Seon-Young Lee1, Penelope J Teoh, Christian F Camm, Riaz A Agha. 1. From the Southampton Medical School (S.-Y.L., P.J.T.), Southampton; New College (C.F.C.), University of Oxford, Oxford; and Department of Plastic Surgery (R.A.A.), Stoke Mandeville Hospital, Aylesbury, Buckinghmanshire, United Kingdom.
Abstract
BACKGROUND: Randomized controlled trials (RCTs) are the criterion standard for assessing new interventions. However, bias can result from poor reporting, which also makes critical appraisal and systematic review challenging. The Consolidated Standards of Reporting Trials (CONSORT) criteria for nonpharmacological trials published in 2008 provided a set of 23 mandatory items that should be reported in an RCT. This is the first study to assess the compliance of RCTs in trauma with the CONSORT criteria for nonpharmacological trials. METHODS: The MEDLINE database was searched using the MeSH term wounds and injuries for English-language articles published between January 2009 and December 2011. Relevant articles were scored by two reviewers and compared against surrogate markers of article quality (such as journal impact factor). RESULTS: Eighty-three articles were deemed suitable for inclusion. The mean CONSORT score was 11.2 of 23 items (49%; range, 3.38-18.17). Compliance was poorest for items relating to the adherence of care providers (0%), abstract (5%), and implementation of randomization (6%). Only 40% declared conflicts of interest, 73% declared permission from an ethics review committee, 43% declared sources of funding, and 10% stated a trial registry number. There was a significant correlation between the CONSORT score and the impact factor of the publishing journal (ρ = 0.37, p = 0.0006) but not for the number of patients or authors or single versus multicentre trials CONCLUSION: The reporting quality of RCTs in trauma surgery needs improvement. We suggest ways by which this could be improved including the following: better education, awareness, and a cohesive strategy among all stakeholders and the hard wiring of compliance through electronic journal submission systems.
BACKGROUND: Randomized controlled trials (RCTs) are the criterion standard for assessing new interventions. However, bias can result from poor reporting, which also makes critical appraisal and systematic review challenging. The Consolidated Standards of Reporting Trials (CONSORT) criteria for nonpharmacological trials published in 2008 provided a set of 23 mandatory items that should be reported in an RCT. This is the first study to assess the compliance of RCTs in trauma with the CONSORT criteria for nonpharmacological trials. METHODS: The MEDLINE database was searched using the MeSH term wounds and injuries for English-language articles published between January 2009 and December 2011. Relevant articles were scored by two reviewers and compared against surrogate markers of article quality (such as journal impact factor). RESULTS: Eighty-three articles were deemed suitable for inclusion. The mean CONSORT score was 11.2 of 23 items (49%; range, 3.38-18.17). Compliance was poorest for items relating to the adherence of care providers (0%), abstract (5%), and implementation of randomization (6%). Only 40% declared conflicts of interest, 73% declared permission from an ethics review committee, 43% declared sources of funding, and 10% stated a trial registry number. There was a significant correlation between the CONSORT score and the impact factor of the publishing journal (ρ = 0.37, p = 0.0006) but not for the number of patients or authors or single versus multicentre trials CONCLUSION: The reporting quality of RCTs in trauma surgery needs improvement. We suggest ways by which this could be improved including the following: better education, awareness, and a cohesive strategy among all stakeholders and the hard wiring of compliance through electronic journal submission systems.
Authors: Riaz Agha; Alexander J Fowler; Seon-Young Lee; Buket Gundogan; Katharine Whitehurst; Hakiran Sagoo; Kyung Jin Lee Jeong; Douglas G Altman; Dennis P Orgill Journal: BMJ Open Date: 2015-10-05 Impact factor: 2.692
Authors: Jean Joel R Bigna; Jean Jacques N Noubiap; Serra Lem Asangbeh; Lewis N Um; Paule Sandra D Sime; Elvis Temfack; Mathurin Cyrille Tejiokem Journal: BMC Med Res Methodol Date: 2016-10-13 Impact factor: 4.615