Bibliographic record
Abstract
The recent report by Pedemonte et al.1 of their substudy of the Minimizing ICU Neurologic Dysfunction with Dexmedetomidine-induced Sleep (MINDDS) study2 emphasized the relationship between electroencephalogram (EEG) burst-suppression during cardiopulmonary bypass and delirium in elderly patients undergoing cardiac surgery. It raises several important points regarding the potential for cerebral monitoring to identify patients who may be at risk for significant postoperative neurologic complications, including delirium and postoperative cognitive dysfunction. However, interpreting these complex relationships requires certain safeguards to minimize the risk of potential false discovery, and thus maximize the confidence in a study’s conclusions. These safeguards include, but are not limited to, clear adherence to the prespecified substudy aims and a priori hypotheses, the development of a data statistical analytic plan before accessing the data, and consideration to the potential moderating effects in the substudy from the intervention of the parent trial. In this case, for example, the data from the substudy were derived from an ongoing randomized controlled trial investigating the potential effects of dexmedetomidine on postoperative delirium. It would seem reasonable then for any analysis in the substudy to be adjusted for the use of dexmedetomidine. Clarification as to whether and how this was done would be useful.Several other aspects of their study might also benefit from additional clarity. For example, adherence of reporting to the ordered prestated hypotheses seems to have been modified. For example, the primary hypothesis stated in their introduction was that “preexisting cognitive impairment accounts for electroencephalogram burst-suppression during CPB.”1 It is curious, then, that the article’s title, and the subsequent analysis and reporting of the study, principally focuses on postoperative delirium as opposed to preexisting cognitive impairment. This is particularly notable because their power analysis states that the “primary objective of the study was to detect the difference in mean preoperative cognitive scores between the burst-suppression and no burst-suppression groups.”1 The current delirium analysis, as they state, was likely underpowered.Although there is a potentially important relationship between preexisting cognitive impairment and delirium, and one that could be plausibly mediated via EEG burst-suppression, the primary analysis reported should have been the relationship between baseline cognition and EEG burst suppression, with the delirium-related analyses being secondary, and/or exploratory, and fully adjusted for multiple comparisons. Indeed, although some mention is made of adjustments to reduce false discovery, it is not clear where and how these were done. Furthermore, as the authors stated that the “data and statistical analyses plans were defined and written after the data were accessed,”1 it is not clear how much data and analyses mining might have been undertaken before these complex analyses were settled on and which results were chosen to be reported. The study’s actual primary objective found that the relationship between preexisting cognition (assessed using the abbreviated Montreal cognitive assessment) and EEG burst-suppression was not statistically significant (P = 0.965 in their table 1).These limitations should not dissuade the reader from considering the potentially important relationships that the authors have described, because they may in fact be quite meaningful. However, without adequate adjustment for the unit of randomization, consideration for the analytical plan being developed after the data was accessed, and the subsequent organization of the results around a hypothesis that was not the primary one, it does raise the question as to whether undue emphasis is being placed on the “positive” results surrounding delirium, as opposed to the “negative” results related to baseline cognition.The author declares no competing interests.
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 | 0.047 | 0.182 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.003 | 0.006 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.003 | 0.006 |
| Scholarly communication | 0.004 | 0.011 |
| Open science | 0.013 | 0.003 |
| Research integrity | 0.035 | 0.052 |
| Insufficient payload (model declined to judge) | 0.010 | 0.009 |
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; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
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".