Chiropractic has been defined the “diagnosis, treatment and prevention of mechanical disorders of the musculoskeletal system and the effects of these disorders on the function on the nervous system and general health”.1 The hallmark intervention of chiropractic is spinal manipulation which is used to adjust spinal “subluxations”. The founder of chiropractic believed that 95% of all diseases are caused by vertebral subluxations.2
Many authors have expressed doubts about the safety of spinal manipulation. A particular concern relates to vascular accidents caused by arterial dissection after upper spinal manipulation.3,4 The main aim of this review is to examine the reporting of adverse effects of chiropractic manipulations in recent randomised clinical trials (RCTs).
Sixty RCTs met our eligibility criteria.5–64 Twenty-nine RCTs (48%) did not report information regarding AEs6,9,13–20,26,29,32,36,40,41,45–49,53,56,60,62,64 and 31 RCTs (52%) did. Of those 31 RCTs, 16 (51.6%) reported that no adverse effect of chiropractic manipulations occurred during the study.5,9,18,20,21,23,24,26,30,38,42–44,59,60,63 Eight of the 31 RCTs mentioning AEs described the method of establishing AEs: interview,31,33 questionnaire,23 questionnaire and telephone interview,38 self-reports,39 diary,61 standardised form,8 and notes.9 Eleven RCTs mentioned the severity of AEs.7,8,17,19,27,29,31,33,34,46,51 Four RCTs reported the duration of AEs.27,29,34,51 One RCT reported the frequency of AEs.38 One RCT (1.6%) reported complete information on AEs, i.e. method of reporting, severity, duration and frequency.38
The source of funding was provided in 47 RCTs (78.4%%)5–10,12,14,15,17–22,24–29,31–37,40–46,51,52,54–60,62–64 and in 13 RCTs (21.6%) this information was missing 11,13,16,23,30,38,39,47–50,53,61 Conflicts of interest were declared in 17 RCTs (28.3%)7,24,27,29–31,37,40,41,43,45,57–60,63,64 and in 43 RCTs (71.7%) this information was missing5–12,16–20,22–51
Table 3 summarises the findings related to AEs as a function of source of funding, conflict of interest and affiliation to chiropractic institutions. Sizable proportions (55.8%) of RCTs without a declaration of interest also failed to mention AEs. The method of reporting AEs was not mentioned in large proportions of RCTs without a statement of conflict of interest (86%) or with affiliations to chiropractic institutions (91.8%).
The severity of AEs was not reported in 83.7% of RCTs without declarations of conflict of interest, and in 81% with affiliations to chiropractic institutions. The duration of AEs was not reported in 95.3% of RCTs without declarations of conflict of interest, and in 91.8% of RCTs with affiliations to chiropractic institutions. The frequency of AEs was not reported in 95.3% of RCTs without declarations of conflict of interest and in 100% with affiliations to chiropractic institutions.
A remarkably low number of RCTs of chiropractic manipulation was published during the last decade: 60 RCTs compared to nearly 300 000 RCTs that emerged in the conventional healthcare during the same period.65 The reasons for this paucity of RCTs might be complex and could involve a shortage of funding and a general lack of insight by chiropractors into the necessity of submitting their therapeutic claims to scientific tests.66
Forty eight per cent of RCTs reviewed here fail to mention AEs and sizable proportions of those which do mention AEs provide no information as to how AEs were recorded, their severity, duration or frequency (Table 2). For instance, of the 31 RCTs reporting AEs, only 8 (25.8%) mentioned the method of reporting, 11 (35.4%) the severity, 4 (12.9%) the duration, one (3.2%) the frequency of AEs. Further analyses (Table 3) seem to suggest that not declaring conflicts of interest and being affiliated to chiropractic institutions might be risk factors for incomplete reporting of AEs.
Guidelines of reporting or designing RCTs strongly emphasise that details on AEs are an ethical imperative in clinical research.e.g.67 Our review seems to indicate that, in chiropractic research, this imperative is frequently ignored. Similarly, the source of funding and any conflicts of interest should be declared in publications of RCTs.67
Our review shows that, in chiropractic research, this is frequently not the case. Our analyses also suggest that being affiliated to a chiropractic institution, arguably a conflict of interest in itself,68 is associated with poor reporting of AEs. Twelve of the RCTs were funded by the US National Center for Complementary and Alternative Medicine,10,11,19,20-22,25,32-35 and 19 by chiropractic organisations.13,17,18,22,32,33,35–38,45–47,51,56–60 Of those funded by chiropractors, 6 (31%) failed to report AEs.64 13,17,22,51,60 Forty-three (71.7%) RCTs failed to report conflicts of interest. Similar deficits in ethical standards have also been noted in other areas of alternative medicine research.69
Several hundred severe complications after upper spinal manipulations have been reported.e.g.70,71 The estimates as to the incidence of these complications vary hugely.72 The opinion of most chiropractors that such complications are extreme rarities is partly based on the fact that clinical trials of chiropractic manipulation fail to demonstrate the existence of such events. Our review shows that authors of such RCTs frequently neglect to mention AEs. Thus the lack of trial evidence for severe complications could well be due to the failure of triallists to report AEs.
Our review also points to an overt contradiction regarding the incidence of mild to moderate AEs. Sixteen of the 31 RCTs mentioning AEs stated that no AEs occurred and the RCTs which did report AEs provided incidence figures between 1.4%21and 50% 61 Numerous prospective studies specifically designed to investigate AEs of chiropractic manipulation agree that about 50% of patients experience mild to moderate AEs after such treatments.72 Method of monitoring AEs in RCTs strongly influences their incidence.73,74 Thus lack of rigorous methods for assessing AEs might have generated spuriously low incidence figures in RCTs.
In conclusion, this review shows that AEs are poorly reported or not mentioned at all in RCTs of chiropractic manipulations. Further concerns relate to the prevalent failure to report conflicts of interest or sources of funding.