Validity, Judgment, and Authority: The Hypotension Prediction Index as a Test Case for Clinical AI Alignment

The Hypotension Prediction Index (HPI) estimates from the arterial waveform the likelihood of impending hypotension and, when the index is high, prompts review of a secondary screen that displays pulse pressure or stroke volume variation (PPV/SVV), peripheral arterial dP/dt, and dynamic arterial elastance (Eadyn) under the headings preload, contractility, and afterload. Its predictive performance is supported by meta-analyses, but the physiological meaning of the secondary-screen variables depends on conditions that the monitor does not verify, so that a numerically correct value can be physiologically uninterpretable. This narrative review summarizes the evidence on the prediction and on the validity conditions of each variable, using a companion conceptual framework that distinguishes the capability to evaluate from the authority to act. We separate the documented functions of the current software from extensions that we propose: a runtime validity gate that would verify, flag, or withhold interpretations whose prerequisites cannot be established, and a bedside sequence for clinicians. The framework, the proposed PPV checklist, and the taxonomy tables are expert-derived, have not been tested against clinical outcomes or usability data, and concern a single manufacturer's product. Within these limits, HPI illustrates a principle that we propose for clinical AI: prediction accuracy does not guarantee interpretation validity, and interpretation validity does not confer therapeutic authority.

Authors

Institutions

Publication Details

Journal
The Journal of Critical Care Medicine
Published
2026-11-16
DOI
https://doi.org/10.62838/jccm-2026-0068
Primary Topic
Hemodynamic Monitoring and Therapy
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Validity, Judgment, and Authority: The Hypotension Prediction Index as a Test Case for Clinical AI Alignment

Christian de Tymowski, Dan Longrois, Stefan Andrei, David Holcman et al.
1 links
The Journal of Critical Care Medicine
Hemodynamic Monitoring and Therapy
article

Validity, Judgment, and Authority: The Hypotension Prediction Index as a Test Case for Clinical AI Alignment

Christian de Tymowski, Dan Longrois, Stefan Andrei, David Holcman, Pierre-Grégoire Guinot
article en

Abstract

The Hypotension Prediction Index (HPI) estimates from the arterial waveform the likelihood of impending hypotension and, when the index is high, prompts review of a secondary screen that displays pulse pressure or stroke volume variation (PPV/SVV), peripheral arterial dP/dt, and dynamic arterial elastance (Eadyn) under the headings preload, contractility, and afterload. Its predictive performance is supported by meta-analyses, but the physiological meaning of the secondary-screen variables depends on conditions that the monitor does not verify, so that a numerically correct value can be physiologically uninterpretable. This narrative review summarizes the evidence on the prediction and on the validity conditions of each variable, using a companion conceptual framework that distinguishes the capability to evaluate from the authority to act. We separate the documented functions of the current software from extensions that we propose: a runtime validity gate that would verify, flag, or withhold interpretations whose prerequisites cannot be established, and a bedside sequence for clinicians. The framework, the proposed PPV checklist, and the taxonomy tables are expert-derived, have not been tested against clinical outcomes or usability data, and concern a single manufacturer's product. Within these limits, HPI illustrates a principle that we propose for clinical AI: prediction accuracy does not guarantee interpretation validity, and interpretation validity does not confer therapeutic authority.

The Journal of Critical Care Medicine
Université Paris Cité (FR), University of Cambridge (GB), Hôpital Louis-Mourier (FR), Université Sorbonne Paris Nord (FR), CHU Dijon Bourgogne (FR), Université Bourgogne Franche-Comté (FR), Institut de Biologie de l'École Normale Supérieure (FR), Laboratoire de Recherche Vasculaire Translationnelle (FR), Hôpital Bichat-Claude-Bernard (FR)
Good health and well-being
Hemodynamic Monitoring and Therapy
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon