Abstract P3-16-02: Enhancing impact and efficiency of financial support for cancer through use of AI: A nascent initiative of Indian cancer society
Bibliographic record
Abstract
Abstract Background: Since 2011, Indian Cancer Society-Cancer Cure Fund (ICS-CCF), established through donations of mutual fund dividends, has funded $27 million in treatment costs for 9800 underprivileged patients. For governance, a Due Diligence Team (DDT) of cancer experts from academic tertiary centers reviewed every planned treatment and authorized covering the treatment costs for evidence-based care. Since expert time is scarce, in 2019, ICS-CCF introduced NAVYA AI (AI) system to scale the prior authorization operation. AI is a clinically validated system that matches clinical data with evidence and expert recommended treatment options. The system can adapt to institutional guidelines such as ICS-CCF authorization criteria on published survival rates. In this prospective study, we explore application of AI in prior authorization, by assessing concordance with DDT expert decisions and potential time savings. Methods: Clinical data of beneficiaries were input in AI and if the output treatment options matched the planned treatment, the application was authorized, otherwise rejected. If AI did not output a decision, the application was deferred to DDT. Concordance between AI and DDT decisions, and DDT time spent on deferred applications, weremeasured. Results: From April 22nd, 2020 to June 9th, 2021, 1994 beneficiary applications were simultaneously reviewed by AI and DDT, of which 601 (30%) were breast cancer patients. AI “authorized” (1563/1994) or “rejected” (16/1994), 79.18% (1579/1267) of applications; and “deferred” 20.81% (415/1994) of applications. AI and DDT were 99.37% (1553/1563) concordant to authorize and 100% (16/16) concordant to reject a beneficiary application. DDT spent an average of 3.4 minutes/deferred application. Conclusions: A clinically validated AI system, 99% concordant with cancer experts in authorizing or rejecting 79% of beneficiary applications, promises a fivefold decrease in experts’ time, the scarcest resource for preauthorization. Subsequent to this study, ICS-CCF is awarding financial support directly based on AI authorize decisions. Similarly, health insurance organizations may consider a prospective study of the AI system in raising quality and efficiency of prior authorization thereby saving cost of time, money and expertise. Citation Format: Tushar Vora, Nehal Khanna, Anant Gokarn, Ann Rawat, Naresh Ramarajan, Gitika Srivastava, Usha Thorat. Enhancing impact and efficiency of financial support for cancer through use of AI: A nascent initiative of Indian cancer society [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P3-16-02.
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 | 0.025 | 0.057 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.003 | 0.006 |
| Science and technology studies | 0.002 | 0.002 |
| Scholarly communication | 0.006 | 0.002 |
| Open science | 0.003 | 0.007 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.012 | 0.002 |
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".