

How achievable are the ideals of evidence in medical practice?

The same diagnosis. It is relatively convenient when diagnoses can be defined very precisely. Usually these are either clearly established through laboratory or genetic tests and similar investigations, or serious and dangerous conditions, often definitively established after death. What about diagnoses such as “dorsalgia”, which simply means back pain? The back hurts in different places, with different movements (or independently of them), and improves or worsens under different circumstances. How homogeneous can a sample united by a diagnosis of dorsalgia be? The good news is that some heterogeneity can be compensated for by sample size, and homogeneity can also be tested separately using statistical methods.

Sample homogeneity: ideally, all participants would be the same age, have had the same previous illnesses, be of the same sex, and have their menstrual cycles in the same phase during the study. Statistical methods could then be applied as correctly as possible. This is why mouse studies use particular genetic strains and, where possible, males. These are all attempts to make samples as homogeneous as possible, while recognising that true homogeneity is unattainable.

Placebo control works well when studying medication and much less well for technically more complex methods. How can you perform a dental procedure that feels indistinguishable from the real one but omits the key elements that determine its result?

Evidence in medicine is therefore very different from proof in physics. There is a saying that “medicine is the second most exact science after theology”. Nevertheless, this kind of evidence is the best we have today, and is sufficient to solve many problems effectively.

Is an evidence base possible in osteopathy?

The short answer is: to a certain extent, usually under the following conditions:

Large, preferably very large samples.
Or, conversely, case studies: individual cases described in detail.
With large samples, the osteopath’s actions should not be restricted by a protocol: osteopaths simply do not work to protocols, so such studies are not appropriate.

Osteopathy is highly personalised: it works with an individual person rather than a diagnosis, taking account of lifestyle, needs and personality. The choice of strategy and specific techniques is therefore very individual. Repeating the same strategy for another person with the same diagnosis makes little sense, because their spinal loading, physical activity, history of surgery and injuries are different.

In osteopathy, unlike science and evidence-based medicine, reproducibility of results is generally not the focus. Just as no two patients are identical, no two osteopaths are identical. We build our approach and select techniques differently. Every case is unique. The only way to check is to retest movement at the end of a session and assess the patient’s subjective experience. In effect, an osteopath compares a person with themselves more than with others.

While medicine without protocols would descend into chaos, working to a protocol in osteopathy represents very low-quality work: protocols are used only for teaching osteopaths. Good osteopathic studies now mainly interpret the osteopath’s work as a black box: there is a group of patients with similar concerns, osteopaths who do what they consider appropriate for each individual rather than perform the same procedure, and an assessment of results.

What research exists on osteopathy’s effectiveness?

Let us consider several papers assessing effectiveness. They address different clinical conditions and use different methodologies. Together, they give a contradictory impression, which is to be expected given the nature of osteopathy, patients presenting with less severe conditions (making statistical analysis harder), and other methodological difficulties discussed at the end of this section.

So, does osteopathy work?

This appears to depend greatly on the individual osteopath’s skills and the individual patient’s condition, as well as research methods, discussed below. This probably explains why papers on the same problem can show either statistically significant improvement or no benefit. Hence the appropriate conclusion of reviews: we do not know whether the method works as a whole.

One major methodological problem is that although osteopaths in the European understanding of the profession train in fairly standardised ways, they work very differently. As usual, the devil is in the detail. Palpation skills, reasoning and experience create enormous differences between individual osteopaths’ results, making general statements such as “osteopathy helps with…” difficult. It is more accurate to say “osteopath Jim Watson works very well with various shoulder problems, but for a headache, see Fred Simmons”. An elegant study by Dr Sabini and colleagues supports this: Sabini, R.C., Leo, C.S., Moore, A.E., 2013. The relation of experience in osteopathic palpation and object identification. Chiropr Man Therap 21, 38. https://doi.org/10.1186/2045-709X-21-38. People with very different experience and training — renowned osteopaths, osteopathy students and other volunteers — were asked to identify small objects by touch through layers of fabric and foam. Results showed no relationship to experience. Some individuals performed very well, but they came from all categories.

Other methodological difficulties in osteopathic research are well described in Dr Cerritelli and colleagues’ paper on osteopathy for neurological problems, cited above.

Great sample heterogeneity. Even when patients share a diagnosis, some feel fairly well while others have a dozen other diagnoses and poor quality of life. Good papers test homogeneity statistically, but many osteopathic papers do not, and some do not demonstrate the homogeneity required to compare treatment and control groups.
Researchers pay little attention to treatment patients receive alongside osteopathy. Information on medication and dosage, lifestyle, physical activity and so on is absent.
Adverse effects of osteopathy are not examined either. Although usually minor, it is inappropriate to omit them.
The authors conclude that osteopathic studies struggle to balance rigorous reasoning and reproducible results with individual experience and clinical intuition.

What should a patient do? Whom should they see?

Above all, osteopathy should not take precedence over evidence-based medicine. There are exceptions: for example, the British NHS recommends osteopaths as a first point of contact for musculoskeletal problems. Responsibility for identifying red flags and referring the patient then lies with the osteopath, and British osteopathic training prepares practitioners for this.

You should certainly not spend time seeing osteopaths in cases of:

Acute injury.
Acute circulatory problems.
Other emergencies.
Broken skin.
Acute psychiatric conditions.
Lack of patient consent.
In other cases, osteopathic work generally involves working alongside doctors. A competent and responsible osteopath always asks what an orthopaedist, gynaecologist, paediatrician or other doctor has said about the concern. Osteopaths cooperate closely with doctors, which is both ethically appropriate and leads to better patient outcomes.

At the same time, many problems can be addressed through osteopathic sessions alone. This mainly concerns musculoskeletal problems but is not limited to them.

A prospective patient effectively faces two questions:

Is osteopathy acceptable to them in principle, knowing that it offers no guarantees but generally involves small risks?
Which particular osteopath should they choose?
Judging by the studies discussed above, practitioner choice is a fundamental factor in success. One may succeed where another does not. Follow recommendations, check the osteopath’s experience and qualifications, and trust your intuition. A practical solution may be to agree on three sessions, for example, and if there is no noticeable effect, try another osteopath or look for a different kind of practitioner. Osteopaths often suggest this themselves.

Who are you to draw these conclusions?

The author is Natalia Ermachenko, osteopath (European School of Osteopathy, UK), PhD in physiology and neurobiology (Lomonosov Moscow State University), founder and head of Still Point osteopathy centre in Yerevan, and president of the professional association of osteopaths registered in the Republic of Armenia.