Conflicting Research Evidence in Evidence-based Practice: Whose Bias do you Seek?

As a health services researcher, I have the opportunity to interact with a number of heath care stakeholders, including patients, health care professionals, funders, administrators, students and educators. I find these opportunities very valuable as it provides me with an insight into how health services research is perceived by these stakeholders and what, if any, issues they confront when utilising research to inform their practice and policies. As someone who has a passion for evidence-based practice generally, and the uptake of evidence (evidence implementation) more specifically, the issues raised at these interactions highlight the real challenges confronting those at the coal face.

Recently, during one such interaction, health professionals raised with me the issue of conflicting research evidence. They remarked that as health professionals it was very difficult for them to keep up to date with research evidence due to time and resource limitations and even when they tried, the evidence base was constantly evolving and often conflicting. These health professionals are not alone as previous research has identified that conflicting research evidence is a barrier in the uptake of research evidence into practice.1,2

So, researchers from the International Centre for Allied Health Evidence (iCAHE) decided to test and verify the issue of conflicting evidence in allied health. The topic of choice was the best evidence underpinning manual therapy, specifically manipulation and mobilisation, for headaches. Headache disorders such as migraine, tension type headache and cervicogenic headache (CGH) are common complaints which can impact on the employment, family and social life of the person suffering from headaches. The financial burden to society from chronic headaches in the form of lost working hours and reduced productivity is estimated to be significant. Common treatments may include massage, trigger point therapy, reflexology, spinal mobilization and spinal manipulation (SM), therapeutic heat and cold, exercise therapy and other miscellaneous approaches. Manual therapy interventions for headache disorders are popular in the clinical settings with manipulation, mobilization, massage and other soft tissue techniques commonly used by health professionals to treat headache disorders.

We undertook a rapid review of the literature and we identified three recently published best evidence sources. We identified two systematic reviews and one evidence-based guidelines, which was underpinned by a systematic approach.3,4,5 The systematic review by Chaibi et al3 summarised the research evidence for manual therapies for migraine. This review included seven randomised controlled trials (RCTs). The findings from the review indicate that massage therapy, physiotherapy, relaxation and spinal manipulative therapy might be equally effective as some medications in the prophylactic management of migraines. However, the authors then make an interesting statement. They conclude by stating that "However, the evaluated RCTs had many shortcomings. Therefore, any firm conclusions will require future, well-conducted RCTs on manual therapies for migraine” (pg. 127).3 This concluding statement indicates that effectiveness of manual therapy for migraine is presently equivocal until further research with sound methodology can contribute to the evidence base.

The systematic review by Posadzki and Ernst summarised the research evidence for spinal manipulations for cervicogenic headaches.4 This review included nine RCTs. The findings from this review indicate that despite some individual RCTs showcasing positive outcomes for spinal manipulations Posadzki and Ernst conclude that "the evidence from RCTs of SM for the treatment of CGH is thus ambiguous and, for several reasons, inconclusive” (pg. 1135).4 Posadzki and Ernst highlight methodological concerns in the included RCTs, and therefore increased opportunities for bias in the current evidence base. Due to these reasons they state that the therapeutic value of spinal manipulation for cervicogenic headaches remain uncertain.4

The evidence-based guideline by Bryans and colleagues summarised the research evidence for chiropractic spinal manipulation in the treatment of headaches in adults. This review included 21 articles comprising of controlled clinical trials, randomised controlled trials and systematic reviews.5 This evidence-based guideline concludes by stating that "Evidence suggests that chiropractic care, including spinal manipulation, improves migraine and cervicogenic headaches” (pg. 274).5 While Bryans and colleagues recognise methodological issues associated with the evidence, similar to Chaibi et al and Posadzki and Ernst, their recommendation differs vastly when compared to the other two systematic reviews.3,4

It is interesting to note that all three secondary evidence sources (systematic reviews and guidelines) were recent (published in 2011), had sound methodological processes and included only high level primary (such as RCTs) and secondary (such as systematic reviews) evidence sources, which informed their recommendations. And yet, their recommendations are conflicting. Are they conflicting due to manner in which the research evidence was accessed, extracted, synthesised and reported? Which evidence source should then be utilised to inform health care policy and practice given that all three, generally, seem to be on equal ground? Who should make the decision as to which source of evidence should be utilised when there is conflicting evidence? As Socrates in Plato's "Euthyphro”, and in more recent times, Jay-Z in the song "no church in the wild”, state "whose bias do you seek”, which we can interpret as which evidence source should we choose as each evidence source may present with inherent limitations.

This practical example highlights the important issue of conflicting research evidence which may confront health care stakeholders when they engage with research evidence. Given the time and resource limitations that confront health care stakeholders, it is important that when they engage with research evidence the end result is a positive outcome. When there is conflicting research evidence, it is important to understand why this to be the case and unravel the evidence base underpinning conflicting findings. This might be in the form of critical appraisals of scientific rigour of each evidence source (so that health care stakeholders can be informed about the risks of methodological bias) and to assist them in interpreting and clarifying the information presented.1 By doing so, health care stakeholders can then become informed consumers of research and can better understand the nuances underpinning conflicting research evidence.


  1. Hilton, S., Bedford, H., Calnan, M., Hunt, K. 2009. Competency, confidence and conflicting evidence: key issues affecting health visitors' use of research evidence in practice, BMC Nursing, vol.8, no.4, doi: 10.1186/1472-6955-8-4.
  2. Taylor, T., Wilkinson, D., Blue, IA, Dollard, JT. 2002. Evidence-based rural general practice: barriers and solutions in South Australia, Rural and Remote Health, vol.2, no.1, pp.116.
  3. Chaibi, A., Tuchin, P., Russell, MB. 2011. Manual therapies for migraine: a systematic review, Journal of Headache and Pain, vol.12, pp.127-133.
  4. Posadzki, P., Ernst, E. 2011, Spinal Manipulations for Cervicogenic Headaches: A Systematic Review of Randomized Controlled Trials, Headache, vol.51, no.7, pp.1132-9. doi: 10.1111/j.1526-4610.2011.01932.x.
  5. Bryans, R., Descarreaux, M., Duranleau, M., Marcoux , H., Potter, B., Ruegg, R., Shaw, L., Watkin, R., White, E. 2011, Evidence-Based Guidelines For the Chiropratic Treatment of Adults with Headache, Journal of Manipulative and Physiological Therapeutics, vol. 34, pp. 274-289.


Submission Location