ChatGPT for assessing risk of bias of randomized trials using the RoB 2.0 tool: A methods study
Bibliographic record
Abstract
Abstract Background Internationally accepted standards for systematic reviews necessitate assessment of the risk of bias of primary studies. Assessing risk of bias, however, can be time- and resource-intensive. AI-based solutions may increase efficiency and reduce burden. Objective To evaluate the reliability of OpenAI’s ChatGPT for performing risk of bias assessments of randomized trials using the revised risk of bias tool for randomized trials (RoB 2.0). Methods We sampled recently published Cochrane systematic reviews of medical interventions (up to October 2023) that included randomized controlled trials and assessed risk of bias using the Cochrane-endorsed RoB 2.0. From each eligible review, we collected data on the risk of bias assessments for the first three reported outcomes. Using ChatGPT-4, we assessed the risk of bias for the same outcomes using three different prompts: a minimal prompt including limited instructions, a maximal prompt with extensive instructions, and an optimized prompt that was designed to yield the best risk of bias judgements. The agreement between ChatGPT’s assessments and those of Cochrane systematic reviewers was quantified using weighted kappa statistics. Results We included 34 systematic reviews with 157 unique trials. We found the agreement between ChatGPT and systematic review authors for assessment of overall risk of bias to be 0.16 (95% CI: 0.01 to 0.3) for the maximal ChatGPT prompt, 0.17 (95% CI: 0.02 to 0.32) for the optimized prompt, and 0.11 (95% CI: −0.04 to 0.27) for the minimal prompt. For the optimized prompt, agreement ranged between 0.11 (95% CI: −0.11 to 0.33) to 0.29 (95% CI: 0.14 to 0.44) across risk of bias domains, with the lowest agreement for the deviations from the intended intervention domain and the highest agreement for the missing outcome data domain. Conclusion Our results suggest that ChatGPT and systematic reviewers only have “slight” to “fair” agreement in risk of bias judgements for randomized trials. ChatGPT-4 cannot be relied upon to judge risk of bias and should not be used for this purpose. Since this study was completed, ChatGPT has advanced and OpenAI has released newer models with different capabilities. These may prove more adept at risk of bias evaluation. There may also be opportunities to use ChatGPT to streamline other aspects of systematic reviews, such as screening of search records or collection of data.
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 | 0.502 | 0.800 |
| Meta-epidemiology (narrow) | 0.006 | 0.006 |
| Meta-epidemiology (broad) | 0.015 | 0.037 |
| Bibliometrics | 0.027 | 0.021 |
| Science and technology studies | 0.002 | 0.005 |
| Scholarly communication | 0.008 | 0.007 |
| Open science | 0.005 | 0.009 |
| Research integrity | 0.006 | 0.006 |
| Insufficient payload (model declined to judge) | 0.024 | 0.003 |
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".