Quality of reporting of randomised controlled trials in chiropractic using the CONSORT checklist

BACKGROUND Reviews indicate that the quality of reporting of randomised controlled trials (RCTs) in the medical literature is less than optimal, poor to moderate, and require improving. However, the reporting quality of chiropractic RCTs is unknown. As a result, the aim of this study was to assess the reporting quality of chiropractic RCTs and identify factors associated with better reporting quality. We hypothesized that quality of reporting of RCTs was influenced by industry funding, positive findings, larger sample sizes, latter year of publication and publication in non-chiropractic journals. METHODS RCTs published between 2005 and 2014 were sourced from clinical trial registers, PubMed and the Cochrane Reviews. RCTs were included if they involved high-velocity, low-amplitude (HVLA) spinal and/or extremity manipulation and were conducted by a chiropractor or within a chiropractic department. Data extraction, and reviews were conducted by all authors independently. Disagreements were resolved by consensus. OUTCOMES a 39-point overall quality of reporting score checklist was developed based on the CONSORT 2010 and CONSORT for Non-Pharmacological Treatments statements. Four key methodological items, based on allocation concealment, blinding of participants and assessors, and use of intention-to-treat analysis (ITT) were also investigated. RESULTS Thirty-five RCTs were included. The overall quality of reporting score ranged between 10 and 33 (median score 26.0; IQR = 8.00). Allocation concealment, blinding of participants and assessors and ITT analysis were reported in 31 (87 %), 16 (46 %), 25 (71 %) and 21 (60 %) of the 35 RCTs respectively. Items most underreported were from the CONSORT for Non-Pharmacological Treatments statement. Multivariate regression analysis, revealed that year of publication (t32 = 5.17, p = 0.000, 95 % CI: 0.76, 1.76), and sample size (t32 = 3.01, p = 0.005, 95 % CI: 1.36, 7.02), were the only two factors associated with reporting quality. CONCLUSION The overall quality of reporting RCTs in chiropractic ranged from poor to excellent, improving between 2005 and 2014. This study suggests that quality of reporting, was influenced by year of publication and sample size but not journal type, funding source or outcome positivity. Reporting of some key methodological items and uptake of items from the CONSORT Extension for Non-Pharmacological Treatments items was suboptimal. Future recommendations were made.