Delirium diagnosis without a gold standard: Evaluating diagnostic accuracy of combined delirium assessment tools
Bibliographic record
Abstract
BACKGROUND: Fluctuating course of delirium and complexities of ICU care mean delirium symptoms are hard to identify or commonly confused with other disorders. Delirium is difficult to diagnose, and clinicians and researchers may combine assessments from multiple tools. We evaluated diagnostic accuracy of different combinations of delirium assessments performed in each enrolled patient. METHODS: Data were obtained from a previously conducted cross-sectional study. Eligible adult patients who remained admitted to ICU for >24 hours with at least one family member present were consecutively enrolled as patient-family dyads. Clinical delirium assessments (Intensive Care Delirium Screening Checklist [ICSDC] and Confusion Assessment Method-ICU [CAM-ICU]) were completed twice daily by bedside nurse or trained research assistant, respectively. Family delirium assessments (Family Confusion Assessment Method and Sour Seven) were completed once daily by family members. We pooled all delirium assessment tools in a single two-class latent model and pairwise (i.e., combined, clinical or family assessments) Bayesian analyses. RESULTS: Seventy-three patient-family dyads were included. Among clinical delirium assessments, the ICDSC had lower sensitivity (0.72; 95% Bayesian Credible [BC] interval 0.54-0.92) and higher specificity (0.90; 95%BC, 0.82-0.97) using Bayesian analyses compared to pooled latent class analysis and CAM-ICU had higher sensitivity (0.90; 95%BC, 0.70-1.00) and higher specificity (0.94; 95%BC, 0.80-1.00). Among family delirium assessments, the Family Confusion Assessment Method had higher sensitivity (0.83; 95%BC, 0.71-0.92) and higher specificity (0.93; 95%BC, 0.84-0.98) using Bayesian analyses compared to pooled latent class analysis and the Sour Seven had higher specificity (0.85; 95%BC, 0.67-0.99) but lower sensitivity (0.64; 95%BC 0.47-0.82). CONCLUSIONS: Results from delirium assessment tools are often combined owing to imperfect reference standards for delirium measurement. Pairwise Bayesian analyses that explicitly accounted for each tool's (performed within same patient) prior sensitivity and specificity indicate that two combined clinical or two combined family delirium assessment tools have fair diagnostic accuracy.
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 | 0.060 | 0.146 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.003 | 0.004 |
| Bibliometrics | 0.007 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.002 | 0.004 |
| Research integrity | 0.002 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
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".