Expectations of Intensive Care Physicians Regarding an AI-Based Decision Support System for Weaning From Continuous Renal Replacement Therapy: Predevelopment Survey Study
Open accessThe DOI matches its Crossref record; 43 of the 43 references it cites were found in a registry or an index; 21 of the 40 sources its claims cite were checked for support.
Abstract
Background: Critically ill patients in intensive care units (ICUs) require continuous monitoring, generating vast amounts of data. Clinical decision support systems (CDSS) leveraging artificial intelligence (AI) technologies have shown promise in improving diagnostic, prognostic, and therapeutic decision-making. However, these models are rarely implemented in clinical practice. Objective: The aim of this study was to survey ICU physicians to understand their expectations, opinions, and level of knowledge regarding a proposed AI-based CDSS for continuous renal replacement therapy (CRRT) weaning, a clinical decision-making process that is still complex and lacking in guidelines. This will be used to guide the development of an AI-based CDSS on which our team is working to ensure user-centered design and successful integration into clinical practice. Methods: A prospective cross-sectional survey of French-speaking physicians with clinical activity in intensive care was conducted between December 2023 and April 2024. The questionnaire consisted of 20 questions structured around 4 axes: overview of the problem and current practices concerning weaning from CRRT, opinion on AI-based CDSS, implementation in daily clinical practice, real-life operation and willingness to adopt the CDSS in everyday practice. Statistical analyses included Wilcoxon rank sum tests for quantitative variables and χ2 or Fisher exact tests for qualitative variables, with multivariate analyses performed using ordinal logistic regression. Results: A total of 171 complete responses were received. Physicians expressed an interest in a CDSS for CRRT weaning, with 70.2% (120/171) viewing AI-based CDSS favorably. Opinions were split regarding the difficulty of the weaning decision itself, with 46.2% (79/171) disagreeing that it is challenging, while 31.6% (54/171) agreed. However, 66.1% (113/171) of respondents supported the value of an AI-based CDSS to assist them in this decision, with younger physicians showing stronger support (81.8%, 27/33 vs 62.3%; 86/138; P=.01). Most respondents (163/171, 95.3%) emphasized the importance of understanding the criteria used by the model to make its predictions. Conclusions: Our findings highlight an optimistic attitude among ICU physicians toward AI-based CDSS for CRRT weaning, emphasizing the need for transparency, integration into existing workflows, and alignment with clinicians' decision-making processes. Actionable recommendations include incorporating key variables such as urine output and biological parameters, defining probability thresholds for recommendations and ensuring model transparency to facilitate the successful adoption and integration into clinical practice. The methodology of this survey may help the development of further predevelopment studies accompanying AI-based CDSS projects.
Overview
CoChat Summary
## TL;DR
This French prospective cross-sectional survey asked ICU physicians about expectations for an AI-based CDSS to support weaning from CRRT. Among 171 complete responses, physicians were broadly favorable toward such a tool, but they strongly emphasized transparency, workflow integration, and clinically meaningful variables.
## Key findings
- 171 complete responses were analyzed; the estimated response rate was about 6.0% (168/2850), based on an approximate survey denominator.
- 70.2% (120/171) viewed AI-based CDSS favorably, and 66.1% (113/171) supported an AI-based CDSS to assist CRRT-weaning decisions.
- Physicians most often identified resumption of diuresis above a certain volume (125/171, 73.1%), variation in biological settings (110/171, 63.3%), and good response to diuretics (82/171, 47.9%) as key variables for weaning.
- Transparency was a major concern: 95.3% (163/171) said it was important to understand the criteria used by the model, and 74.9% (128/171) wanted the CDSS to provide a percentage of certainty.
- Preferred implementation features included a prediction rendered at a specific time of day (108/171, 63.2%), integration into the ICU EHR (124/171, 72.5%), and a median manual-entry allowance of 5 variables (IQR 4–6).
- Physicians indicated a median 50% (IQR 30–60) probability threshold below which they would not consider weaning, and 80% (IQR 73–85) above which they would consider weaning.
## Methods
- Design: Prospective cross-sectional survey conducted in France from December 1, 2023 to April 30, 2024 among French-speaking physicians with ICU clinical activity.
