Scholarlink

Objective <p>To construct symptom network for acute myeloid leukemia (AML) patients at different stages of treatment, explore the longitudinal relationships among 31 symptoms at different points, identify key causal symptoms, and determine optimal intervention time windows.</p> Methods <p>AML patients admitted between July 2024 and May 2025 for chemotherapy were enrolled. Symptom incidence was assessed using the Chinese version of the Memorial Symptom Assessment Scale (MSAS-Ch) at three points: the day before chemotherapy initiation (T1), chemotherapy completion day (T2), and seven days post-chemotherapy (T3).Cross-lagged panel network (CLPN) modeling was applied to analyze longitudinal mechanisms among 31 core symptoms across three critical treatment phases.</p> Results <p>Of 275 enrolled AML patients (mean age = 56.02 ± 13.96&#xa0;years), fatigue and feeling sad were the most prevalent symptoms across T1, T2, and T3. CLPN analysis identified key cross-lagged paths: difficulty swallowing (S22) → loss of appetite (S20) from T1 → T2 and weight loss (S26) → loss of appetite (S20) from T2 → T3. Predictive analysis showed feeling sad (S16) had the highest out-predictive strength in T1 → T2, while weight loss (S26) had the highest out-predictive strength in T2 → T3. Shortness of breath (S14) exhibited the highest in-predictive strength in T1 → T2, and loss of appetite (S20) had the highest in-predictive strength in T2 → T3. Centrality analysis revealed feeling sad (S16) and mental tension (S5) as the top out-strength nodes in T1 → T2, whereas mouth ulcers (S24) and weight loss (S26) were dominant in T2 → T3.</p> Conclusions <p>AML symptom management should focus on the dynamic interplay of core symptoms, particularly during T2, providing targeted nutritional and psychological support guided by causal pathways to improve patient outcomes and quality of life. Future models integrating physiological data and treatment plans could enhance symptom prediction and management.</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

Symptom dynamics through multiphase treatment in acute myeloid leukemia: a cross-lagged panel network analysis

  • 期刊论文

  • 发表日期：

    2025年11月29日

查看全文
  • Qi Liu,
  • Yongxia Chen,
  • Zihan Liu,
  • Yizhu Xue,
  • Wenxiu Qi,
  • Xiangqin Xing,
  • Xi Wang

摘要

Objective

To construct symptom network for acute myeloid leukemia (AML) patients at different stages of treatment, explore the longitudinal relationships among 31 symptoms at different points, identify key causal symptoms, and determine optimal intervention time windows.

Methods

AML patients admitted between July 2024 and May 2025 for chemotherapy were enrolled. Symptom incidence was assessed using the Chinese version of the Memorial Symptom Assessment Scale (MSAS-Ch) at three points: the day before chemotherapy initiation (T1), chemotherapy completion day (T2), and seven days post-chemotherapy (T3).Cross-lagged panel network (CLPN) modeling was applied to analyze longitudinal mechanisms among 31 core symptoms across three critical treatment phases.

Results

Of 275 enrolled AML patients (mean age = 56.02 ± 13.96 years), fatigue and feeling sad were the most prevalent symptoms across T1, T2, and T3. CLPN analysis identified key cross-lagged paths: difficulty swallowing (S22) → loss of appetite (S20) from T1 → T2 and weight loss (S26) → loss of appetite (S20) from T2 → T3. Predictive analysis showed feeling sad (S16) had the highest out-predictive strength in T1 → T2, while weight loss (S26) had the highest out-predictive strength in T2 → T3. Shortness of breath (S14) exhibited the highest in-predictive strength in T1 → T2, and loss of appetite (S20) had the highest in-predictive strength in T2 → T3. Centrality analysis revealed feeling sad (S16) and mental tension (S5) as the top out-strength nodes in T1 → T2, whereas mouth ulcers (S24) and weight loss (S26) were dominant in T2 → T3.

Conclusions

AML symptom management should focus on the dynamic interplay of core symptoms, particularly during T2, providing targeted nutritional and psychological support guided by causal pathways to improve patient outcomes and quality of life. Future models integrating physiological data and treatment plans could enhance symptom prediction and management.