What are the symptom heterogeneity and network characteristics among lung cancer survivors in China? A cross-sectional latent profile and network analysis.
To identify latent symptom subgroups, compare symptom network characteristics across subgroups and examine factors associated with subgroup membership among lung cancer survivors.
Cross-sectional study using latent profile analysis and symptom network analysis.
Four hospitals in Shanghai, China, including a national thoracic oncology centre, two tertiary general hospitals and one regional general hospital.
A total of 942 lung cancer survivors who had completed surgical treatment or received at least one course of initial antitumour therapy and were in a stable follow-up phase or treatment interval. The mean age was 64.43±10.70 years and 65.10% were male.
Primary outcomes were latent symptom subgroups identified by latent profile analysis and symptom network characteristics derived from partial correlation networks. Secondary outcomes were socio-demographic, clinical, functional and psychosocial factors associated with subgroup membership, assessed using multivariable multinomial logistic regression following least absolute shrinkage and selection operator (LASSO) variable selection.
Three symptom subgroups were identified: a low-symptom group (n=488, 51.80%), a moderate-symptom group (n=364, 38.64%) and a high-symptom group (n=90, 9.55%). Network density was descriptively higher in the high-symptom group than in the low-symptom group (0.468 vs 0.175). Central symptoms differed across subgroups, with cough in the low-symptom group, vomiting in the moderate-symptom group and distress in the high-symptom group. After LASSO selection and collinearity assessment, 22 predictors were entered into the final multivariable multinomial logistic regression model. Compared with the low-symptom group, surgery with adjuvant therapy was associated with higher odds of membership in the moderate-symptom group (adjusted OR (aOR)=4.054, 95% CI 2.094 to 7.850), whereas better exercise capacity was associated with lower odds of membership in the high-symptom group (6-minute walk distance ≥450 m: aOR=0.101, 95% CI 0.027 to 0.372). The final model had a Nagelkerke pseudo-R² of 0.522.
Symptom burden among lung cancer survivors is heterogeneous and differs in both severity profiles and network structure. Integrating latent profile analysis with symptom network analysis may provide a useful framework for stratified symptom assessment and individualised symptom management in survivorship care.
MR-31-24-027806 (https://www.medicalresearch.org.cn/clinicalResearch/researchInfo?id=eb897346-a67b-4465-a553-1d1a3488bc30).
Cross-sectional study using latent profile analysis and symptom network analysis.
Four hospitals in Shanghai, China, including a national thoracic oncology centre, two tertiary general hospitals and one regional general hospital.
A total of 942 lung cancer survivors who had completed surgical treatment or received at least one course of initial antitumour therapy and were in a stable follow-up phase or treatment interval. The mean age was 64.43±10.70 years and 65.10% were male.
Primary outcomes were latent symptom subgroups identified by latent profile analysis and symptom network characteristics derived from partial correlation networks. Secondary outcomes were socio-demographic, clinical, functional and psychosocial factors associated with subgroup membership, assessed using multivariable multinomial logistic regression following least absolute shrinkage and selection operator (LASSO) variable selection.
Three symptom subgroups were identified: a low-symptom group (n=488, 51.80%), a moderate-symptom group (n=364, 38.64%) and a high-symptom group (n=90, 9.55%). Network density was descriptively higher in the high-symptom group than in the low-symptom group (0.468 vs 0.175). Central symptoms differed across subgroups, with cough in the low-symptom group, vomiting in the moderate-symptom group and distress in the high-symptom group. After LASSO selection and collinearity assessment, 22 predictors were entered into the final multivariable multinomial logistic regression model. Compared with the low-symptom group, surgery with adjuvant therapy was associated with higher odds of membership in the moderate-symptom group (adjusted OR (aOR)=4.054, 95% CI 2.094 to 7.850), whereas better exercise capacity was associated with lower odds of membership in the high-symptom group (6-minute walk distance ≥450 m: aOR=0.101, 95% CI 0.027 to 0.372). The final model had a Nagelkerke pseudo-R² of 0.522.
Symptom burden among lung cancer survivors is heterogeneous and differs in both severity profiles and network structure. Integrating latent profile analysis with symptom network analysis may provide a useful framework for stratified symptom assessment and individualised symptom management in survivorship care.
MR-31-24-027806 (https://www.medicalresearch.org.cn/clinicalResearch/researchInfo?id=eb897346-a67b-4465-a553-1d1a3488bc30).
Authors
Cao Cao, Wang Wang, Li Li, Zhang Zhang, Ye Ye, Dang Dang, Tian Tian, Hao Hao, Cao Cao, Ma Ma, Hou Hou
View on Pubmed