Poor Validation of Medical Record ICD-9 Diagnoses of Gout in a Veterans Affairs Database
Bibliographic record
Abstract
OBJECTIVE: Diagnostic codes based on medical records or claims data have been used to identify patient populations with gout for important epidemiologic and clinical studies. We evaluated whether we can document the accuracy of such diagnoses by review of medical records and then on direct interviews with a subset of patients. METHODS: Electronic medical records of 289 patients with 2 visits with ICD-9 codes for gout were extensively reviewed to search for documentation of features that would classify patients as having gout by 3 sets of proposed criteria, the American College of Rheumatology (ACR), New York, or Rome criteria. Records of patients who had been seen by rheumatologists were compared with all others. A subset of patients seen in clinic were directly interviewed for comparison with the results from the records. RESULTS: Based on medical records review there was documentation of gout by the ACR criteria in only 36%, Rome criteria in 30%, and New York criteria in 33%. Records of patients who had seen rheumatologists had better documentation of classification features. Interview in clinic of 37 patients also improved documentation of the 3 sets of criteria features of gout in 65%-81% of those with ICD-9 codes for gout. CONCLUSION: We found it difficult to confirm ICD-9 coded diagnoses of gout using currently available proposed criteria from details recorded in medical records. This may reflect a problem with available criteria and with documentation. Direct interview of patients may be needed to confirm the presence of typical features when high specificity is desired.
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 | 0.046 | 0.234 |
| Meta-epidemiology (narrow) | 0.000 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.004 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.001 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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; 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".