Searching for trial protocols: A comparison of methods

Abstract

INTRODUCTION: Registration and publication of trial protocols has become increasingly important and a requirement in some sources of funding and publication. Increased access to protocols yields many potential benefits, but there are issues regarding identification of published protocols. The aim of this investigation is to compare methods of retrieval for identifying trial protocols in a systematic review. METHODS: Six stages of searching (checking published trial reports, searching journal Web sites, Internet searching, trial registers, bibliographic databases, and contact with authors) were completed to identify 74 trial protocols. RESULTS: Fifty-seven percent of the trial protocols were identified upon completion of all 6 stages of searching. The most comprehensive method was searching trial registers that identified 51% of the protocols. Contact with authors was most effective at uniquely identifying protocols; 12% were retrieved via this single method. Contact with authors was the only effective method of identifying protocols for trials pre-2005. DISCUSSION: When attempting to identify trial protocols to include in systematic reviews, some methods are relatively quick to undertake but deliver a low yield. The most effective search strategy for most sources was retrieval by trial registration number where available. CONCLUSIONS: For protocols of trial results published pre-2005, review authors should contact authors as a priority. For protocols post-2005, they should check the trial publication for protocol details, search trial registers, and contact authors, ceasing searching once a predetermined point of diminishing returns has been reached.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 6 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.