- Sample/data: 171 complete questionnaires from attending physicians, fellows, and junior doctors; respondents had a median age of 39 years (IQR 33–45) and median ICU experience of 9 years (IQR 3–14).
- Analysis: Mandatory questionnaire responses were analyzed in complete-case fashion; the paper used Wilcoxon rank-sum tests, χ² or Fisher exact tests, and ordinal logistic regression with covariates including age, ICU experience, specialty, and workplace setting.
## Limitations
- The paper states that the response rate was low, raising selection bias concerns and limiting generalizability.
- The authors note that the survey link was distributed through newsletters and social media and that the exact number of physicians reached was not exhaustively documented; this makes the response rate and representativeness uncertain.
- The paper states the sample was younger and had more university-hospital physicians than expected, which may have biased results toward greater openness to AI.
- The paper notes self-reporting bias, possible comprehension issues for some questions, and that no validated tool exists for predevelopment surveys of AI-based CDSS.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures




Verification
Claims the paper makes
19 Valid · 0 Not valid · 18 Not checked · 59 Own results and interpretations
Every source the paper cites for this claim was checked and supports it.
In the paper
“In a review of 89 articles on CDSS in various medical fields, Jankovic et al [38] showed that poorly performing CDSS with frequent alerts could contribute to clinicians’ frustration and burnout.”
§ Expectations of Intensive Care Physicians Regarding an AI-Based Decision Support System for Weaning From Continuous Renal Replacement Therapy: Predevelopment Survey Study > Discussion > Preferences for Implementation
What the sources say
- Supportedcites [38] Jankovic, Clinical decision support and implications for the clinician burnout crisis, Yearb Med Inform source
The Results section states that 89 articles met the review’s inclusion criteria. Results §3.3 links irrelevant alerts to alarm fatigue and negative effects on clinician wellbeing; the Discussion describes high alert volume, poor relevance, clinician dissatisfaction, and potential contributions to burnout. This supports the claim’s cautious wording that poorly performing CDSS with frequent alerts could contribute to frustration and burnout. (LLM, conf=0.91)
hedgedstates a possible cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“This positive outlook towards AI is consistent with other studies showing clinicians’ openness to AI-based CDSS if they enhance clinical practice and improve patient outcomes [37].”
§ Expectations of Intensive Care Physicians Regarding an AI-Based Decision Support System for Weaning From Continuous Renal Replacement Therapy: Predevelopment Survey Study > Discussion > Attitudes Toward AI and CDSS
What the sources say
- Partly supportedcites [37] Oh, Physician confidence in artificial intelligence: an online mobile survey, J Med Internet Res sourceExperimental — accuracy not yet measured
The Discussion (page unavailable) reports that physicians and medical students regarded AI as useful, particularly for diagnosis and treatment planning. It does not establish that clinicians were open specifically to AI-based CDSS on the condition that they improve clinical practice or patient outcomes. (LLM, conf=0.86)
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“These factors align with a previous systematic review, which identified similar predictors for successful RRT weaning, such as the urine output before RRT discontinuation and biochemical criteria (eg, serum urea, serum creatinine and creatinine clearance) [13].”
§ Expectations of Intensive Care Physicians Regarding an AI-Based Decision Support System for Weaning From Continuous Renal Replacement Therapy: Predevelopment Survey Study > Discussion > Factors Influencing Weaning Decision
What the sources say
- Supportedcites [13] Katulka, Determining the optimal time for liberation from renal replacement therapy in critically ill patients: a systematic review and meta-analysis (DOnE RRT), Crit Care source
The Discussion (Key findings) identifies urine output and conventional biochemical markers—including creatinine, urea, and estimates of GFR—as parameters evaluated for predicting successful RRT discontinuation. The Results (Urine output prior to discontinuation) also reports a pooled analysis of urine output measured before RRT was stopped. (LLM, conf=0.92)
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“However, 66.1% (113/171) of respondents supported the value of an AI-based CDSS to assist them in this decision”
§ Expectations of Intensive Care Physicians Regarding an AI-Based Decision Support System for Weaning From Continuous Renal Replacement Therapy: Predevelopment Survey Study > Abstract > Results
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“with 70.2% (120/171) viewing AI-based CDSS favorably”
§ Expectations of Intensive Care Physicians Regarding an AI-Based Decision Support System for Weaning From Continuous Renal Replacement Therapy: Predevelopment Survey Study > Abstract > Results
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
33 more claims
Create a free account to see all 96 claims and every source verdict.
