Clinical trial registries are integral to the world of evidence-based medicine. They provide a database of information, assist in monitoring selective reporting of outcome measures, disclosing publication bias, and preventing duplication of trials. Researchers can also observe the records of non-published trials when investigating a similar theme or hypothesis. These databases are like a super resource and can act as an accountability tool of sorts. Research is as important for chiropractic as it is for any other discipline. Thus, indexing such research on these databases is essential. But how does chiropractic literature stack up? A paper published in the Journal of the Canadian Chiropractic Association has taken aim at this important issue.
In the world of research and academia, there is an adage: publish or perish. The pressure placed on researchers and academics to get articles into journals is very real. Interestingly though, reviews of Randomised Controlled Trials in medical research have revealed a disappointing number of registered trials that altered the primary outcome measures in their publications from thoseoriginally recorded in the clinical trial protocol, and a disappointing rate of unpublished trials [1]. Essentially, we are looking at the biases that exist in publication and the pressure placed on research teams to return statistically significant results. Publish or perish, but you can only publish if it is significant.
Other studies have compared the primary outcomes of initial entries to entries obtaining a completed status and have shown a large percentage (31.7% of registered interventional studies on clinicaltrials.gov between 1999-2012, and 92.5% of completed interventional studies between 1999-2014 that have published results) changed their primary outcome measure. While these studies did not compare between trial registry and subsequent publication, it does highlight the large proportion that were not consistent with their initial goals.
In terms of medical/biomedical studies in existing literature, the following statistics have been reported [1]:
- Coherence between trial registry and trial publication: 38.8% of registered Randomised Controlled Trials had changes to the ‘primary outcome’ field in the publication, demonstrating a discrepancy in completeness and agreement between protocols registered on clinicaltrials.gov and their ICMJE publication.
- Publication rate of trials registered ongov and European Union Clinical Trials Register: 44.4%
The current review of Chiropractic-related studies revealed [1]:
- Coherence between trial registry and publication: primary outcome agreement between clinicaltrials.gov entries and the publish papers was 90.2%
- Publication rate of completed trials: 59.6%
- Rate of displaying results on the clinicaltrials.gov website: 14.6%
In plain terms, chiropractic has far outperformed other modalities in terms of matching trial registration and publication. There has long been a rumour about chiropractic research not performing as well as other modalities of health care, but this piece of research frankly proves it to be false.
How did the researchers arrive at this result? In order to analyse and compare chiropractic care with known publication and registry rates across other modalities, 171 chiropractic-related protocols were found on clinicaltrials.gov using the search terms “chiropractic” (160) and “chiropractor” (11). Some twenty-five of these had results posted on clinicaltrials.gov and one hundred and two were published. (The paper’s authors were essentially comparing the agreement of primary outcome measures between the register and publication.)
But what could be the possible reasons for the more favourable results in chiropractic-related studies?
As the effort and push for significant research into the chiropractic discipline has increased in recent years, we have likely learned from the advances (and even errors) made in the biomedical research community.
Another potential reason is that, as there is less funding provided to the chiropractic profession when compared with the funding provided to the biomedical field, there is less financial pressure to produce statistically significant outcomes.
It is sad to see that only 14.6% of studies had included results on clinicaltrials.gov. The purpose of the registry is to have a database of trials that have and have not been published. It is essentially the master list of clinical trial results, regardless of publication status or success. We lose the information the studies obtained if the results are not published.
The Canadian Paper’s authors made a few recommendations to continue chiropractic’s record of registry and publication coherence [1]. They recommended that chiropractic investigators should continue to complete studies and upload results and seek publication regardless of the study findings.
The remark about publishing regardless of results is an important one in terms of bias. Publishing negative results avoids introducing positive bias into meta-analysis. “Both positive and negative findings are important when evaluating treatments and determining the best care for patients,” the authors remarked [1]. We lose the information the studies obtained if the results are not published. This data then cannot inform future research.
They also suggested that the national clinical trial identifier should be included in published papers to better link with the clinical trial registry. Lastly, they suggested that a future review involving a broader reach of terms to include a greater proportion of chiropractic-related studies would provide more accurate data and a full representation of the chiropractic clinical research.
Reference:
- Coté R, Perle S, Martin D. Agreement of primary outcomes in chiropractic- related clinical trials registered in clinicaltrials.gov with corresponding publication. J Can Chiropr Assoc. 2021. 65(2):207.
