The reproducibility of psychiatric evaluations of work disability: two reliability and agreement studies
Bibliographic record
Abstract
Expert psychiatrists conducting work disability evaluations often disagree on work capacity (WC) when assessing the same patient. More structured and standardised evaluations focusing on function could improve agreement. The RELY studies aimed to establish the inter-rater reproducibility (reliability and agreement) of ‘functional evaluations’ in patients with mental disorders applying for disability benefits and to compare the effect of limited versus intensive expert training on reproducibility. We performed two multi-centre reproducibility studies on standardised functional WC evaluation (RELY 1 and 2). Trained psychiatrists interviewed 30 and 40 patients respectively and determined WC using the Instrument for Functional Assessment in Psychiatry (IFAP). Three psychiatrists per patient estimated WC from videotaped evaluations. We analysed reliability (intraclass correlation coefficients [ICC]) and agreement (‘standard error of measurement’ [SEM] and proportions of comparisons within prespecified limits) between expert evaluations of WC. Our primary outcome was WC in alternative work (WCalternative.work), 100–0%. Secondary outcomes were WC in last job (WClast.job), 100–0%; patients’ perceived fairness of the evaluation, 10–0, higher is better; usefulness to psychiatrists. Inter-rater reliability for WCalternative.work was fair in RELY 1 (ICC 0.43; 95%CI 0.22–0.60) and RELY 2 (ICC 0.44; 0.25–0.59). Agreement was low in both studies, the ‘standard error of measurement’ for WCalternative.work was 24.6 percentage points (20.9–28.4) and 19.4 (16.9–22.0) respectively. Using a ‘maximum acceptable difference’ of 25 percentage points WCalternative.work between two experts, 61.6% of comparisons in RELY 1, and 73.6% of comparisons in RELY 2 fell within these limits. Post-hoc secondary analysis for RELY 2 versus RELY 1 showed a significant change in SEMalternative.work (− 5.2 percentage points WCalternative.work [95%CI − 9.7 to − 0.6]), and in the proportions on the differences ≤ 25 percentage points WCalternative.work between two experts (p = 0.008). Patients perceived the functional evaluation as fair (RELY 1: mean 8.0; RELY 2: 9.4), psychiatrists as useful. Evidence from non-randomised studies suggests that intensive training in functional evaluation may increase agreement on WC between experts, but fell short to reach stakeholders’ expectations. It did not alter reliability. Isolated efforts in training psychiatrists may not suffice to reach the expected level of agreement. A societal discussion about achievable goals and readiness to consider procedural changes in WC evaluations may deserve considerations.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.006 | 0.003 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 teacher head, 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".