[Defining substance related disorders in administrative health databanks].
Bibliographic record
Abstract
Introduction Epidemiogical surveys in the general population can provide relevant information on substance use and substance-related disorders (SRD). However, because of time and resource constraints, this data is limited in its scope. Health administrative databanks consist of routinely collected data covering a large sample size, often representative of the general population. They allow for further longitudinal analyses of comorbidities patterns and health services utilization over decades in individuals with SRD. Developing algorithms to identify these individuals is crucial before being able to tap into these databanks. Objective To present and to reflect on the methodological process leading to the creation of SRD case definitions in administrative health databanks. Methods The Quebec Integrated Chronic Disease Surveillance System (QICDSS) contains five linked administrative health databanks that are updated annually and covers over 98% of the general population. Codes from the 9th and 10th revisions of the International Classification of Diseases (ICD-9 and ICD-10) were used to define individuals who have a SRD, according to diagnoses made by a physician. First, all ICD codes that could potentially define a SRD were identified through a literature review. Second, relevant codes were selected. Third, case definition algorithms were created by grouping codes that describe a similar concept. These three steps were performed by comparing our codes with previous propositions from other teams, and through group discussions with a committee of experts (one psychiatrist, two general practitioners, one emergency doctor, and two researchers). Results Relevant ICD codes were found in specific chapters on SRD, but also in different sections concerning physical diseases that are induced by substance use or concerning poisoning and intoxication. In total, 89 ICD-9 codes and 197 ICD-10 codes were identified. From this list, codes that were almost never used in the QICDSS, codes that were almost never reported by other research teams, codes that were not specific to substance use, and codes related to tobacco use were all excluded. Codes were first categorized if they were related to alcohol or to another substance. No distinction could be made according to a specific substance, mainly because of imprecision surrounding ICD-9 coding. From this retained list, six case definitions were created: 1) alcohol use disorders (i.e. abuse or dependence); 2) drug use disorders; 3) alcohol induced disorders (i.e. withdrawal, induced psychotic disorders and other mental disorders, physical diseases 100% attributable to alcohol); 4) drug induced disorders; 5) alcohol intoxication; 6) drug intoxication. Discussion and conclusion Although unanimous consensus by the expert committee had to be obtained during code selection and grouping to create these case definitions for SRD, further validation needs to be conducted to determine if these algorithms identify appropriately individuals with SRD. Once tested in other databanks using the ICD system, these case definitions can be used to perform analyses concerning prevalence and incidence, comorbidities patterns and health services utilization to obtain a more complete picture of SRD.
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 | 0.015 | 0.059 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.026 | 0.045 |
| Science and technology studies | 0.002 | 0.001 |
| Scholarly communication | 0.004 | 0.003 |
| Open science | 0.003 | 0.004 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.015 | 0.008 |
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".