Network structure and temporal stability of symptoms during perioperative period among gastrointestinal cancer...

Symptom networks can provide empirical evidence for developing personalized and precise symptom management strategies. However, the network structure and temporal stability of perioperative symptoms among gastrointestinal cancer patients remain unknown. This study aims to explore the dynamic connections between symptoms and accurately identify core symptoms to support clinical decision-making.

The measurement points included T0 (2 days before surgery), T1 (2 days after surgery), T2 (6 days after surgery), T3 (10 days after surgery), and T4 (14 days after surgery). Measurement tools included M.D. Anderson Symptoms Inventory-Gastrointestinal Cancer Module (MDASI-GI). A Multilevel Vector Autoregressive Model (mlVAR) was used to build the temporal and contemporaneous networks.

A total of 241 gastrointestinal cancer participants were recruited, primarily with colorectal cancer type. In the temporal network, sadness had the strongest predictive effect on appetite change, with a value of -0.231. Additionally, dry mouth was identified as the core symptom with the highest outward strength centrality (1.234) and positively predicted pain, sadness, difficulty swallowing, distress, and fatigue (EW = 0.157 ~ 0.230, Ps < 0.001). In the contemporaneous network, depression was a core symptom with the highest strength centrality (1.001). The strongest correlation was found between distress and sadness (EW = 0.645, Ps < 0.05), followed by dry mouth and difficulty swallowing (EW = 0.363, Ps < 0.05). At five time points, the core symptoms within the perioperative symptom network encompassed appetite loss (at T0, with a value of 0.943), distress (at T1, with a value of 1.225; at T2, with a value of 1.057; and at T3, with a value of 0.858), and sadness (at T4, with a value of 1.238).

There exist a prevalent occurrence of positive predictive and associative effects among symptoms. Moreover, emotional and gastrointestinal symptoms, particularly depression and dry mouth, hold significant positions in the perioperative symptom network and should be prioritized in symptom management strategies.

This study uncovers the underlying patterns of widespread positive predictive and associative effects among symptoms, and provides targeted clinical guidance for managing core symptoms such as dry mouth in perioperative care for cancer patients.