Scholarlink

Purpose <p>Glycemic emergencies are a frequent cause of hospital admissions and can lead to severe complications, particularly in older or medically complex patients. Anticipating these events is essential for timely intervention and personalized care. This study aimed to identify patients at risk of hypoglycemia or hyperglycemia using routinely collected data from emergency department of 11 hospitals in Spain.</p> Methods <p>A comprehensive modeling framework was designed to identify glycemic events from clinical data. Multiple supervised learning algorithms were trained and validated using routinely collected patient variables. Model explainability was ensured through the integration of XAI methods, which quantified the contribution of individual clinical features to prediction outcomes. This approach enabled transparent model behavior, supporting clinical understanding and facilitating patient risk stratification.</p> Results <p>The developed models achieved predictive accuracies between 70% and 74%. Explainability analyses revealed distinct glycemic risk patterns: patients aged 87 years and above were predominantly hypoglycemic, while among younger individuals, those with a body temperature exceeding 36 <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(^{\circ}\)</EquationSource> </InlineEquation>C, Chronic Kidney Disease (CKD) (creatinine <InlineEquation ID="IEq2"> <EquationSource Format="TEX">\( &gt; 1.5 \textrm{mg/dL}\)</EquationSource> </InlineEquation>), and platelet counts below <InlineEquation ID="IEq3"> <EquationSource Format="TEX">\(200,000 /\mu \textrm{L}\)</EquationSource> </InlineEquation> were more likely to be hyperglycemic, whereas others tended toward hypoglycemia.</p> Conclusions <p>These results highlight the predictive value of age, thermoregulation, renal function, and hematologic parameters in assessing glycemic risk. The combination of machine learning and explainability provides interpretable, actionable insights to support early risk stratification and improve outcomes in diabetes care.</p> Trial registration <p>This retrospective study was approved by the Comité de Ética de la Investigación con medicamentos (CEIm) of Hospital Clínico San Carlos (Madrid, Spain) (Approval code: 19/332-E). The requirement for informed consent was waived. All procedures followed institutional ethics standards, the Declaration of Helsinki, and applicable national regulations. <b>Retrospectively registered</b>.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Exploring the potential of XAI methods in generating clinically meaningful explanations for glycemia prediction in diabetes patients

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年3月14日

查看全文
  • Sayna Rotbei,
  • Pablo Matías Soler,
  • Beatriz Merino-Barbancho,
  • Laura Lopez-Perez,
  • Arturo Corbatón Anchuelo,
  • Luis Picazo García,
  • Ricardo Mesanza Forés,
  • Laura Mariel Matus,
  • Ricardo Muñoz Albert,
  • Aitor Odiaga Andicoechea,
  • Raquel Piñero Panadero,
  • María Ángeles San Martín Díez,
  • Ainhoa Burzaco Sánchez,
  • Rosana Soriano Barrón,
  • Andrea Irimia,
  • Esther Ruescas Esculano,
  • Mireia Cramp Vinceixo,
  • F. Beddar Chaib,
  • Hania Tourab,
  • Giuseppe Fico,
  • Alessio Botta

摘要

Purpose

Glycemic emergencies are a frequent cause of hospital admissions and can lead to severe complications, particularly in older or medically complex patients. Anticipating these events is essential for timely intervention and personalized care. This study aimed to identify patients at risk of hypoglycemia or hyperglycemia using routinely collected data from emergency department of 11 hospitals in Spain.

Methods

A comprehensive modeling framework was designed to identify glycemic events from clinical data. Multiple supervised learning algorithms were trained and validated using routinely collected patient variables. Model explainability was ensured through the integration of XAI methods, which quantified the contribution of individual clinical features to prediction outcomes. This approach enabled transparent model behavior, supporting clinical understanding and facilitating patient risk stratification.

Results

The developed models achieved predictive accuracies between 70% and 74%. Explainability analyses revealed distinct glycemic risk patterns: patients aged 87 years and above were predominantly hypoglycemic, while among younger individuals, those with a body temperature exceeding 36 \(^{\circ}\) C, Chronic Kidney Disease (CKD) (creatinine \( > 1.5 \textrm{mg/dL}\) ), and platelet counts below \(200,000 /\mu \textrm{L}\) were more likely to be hyperglycemic, whereas others tended toward hypoglycemia.

Conclusions

These results highlight the predictive value of age, thermoregulation, renal function, and hematologic parameters in assessing glycemic risk. The combination of machine learning and explainability provides interpretable, actionable insights to support early risk stratification and improve outcomes in diabetes care.

Trial registration

This retrospective study was approved by the Comité de Ética de la Investigación con medicamentos (CEIm) of Hospital Clínico San Carlos (Madrid, Spain) (Approval code: 19/332-E). The requirement for informed consent was waived. All procedures followed institutional ethics standards, the Declaration of Helsinki, and applicable national regulations. Retrospectively registered.