Introduction: Since 2011, Indian Cancer Society Cancer Cure Fund (ICS-CCF) has contributed over $24 million to cover the treatment costs of more than 12,000 underprivileged patients across 19 empanelled hospitals in India. A mandate of this philanthropic funding is to carefully choose beneficiaries who have a high chance of cure with guideline based care. From 2011-2021, a team of expert oncologists reviewed every beneficiary application for cure rates/survival. To augment and scale up this process, in February 2021, ICS-CCF evaluated and implemented the use of Artificial Intelligence (AI) in reviewing the applications for recommendation (prior-Authorizations). Navya Al platform is a clinically validated Al model that matches clinical data of beneficiary applicants with available evidence and registry data to predict survival, and generates guidelines based optimal treatment plans. 80% of applications are adjudicated by Navya AI, and the remainder of complex cases are referred to the experts. A goal for implementation of Navya AI is to standardize patient selection for funding, with reproducible rationale for approval based on predicted survival as compared to variable rationale from a rotational volunteer panel of human experts. We hypothesized that Navya AI plus experts would lead to more reproducible survival estimates, leading to better patient selection for funding, and therefore improved survival of the funded cohorts compared to the experts alone. This study evaluates changes in survival of cohorts funded by the ICS-CCF before and after the implementation of Navya AI as a direct measure of success of the funding initiative's mandates.
Methods: As part of routine process, all consecutive hematolymphoid cancer patients who were funded by ICS between January 2018 and December 2023 had been contacted by prospective phone follow to obtain survival outcomes. Given three years since the implementation of Navya AI, all age/risk/diagnosis matched cohorts with 3 years data were examined. Acute Myeloid leukemia (AML) was chosen as a common diagnosis, and exemplary for analysis of the benefit of Navya AI.
Results: 403 adult AML patients (age 18-60 years) were evaluated by ICS for funding between January 2018 and December 2023 with a median follow up of 14 months. Cohorts 3 years before the introduction of Navya AI with experts alone (157/403) was compared to a cohort 3 years after the introduction of Navya AI plus experts (246/403) in a matched pre-post analysis. Notably, Navya AI plus experts was more effective in assessing risk and selecting patients with favorable or intermediate prognosis categories 207/246 (84%) compared to experts alone 123/157 (78%). Navya AI plus experts funded less high risk patients 39/246 (16%) compared to experts alone 34/157 (22%). Correspondingly, the number of patients reported alive in the Navya AI plus experts cohort was higher 195/246 (79%) than experts alone cohort 98/157 (62%) at a median follow up of 14 months. However, 34/157 patients (22%) were unable to be confirmed alive by phone follow up in the experts alone group as compared to only 9/246 (4%) in the Navya AI plus experts group. Confirmed deaths in Navya AI plus expert group 42/246 (17%) and 25/157 (16%) were comparable.
Conclusion: Given more favorable and intermediate risk profiles of patients with AML in the Navya AI plus experts group, there is evidence to show better selection of patients for philanthropic funding who are likely to withstand and survive treatment, compared to the experts alone group. The application of Navya AI in patient selection may help identify the neediest hematolymphoid patients that would most benefit from standardized treatment and funding, maximizing impact of philanthropic funding.
No relevant conflicts of interest to declare.
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal