Trish Groves: Let SPIRIT take you … towards much clearer trial protocols

Trish Groves
Reporting statements like the CONSORT and STROBE statements are making an important and demonstrable difference to the quality of research papers by helping authors report exactly what happened in their studies. But these statements can’t fix studies that were inadequately designed or poorly conducted. They’re the ambulance at the bottom of the cliff, rather than the fence at the top (apologies to Fiona Godlee, from whom I pinched this analogy).

A new reporting statement, SPIRIT, is coming to the rescue, at least for trials. It stands for Standardized Protocol Items for Randomized Trials and it’s been developed using evidence from two systematic reviews and a robust consensus process. In its current draft SPIRIT’s checklist comprises 34 essential items to ensure that researchers write and register their trial protocols with scientific, ethical, and professional integrity and transparency. It’s a minimum yet comprehensive dataset. It should be published in 2010, hopefully earlier rather than later.

The SPIRIT checklist includes scientific items that closely mirror the latest version of CONSORT (also coming soon) and it also incorporates the information required in existing general guidance, for example from the International Conference on Harmonisation (ICH), and the WHO International Clinical Trials Registry Platform minimum dataset. And, importantly, SPIRIT prompts authors to explicitly report in the protocol their plans for gaining consent, protecting confidentiality (including for further use or sharing of trial data), seeking funding and sponsorship, declaring competing interests, disseminating results, and publication.

Because the SPIRIT statement is not yet published , there’s not yet much information in the public domain about it. But here’s a summary from SPIRIT’s chair, An-Wen Chan at the University of Toronto. This ends with a section on impact. I’d add that, if SPIRIT stops meaningless or unethical trials before they start it will save time, money, and even lives.

Trish Groves, deputy editor BMJ

Competing interests: I participated in the two consensus meetings and Delphi exercise that developed SPIRIT, and the BMJ’s senior statistics editor Doug Altman ( ) is one of its originators.

Summary of the SPIRIT initiative
The protocol for a randomized trial serves as the origin of study conduct and reporting. With recent international policies and legislation mandating public access to information from trial protocols, these documents have become increasingly important for transparency and interpretation of results. Only with adequate appraisal of trial methods described in protocols can the study results be properly evaluated and placed into context.

However, the quality of protocols varies greatly, partly due to variable standards and guidelines. Previous studies have demonstrated that a high percentage of trial protocols do not address important issues such as primary outcomes, data analysis, competing interests, and publication restrictions.

The international SPIRIT initiative will produce evidence-based guidance for key items to be addressed in trial protocols, leading to improved quality of protocols and enabling accurate interpretation of trial results. The SPIRIT methodology includes two systematic reviews of existing protocol guidelines and empiric evidence, as well as a Delphi consensus process (3 survey rounds and two expert meetings).

The SPIRIT initiative has distinct strengths:

  • Evidence-based approach with reference to systematically-identified empiric data
  • Broad international representation, consultation, and endorsement across key stakeholder groups
  • Methodology based on the latest recommendations for developing reporting guidelines
  • Adherence to modern principles of transparency
  • Pilot-testing coupled with integration of an evaluative component to assess future impact.

SPIRIT Checklist – Section headings
Administrative information
Methods: Participants and interventions
Methods: Assignment of interventions
Methods: Data collection, management, analysis
Methods: Trial monitoring
Ethical considerations & dissemination

The evidence-based SPIRIT recommendations will benefit researchers, trial participants, sponsors, funders, research ethics committees, peer reviewers, trial registries, journals, policymakers and other key stakeholders by standardizing the core content of trial protocols and improving their quality; facilitating critical appraisal of trials; and ultimately enhancing transparency in clinical trials research.