Retrospective use of the Pragmatic-Explanatory Continuum Indicator Summary-2 trial design tool to assess design choices in randomized controlled trials: an empirical review
Bibliographic record
Abstract
BACKGROUND AND OBJECTIVE: The Pragmatic-Explanatory Continuum Indicator Summary-2 (PRECIS-2) tool has been widely used to help investigators design randomized trials, facilitating the task of aligning design choices with an explanatory or pragmatic primary trial intention. PRECIS-2 is increasingly being used to retrospectively assess the degree of pragmatism or explanatoriness among published trials within reviews. There is little information on the interrater reliability of the tool and no consensus on the preferred method of achieving an accurate and reliable judgment of trial "pragmatism" when using PRECIS-2 retrospectively. The aims of this study were to assess the level of pragmatism or explanatoriness of trials that cite PRECIS-2 and to assess interrater reliability of PRECIS-2 using different scoring approaches. We compared agreement between two independent ratings within a single pair with agreement between consensus scores reached by two independent pairs of reviewers and whether widening the agreement criteria increased interrater reliability. METHODS: Thirty randomized controlled trials (RCTs) were randomly selected from trials citing the PRECIS-2 tool. Two pairs of reviewers, a clinician paired with a methodologist in each case, were trained and independently scored each trial and reached a consensus score within pairs. Agreement between reviewers within pairs and between consensus scores across pairs was assessed using kappa statistics for each of the nine PRECIS-2 domains. RESULTS: RCTs citing PRECIS-2 had predominantly pragmatic design features. Interrater reliability within pairs was low across all domains, with the highest levels found in the two domains of analysis (0.32) and follow-up (0.33). Agreement across pairs on the consensus scores was similarly low. Agreement between reviewers and reviewer pairs was above 70% when agreement was reclassified as "within 1-point difference on the scoring scale" for eight domains, but no improvement was obtained for the remaining domain. CONCLUSION: Trials citing PRECIS-2 tend to have predominantly pragmatic design features. When using PRECIS-2 to retrospectively score trial publications, agreement between consensus scores across pairs of reviewers was no better than agreement within pairs. Reconfiguring the PRECIS scoring scale and improving scoring guidance may provide a more meaningful, easily interpreted measure of "pragmatism" for trialists wishing to use PRECIS-2 as a review tool. PLAIN LANGUAGE SUMMARY: The Pragmatic-Explanatory Continuum Indicator Summary-2 (PRECIS-2) tool is designed to help researchers match their design decisions to the intended purpose of their trial. The intention of a trial can be "explanatory," which improves our understanding of how an intervention works, or "pragmatic," which supports decision-making in health care. Increasingly, the tool has been used for a secondary purpose: in systematic reviews. Here the tool is used to judge the level of "pragmatism" or "explanatoriness" of trials included in the review to aid the understanding of trial results. However, there is debate on the most reliable means of making this judgment. Sometimes judgements are made using one reviewer; other times, multiple reviewers. Our study evaluated interrater reliability of two methods of scoring trial publications using PRECIS-2: individual reviewer scores and pairs of reviewers agreeing on a consensus score. We also found that neither method we tested produced a reliable judgment using PRECIS-2, and the scores from two reviewers agreeing on a consensus were no more reliable than scores from a single reviewer. We performed an additional analysis that showed that simplifying the scoring from the original five-point scale to a three-point scale may give a more reliable judgment of the "pragmatism" or "explanatioriness" of published trials. This simpler method of scoring should be encouraged for retrospective use of PRECIS-2 in systematic reviews.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.585 | 0.853 |
| Meta-epidemiology (narrow) | 0.002 | 0.003 |
| Meta-epidemiology (broad) | 0.006 | 0.008 |
| Bibliometrics | 0.013 | 0.016 |
| Science and technology studies | 0.001 | 0.004 |
| Scholarly communication | 0.008 | 0.009 |
| Open science | 0.004 | 0.004 |
| Research integrity | 0.004 | 0.003 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; the direct Gemma label and the distilled Codex classifier agree on what is shown here.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".