D. Guidelines for Reporting Primary and Secondary Research

The conduct of systematic reviews, meta-analysis, and related integrative studies requires systematic examination of the reports of primary data studies as well as other integrative methods. As integrative methods have taken on more central roles in HTA and other forms of evaluation, methodological standards for conducting and reporting these studies have risen (Egger, Smith, Altman 2001; Moher 1999; Petitti 2001; Stroup 2000). In addition to the PRISMA instrument for systematic reviews and meta-analyses noted above, there are other instruments for assessing the reporting of clinical trials, systematic reviews, meta-analyses of trials, meta-analyses of observational studies, and economic analyses. Some of these are listed in Box IV-3. HTA programs that use the inclusion/exclusion rules and other aspects of these instruments are more likely to conduct more thorough and credible assessments. In addition to their primary purpose of improving reporting of research, these guidelines are helpful forplanning studies of these types and in reviewing studies as part of systematic reviews and other integrative methods. See also Research Reporting Guidelines and Initiatives compiled by the US NLM at: http://www.nlm.nih.gov/services/research_report_guide.html.

Box IV-3. Guidelines for Reporting Research

• AMSTAR (Assessment of Multiple Systematic Reviews) (Shea 2009)
• CHEERS (Consolidated Health Economic Evaluation Reporting Standards) (Husereau 2013)
• CONSORT (Consolidated Standards of Reporting Trials) (Turner 2012)
• GRACE (Good ReseArch for Comparative Effectiveness) (Dreyer 2014)
• MOOSE (Meta-analysis of Observational Studies in Epidemiology) (Stroup 2000)
• PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) (Moher 2009)
• QUOROM (Quality Of Reporting Of Meta-analyses) (Moher 1999)
• STARD (Standards for Reporting of Diagnostic Accuracy) (Bossuyt 2003)
• STROBE (Strengthening the Reporting of OBservational Studies in Epidemiology) (von Elm 2008)
• TREND (Transparent Reporting of Evaluations with Nonrandomized Designs) (Des Jarlais 2004)