- It may state two separate claims as one. 34
- Probably not a substantial claim. 35
- The quote only partly states this claim. 1
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
“Hence, the estimated response rate for the survey is approximately 6.0% (168/2850, considering that 2 respondents were from Belgium and 1 from Switzerland).”
§ Expectations of Intensive Care Physicians Regarding an AI-Based Decision Support System for Weaning From Continuous Renal Replacement Therapy: Predevelopment Survey Study > Results > Respondent Characteristics
Reported 6 · recomputed 5.894737
Citations
Cites
User Acceptance of Information Technology: Toward A Unified View1
Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003) · MIS Quarterly · doi:10.2307/30036540 · Cited by 45,035 · cited 1×
“The questions also considered acceptance concepts defined by the Unified Theory of Acceptance and Use of Technology (UTAUT) model [30].”
Key challenges for delivering clinical impact with artificial intelligence
Kelly, C. J., Karthikesalingam, A., Suleyman, M., Corrado, G., & King, D. (2019) · BMC Medicine · doi:10.1186/s12916-019-1426-2 · Cited by 3,136 · cited 2×
“Despite these advances, most AI-based CDSS models fail to progress beyond the prototype stage and are rarely deployed into clinical practice, even though their performance is sometimes promising [6].”
A Consensus-Based Checklist for Reporting of Survey Studies (CROSS)
Sharma, A., Minh Duc, N. T., Luu Lam Thang, T., Nam, N. H., Ng, S. J., Abbas, K. S., Huy, N. T., Marušić, A., Paul, C. L., Kwok, J., et al. (2021) · Journal of General Internal Medicine · doi:10.1007/s11606-021-06737-1 · Cited by 2,292 · cited 1×
“This study was reported in accordance with the consensus-based Checklist for Reporting of Survey Studies (CROSS) checklist (Multimedia Appendix 1) [24].”
Global epidemiology and outcomes of acute kidney injury
Hoste, E. A. J., Kellum, J. A., Selby, N. M., Zarbock, A., Palevsky, P. M., Bagshaw, S. M., Goldstein, S. L., Cerdá, J., & Chawla, L. S. (2018) · Nature Reviews Nephrology · doi:10.1038/s41581-018-0052-0 · Cited by 1,486
Initiation strategies for renal-replacement therapy in the intensive care unit.
Gaudry, S., Hajage, D., Schortgen, F., Martin-Lefevre, L., Pons, B., Boulet, E., Boyer, A., Chevrel, G., Lerolle, N., Carpentier, D., et al. (2016) · New England Journal of Medicine · doi:10.1056/nejmoa1603017 · Cited by 1,243 · cited 1×
“While significant research has focused on the optimal timing for initiating CRRT [11,12], much less attention has been given to the question of when and how to discontinue this therapy [13].”
Clinical decision support in the era of artificial intelligence
Shortliffe, E. H., & Sepúlveda, M. J. (2018) · JAMA · doi:10.1001/jama.2018.17163 · Cited by 920 · cited 1×
“The first step is ensuring that the AI-based CDSS addresses the specific needs of clinicians [41].”
Timing of renal-replacement therapy in patients with acute kidney injury and sepsis.
Barbar, S. D., Clere-Jehl, R., Bourredjem, A., Hernu, R., Montini, F., Bruyère, R., Lebert, C., Bohé, J., Badie, J., Eraldi, J.-P., et al. (2018) · New England Journal of Medicine · doi:10.1056/nejmoa1803213 · Cited by 777 · cited 1×
“While significant research has focused on the optimal timing for initiating CRRT [11,12], much less attention has been given to the question of when and how to discontinue this therapy [13].”
