Vulnerability and core intervention targets in the symptom burden for personalized symptom management in hemodialysis care: A computational simulation modeling study

Background

End-stage kidney disease (ESKD) patients on hemodialysis (HD) endure a high symptom burden. Despite the clinical importance of symptom management, traditional methods fail to predict intervention effects or identify optimal targets due to limited consideration of symptom interactions. Computational simulations with symptom network analysis to identify core intervention targets for personalized care is needed.

Design

Computational simulated modeling study using cross-sectional observational data.

Setting

Across five regions of China—northern, southern, western, eastern, and central.

Participants

A total of 1866 hemodialysis patients participated in the study.

Methods

A combination of variable-centered and person-centered approaches was used to simplify the symptoms measured by real patient-reported data, followed by latent profile analysis to classify patients into symptom burden profiles. In silico interventions were performed using the NodeIdentifyR algorithm to simulate the effects of alleviating and aggravating interventions on the symptom network.

Results

Four symptom clusters were identified: uremic toxin, water-electrolyte, psychological, and gastrointestinal. Latent profile analysis revealed two distinct patient profiles: severe and mild symptom groups. Network analysis highlighted key symptoms such as pruritus, fatigue, anxiety, and easy awakening as central nodes. The in silico interventions in overall groups showed that alleviating interventions targeting easy awakening, fatigue and pruritus as core targets. Treating these may reduce the symptom burden by 10.25 %, 10.00 % and 9.82 %. Aggravating interventions identified pruritus, dry skin and easy awakening as pivotal targets. Preventing the presence of these may separately reduce the symptom burden by 24.69 %, 23.05 % and 22.21 %.

Conclusion

This study provides critical insights into the symptom burden of hemodialysis patients, offering potential targets for personalized care plans for nurses. The results of computational simulations, if go through further longitudinal study for verification, the study has potential to advance the development of more tailored and effective symptom management care approaches, which could enhance patient care and quality of life.