Article type
Abstract
Background: Results from well conducted and adequately reported randomised trials are considered a gold standard for evaluating effectiveness of interventions. SPIRIT 2013 for trial protocols and CONSORT 2010 for full trial publications are guidelines developed to aid adequate reporting.
Objective: To explore factors associated with variability in adherence to SPIRIT 2013 and CONSORT 2010.
Methods: We searched Embase and MEDLINE for empirical studies of any design investigating factors for variability in adherence to SPIRIT 2013 or CONSORT 2010. We performed additional searches in, for example, Google Scholar and conference proceedings. Two authors independently screened studies, assessed methodological quality, and extracted data. We provided a summary of main findings for each identified factor for variability and used vote counting to present an overview of the direction of effect.
Results: We included 32 studies. Most of the studies had inadequate/unclear methodological quality, as they often had no publicly available protocol, which prevented us from assessing the risk of selective outcome reporting. The studies investigated numerous factors for variability (Table 1). For SPIRIT 2013, 12 factors were investigated. Multicenter trials tended to have better adherence to SPIRIT 2013 (two of two studies). For CONSORT 2010, 22 factors were investigated. Higher journal impact factor tended to improve adherence to CONSORT 2010 (seven of ten studies), as did journal endorsement (two of three studies), more trial citations (four of four studies), larger sample size (five of seven studies), and better methodological quality (four of five studies). The remaining factors were either investigated in single studies based on limited numbers of trial protocols or full publications, or the direction and magnitude of the association with adherence were inconsistent across the studies.
Conclusion: The included studies investigated several factors for variability in adherence to SPIRIT 2013 and CONSORT 2010. Adherence tended to vary slightly according to multicenter status, journal impact factor, journal endorsement, trial citations, trial sample size, and methodological quality.
Impact on patients: Randomised trials often have profound impact on patient care. Adequate reporting is a prerequisite for adequate implementation of trial findings; it is therefore important to understand how reporting can be improved.
Objective: To explore factors associated with variability in adherence to SPIRIT 2013 and CONSORT 2010.
Methods: We searched Embase and MEDLINE for empirical studies of any design investigating factors for variability in adherence to SPIRIT 2013 or CONSORT 2010. We performed additional searches in, for example, Google Scholar and conference proceedings. Two authors independently screened studies, assessed methodological quality, and extracted data. We provided a summary of main findings for each identified factor for variability and used vote counting to present an overview of the direction of effect.
Results: We included 32 studies. Most of the studies had inadequate/unclear methodological quality, as they often had no publicly available protocol, which prevented us from assessing the risk of selective outcome reporting. The studies investigated numerous factors for variability (Table 1). For SPIRIT 2013, 12 factors were investigated. Multicenter trials tended to have better adherence to SPIRIT 2013 (two of two studies). For CONSORT 2010, 22 factors were investigated. Higher journal impact factor tended to improve adherence to CONSORT 2010 (seven of ten studies), as did journal endorsement (two of three studies), more trial citations (four of four studies), larger sample size (five of seven studies), and better methodological quality (four of five studies). The remaining factors were either investigated in single studies based on limited numbers of trial protocols or full publications, or the direction and magnitude of the association with adherence were inconsistent across the studies.
Conclusion: The included studies investigated several factors for variability in adherence to SPIRIT 2013 and CONSORT 2010. Adherence tended to vary slightly according to multicenter status, journal impact factor, journal endorsement, trial citations, trial sample size, and methodological quality.
Impact on patients: Randomised trials often have profound impact on patient care. Adequate reporting is a prerequisite for adequate implementation of trial findings; it is therefore important to understand how reporting can be improved.