Fleuren, L. M., Klausch, T. L. T., Zwager, C. L., Schoonmade, L. J., Guo, T., Roggeveen, L. F., Swart, E. L., Girbes, A. R. J., Thoral, P., Ercole, A., et al. (2020) · Intensive Care Medicine · doi:10.1007/s00134-019-05872-y · Cited by 755 · cited 1×
“For instance, AI models for sepsis prediction can alert clinicians to septic events hours before clinical signs emerge, enabling timely interventions [5].”
Barriers and facilitators to utilizing digital health technologies by healthcare professionals
Borges do Nascimento, I. J., Abdulazeem, H., Vasanthan, L. T., Martinez, E. Z., Zucoloto, M. L., Østengaard, L., Azzopardi-Muscat, N., Zapata, T., & Novillo-Ortiz, D. (2023) · npj Digital Medicine · doi:10.1038/s41746-023-00899-4 · Cited by 673
de Hond, A. A. H., Leeuwenberg, A. M., Hooft, L., Kant, I. M. J., Nijman, S. W. J., van Os, H. J. A., Aardoom, J. J., Debray, T. P. A., Schuit, E., van Smeden, M., et al. (2022) · npj Digital Medicine · doi:10.1038/s41746-021-00549-7 · Cited by 499
Cited by
Recent approaches of artificial intelligence in intensive care unit: A review
Havelikar, U. A., Bembde, A., Patil, D., Shinde, A., Sharma, R. K., Patel, V. P., & Chaudhari, N. B. (2025) · Intelligent Hospital · doi:10.1016/j.inhs.2025.100030 · Cited by 17
Role of Artificial Intelligence in Critical Care Medicine: A Literature Review
ABDELBAKY, A., Elmasry, W. G., Awad, A. H., & Khan, S. (2025) · Cureus · doi:10.7759/cureus.90149 · Cited by 14
Popoff, B., Delange, B., Nicolas, J., Le Gall, A., Dahamna, B., Cuggia, M., Clavier, T., & Bouzillé, G. (2025) · Scientific Reports · doi:10.1038/s41598-025-26774-8 · Cited by 2
Self-citationChen, H., Xiaowen, Y., Peng, J., Xu, X., Li, Y., Tong, Z., & Chunhong, G. (2026) · Risk Management and Healthcare Policy · doi:10.2147/rmhp.s594033 · Cited by 1
“9,10 Notably, a multicenter study demonstrated that CDSS-informed protocols reduced UE incidence by 30.3% in high-risk populations through structured alerting and standardized care pathways.”
InfluentialHu, Z., Yu, X., Liu, X., Jiang, T., & Chen, L. (2026) · Frontiers in Medicine · doi:10.3389/fmed.2026.1882567
Metadata
- Authors
- Popoff, B., Cabon, S., Cuggia, M., Bouzillé, G., & Clavier, T.
- Affiliations
- Inserm (FR)
- Corresponding author
- Popoff, Benjamin
- Venue
- JMIR Medical Informatics
- Publisher
- JMIR Publications Inc.
- Published
- 2025-04-23
- Volume
- 13
- Pages
- e63709-e63709
- ISSN
- 2291-9694 (ISSN-L)
- PMC
- PMC12043247
- OpenAlex
- W4409720695
- Cited by
- 6
- References
- 43
- Citations by year
- 2026: 3 · 2025: 3
- FWCI
- 1.43
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by
- Locations
- JMIR Medical Informatics (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- HAL (submittedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- Intensive Care Society
- Société de Réanimation de Langue Française
- Topics
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Health Sciences › Health Professions › Health Information Management › Electronic Health Records Systems
- Health Sciences › Medicine › Nephrology › Acute Kidney Injury Research
- Keywords
- artificial intelligence, clinical decision support systems, continuous renal replacement therapy, intensive care physicians, physician attitudes, artificial intelligence adoption, model transparency, explainability, clinical workflow integration, user-centered design, intensive care, physician-patient communication, physician-patient relationship
- MeSH terms
- Continuous Renal Replacement Therapy, Adult, Artificial Intelligence, Attitude of Health Personnel, Critical Care, Cross-Sectional Studies, Female, Humans, Intensive Care Units, Male, Middle Aged, Physicians, Prospective Studies, Surveys and Questionnaires, Decision Support Systems, Clinical
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.