@inproceedings{kumar-etal-2026-sentryline,
title = "{S}entry{L}ine: Evidence-Grounded Question Answering over Evolving Documents in Oncology Care",
author = "Kumar, Tampu Ravi and
Najpande, Gaurav and
Khan, Muhammad Ali and
Khakwani, Kaneez Zahra Rubab and
Kathuria, Karan and
Moses, Shorya Azriel and
Kalia, Yuvraj and
Sonbol, M Bassam and
Riaz, Irbaz Bin and
Gupta, Vivek",
editor = "Agrawal, Sweta and
Abad, Alberto",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: System Demonstrations",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-demos.85/",
pages = "901--911",
ISBN = "979-8-89176-461-3",
abstract = "Oncology care operates at constant pressure of absorbing rapidly evolving evidence base in biomedicine. The American Society of Clinical Oncology (ASCO) addresses this through living guidelines, but the format introduces a new burden: any recommendation can change at any point, across multiple versioned documents. We present SENTRYLINE, a living guideline-aware clinical question answering system. SENTRYLINE retrieves guideline passages through a vectorless hierarchical RAG pipeline and returns a role-specific answer with inline citations, factual and temporal verification reports, and drift detection notes that surface when a guideline has been updated. We construct ASCOBENCH, a benchmark of 405 three-turn conversations across four question categories with gold answers from expert annotators(clinicians), and use test set to evaluate SENTRYLINE against five baselines under an LLM-as-judge framework. Experiments across three generation backbones show consistent improvements over four retrieval baselines and ASCO{'}s guideline assistant, with particularly strong gains on Reasoning and Role-Specific questions where multi-hop synthesis and register adaptation are required."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="kumar-etal-2026-sentryline">
<titleInfo>
<title>SentryLine: Evidence-Grounded Question Answering over Evolving Documents in Oncology Care</title>
</titleInfo>
<name type="personal">
<namePart type="given">Tampu</namePart>
<namePart type="given">Ravi</namePart>
<namePart type="family">Kumar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Gaurav</namePart>
<namePart type="family">Najpande</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Muhammad</namePart>
<namePart type="given">Ali</namePart>
<namePart type="family">Khan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kaneez</namePart>
<namePart type="given">Zahra</namePart>
<namePart type="given">Rubab</namePart>
<namePart type="family">Khakwani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Karan</namePart>
<namePart type="family">Kathuria</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shorya</namePart>
<namePart type="given">Azriel</namePart>
<namePart type="family">Moses</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yuvraj</namePart>
<namePart type="family">Kalia</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">M</namePart>
<namePart type="given">Bassam</namePart>
<namePart type="family">Sonbol</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Irbaz</namePart>
<namePart type="given">Bin</namePart>
<namePart type="family">Riaz</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vivek</namePart>
<namePart type="family">Gupta</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: System Demonstrations</title>
</titleInfo>
<name type="personal">
<namePart type="given">Sweta</namePart>
<namePart type="family">Agrawal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alberto</namePart>
<namePart type="family">Abad</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-461-3</identifier>
</relatedItem>
<abstract>Oncology care operates at constant pressure of absorbing rapidly evolving evidence base in biomedicine. The American Society of Clinical Oncology (ASCO) addresses this through living guidelines, but the format introduces a new burden: any recommendation can change at any point, across multiple versioned documents. We present SENTRYLINE, a living guideline-aware clinical question answering system. SENTRYLINE retrieves guideline passages through a vectorless hierarchical RAG pipeline and returns a role-specific answer with inline citations, factual and temporal verification reports, and drift detection notes that surface when a guideline has been updated. We construct ASCOBENCH, a benchmark of 405 three-turn conversations across four question categories with gold answers from expert annotators(clinicians), and use test set to evaluate SENTRYLINE against five baselines under an LLM-as-judge framework. Experiments across three generation backbones show consistent improvements over four retrieval baselines and ASCO’s guideline assistant, with particularly strong gains on Reasoning and Role-Specific questions where multi-hop synthesis and register adaptation are required.</abstract>
<identifier type="citekey">kumar-etal-2026-sentryline</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-demos.85/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>901</start>
<end>911</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T SentryLine: Evidence-Grounded Question Answering over Evolving Documents in Oncology Care
%A Kumar, Tampu Ravi
%A Najpande, Gaurav
%A Khan, Muhammad Ali
%A Khakwani, Kaneez Zahra Rubab
%A Kathuria, Karan
%A Moses, Shorya Azriel
%A Kalia, Yuvraj
%A Sonbol, M. Bassam
%A Riaz, Irbaz Bin
%A Gupta, Vivek
%Y Agrawal, Sweta
%Y Abad, Alberto
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: System Demonstrations
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-461-3
%F kumar-etal-2026-sentryline
%X Oncology care operates at constant pressure of absorbing rapidly evolving evidence base in biomedicine. The American Society of Clinical Oncology (ASCO) addresses this through living guidelines, but the format introduces a new burden: any recommendation can change at any point, across multiple versioned documents. We present SENTRYLINE, a living guideline-aware clinical question answering system. SENTRYLINE retrieves guideline passages through a vectorless hierarchical RAG pipeline and returns a role-specific answer with inline citations, factual and temporal verification reports, and drift detection notes that surface when a guideline has been updated. We construct ASCOBENCH, a benchmark of 405 three-turn conversations across four question categories with gold answers from expert annotators(clinicians), and use test set to evaluate SENTRYLINE against five baselines under an LLM-as-judge framework. Experiments across three generation backbones show consistent improvements over four retrieval baselines and ASCO’s guideline assistant, with particularly strong gains on Reasoning and Role-Specific questions where multi-hop synthesis and register adaptation are required.
%U https://aclanthology.org/2026.emnlp-demos.85/
%P 901-911
Markdown (Informal)
[SentryLine: Evidence-Grounded Question Answering over Evolving Documents in Oncology Care](https://aclanthology.org/2026.emnlp-demos.85/) (Kumar et al., EMNLP 2026)
ACL
- Tampu Ravi Kumar, Gaurav Najpande, Muhammad Ali Khan, Kaneez Zahra Rubab Khakwani, Karan Kathuria, Shorya Azriel Moses, Yuvraj Kalia, M Bassam Sonbol, Irbaz Bin Riaz, and Vivek Gupta. 2026. SentryLine: Evidence-Grounded Question Answering over Evolving Documents in Oncology Care. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, pages 901–911, Budapest, Hungary. Association for Computational Linguistics.