@inproceedings{garg-etal-2026-labels,
title = "Before the Labels: How Dataset Construction Shapes Suicidality Detection in Clinical Text",
author = "Garg, Priyanshi and
Rao, Ishita and
Ding, Jieqiong and
Paullada, Amandalynne",
editor = "Zirikly, Aya and
Bar, Kfir and
MacAvaney, Sean and
Ireland, Molly and
Ophir, Yaakov and
Atzil-Slonim, Dana and
Varadarajan, Vasudha and
Bedrick, Steven and
Desmet, Bart",
booktitle = "Proceedings of the 10th Workshop on Computational Linguistics and Clinical Psychology ({CLP}sych 2026)",
month = jul,
year = "2026",
address = "San Diego, California, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.clpsych-1.9/",
pages = "119--127",
ISBN = "979-8-89176-421-7",
abstract = "Clinical NLP increasingly relies on electronic health record (EHR) datato detect suicidal behaviors, treating clinical documentation as morereliable ground truth than social media. We argue that this framingobscures how EHR-based suicidality datasets encode a particularoperationalization of suicidality, shaped by who authors the data,how episodes are bounded, and how ambiguity is resolved. We groundthis argument in a case study of the ScAN dataset,built over MIMIC-III clinical notes. We show how governanceconstraints, ICD-based cohort selection, single-annotator labeling,and hospital-stay-level aggregation produce labels that foregroundclinician judgment, treat suicidality as a bounded episode, andassume that intent can be reliably inferred from documentation. Alinguistic analysis demonstrates that identical labels subsumeheterogeneous clinical framings differing in temporality, negation,and uncertainty, and that labeling patterns differ across insurancestatus. We argue the clinical NLP community should examine theassumptions embedded in suicidality datasets before interpretingtheir labels as ground truth."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="garg-etal-2026-labels">
<titleInfo>
<title>Before the Labels: How Dataset Construction Shapes Suicidality Detection in Clinical Text</title>
</titleInfo>
<name type="personal">
<namePart type="given">Priyanshi</namePart>
<namePart type="family">Garg</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ishita</namePart>
<namePart type="family">Rao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jieqiong</namePart>
<namePart type="family">Ding</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Amandalynne</namePart>
<namePart type="family">Paullada</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 10th Workshop on Computational Linguistics and Clinical Psychology (CLPsych 2026)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Aya</namePart>
<namePart type="family">Zirikly</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kfir</namePart>
<namePart type="family">Bar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sean</namePart>
<namePart type="family">MacAvaney</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Molly</namePart>
<namePart type="family">Ireland</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yaakov</namePart>
<namePart type="family">Ophir</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dana</namePart>
<namePart type="family">Atzil-Slonim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vasudha</namePart>
<namePart type="family">Varadarajan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Steven</namePart>
<namePart type="family">Bedrick</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bart</namePart>
<namePart type="family">Desmet</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-421-7</identifier>
</relatedItem>
<abstract>Clinical NLP increasingly relies on electronic health record (EHR) datato detect suicidal behaviors, treating clinical documentation as morereliable ground truth than social media. We argue that this framingobscures how EHR-based suicidality datasets encode a particularoperationalization of suicidality, shaped by who authors the data,how episodes are bounded, and how ambiguity is resolved. We groundthis argument in a case study of the ScAN dataset,built over MIMIC-III clinical notes. We show how governanceconstraints, ICD-based cohort selection, single-annotator labeling,and hospital-stay-level aggregation produce labels that foregroundclinician judgment, treat suicidality as a bounded episode, andassume that intent can be reliably inferred from documentation. Alinguistic analysis demonstrates that identical labels subsumeheterogeneous clinical framings differing in temporality, negation,and uncertainty, and that labeling patterns differ across insurancestatus. We argue the clinical NLP community should examine theassumptions embedded in suicidality datasets before interpretingtheir labels as ground truth.</abstract>
<identifier type="citekey">garg-etal-2026-labels</identifier>
<location>
<url>https://aclanthology.org/2026.clpsych-1.9/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>119</start>
<end>127</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Before the Labels: How Dataset Construction Shapes Suicidality Detection in Clinical Text
%A Garg, Priyanshi
%A Rao, Ishita
%A Ding, Jieqiong
%A Paullada, Amandalynne
%Y Zirikly, Aya
%Y Bar, Kfir
%Y MacAvaney, Sean
%Y Ireland, Molly
%Y Ophir, Yaakov
%Y Atzil-Slonim, Dana
%Y Varadarajan, Vasudha
%Y Bedrick, Steven
%Y Desmet, Bart
%S Proceedings of the 10th Workshop on Computational Linguistics and Clinical Psychology (CLPsych 2026)
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, USA
%@ 979-8-89176-421-7
%F garg-etal-2026-labels
%X Clinical NLP increasingly relies on electronic health record (EHR) datato detect suicidal behaviors, treating clinical documentation as morereliable ground truth than social media. We argue that this framingobscures how EHR-based suicidality datasets encode a particularoperationalization of suicidality, shaped by who authors the data,how episodes are bounded, and how ambiguity is resolved. We groundthis argument in a case study of the ScAN dataset,built over MIMIC-III clinical notes. We show how governanceconstraints, ICD-based cohort selection, single-annotator labeling,and hospital-stay-level aggregation produce labels that foregroundclinician judgment, treat suicidality as a bounded episode, andassume that intent can be reliably inferred from documentation. Alinguistic analysis demonstrates that identical labels subsumeheterogeneous clinical framings differing in temporality, negation,and uncertainty, and that labeling patterns differ across insurancestatus. We argue the clinical NLP community should examine theassumptions embedded in suicidality datasets before interpretingtheir labels as ground truth.
%U https://aclanthology.org/2026.clpsych-1.9/
%P 119-127
Markdown (Informal)
[Before the Labels: How Dataset Construction Shapes Suicidality Detection in Clinical Text](https://aclanthology.org/2026.clpsych-1.9/) (Garg et al., CLPsych 2026)
ACL