Can we accurately identify chemotherapy-related acute care visits in administrative data?
Bibliographic record
Abstract
185 Background: Administrative data is increasingly being used to study treatment related complications that lead to acute care visits such as emergency department visits or hospitalizations (ED+H). We evaluated the accuracy of diagnosis codes for identifying chemotherapy related acute care visits (CRVs) among women with breast cancer. Methods: We prospectively developed algorithms to identify CRVs from administrative data in women receiving adjuvant chemotherapy for breast cancer in Ontario, Canada. Sensitivity (SN) and specificity (SP) were calculated for 3 scenarios: chemotherapy related ED visit, chemotherapy related H, and febrile neutropenia (FN) related visit using the chart as the gold standard. Since there is no specific diagnosis code for FN, three definitions of FN were considered: general (defined as fever or infection or neutropenia as main reason for visit), moderate (neutropenia as main reason for visit) or strict (fever or infection plus neutropenia). The population based cohort was generated by linking several health databases to identify women who had at least one ED+H during adjuvant chemotherapy for breast cancer between 2007-2009. The validation cohort consisted of 490 randomly selected cases from this cohort. Results: The population-based cohort consisted of 8,359 patients of whom 43.4% had at least one ED+H including 1,496 women who had multiple visits resulting in 6,293 unique ED+H. Of these, 73.1% were considered CRVs based on our algorithm. The algorithm performed well in identifying CRVs that included an H either from ED (SN 90%, SP 100%) or directly from home (SN 91%, SP 93%) but less well for ED visits that did not result in H (SN 65%, SP 80%). Depending on which FN algorithm was used, 1.4-24% of visits were considered FN related. The general FN algorithm had excellent SN regardless of whether the visit involved H (94-98%) but SP was moderate (66-80%). The strict FN algorithm had good SP (78-99%) but SN was highly variable (13-89%). The moderate FN algorithm provided the best tradeoff between SN (69-97%) and SP (76-98%). Conclusions: CRVs can be identified from administrative data with reasonable confidence, obviating the need for chart abstraction to evaluate chemotherapy related serious events.
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 | 0.026 | 0.165 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.005 | 0.007 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.002 |
| Open science | 0.002 | 0.001 |
| 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".