Graphical modeling can be used to illustrate associations between variables describing functioning in head and neck cancer patients

ObjectiveTo examine the associations between variables of functioning measured by the International Classification of Functioning, Disability and Health (ICF) in head and neck cancer (HNC) patients by means of graphical modeling.Study Design and SettingGraphical modeling was used on a data set of a cross-sectional multicentric study of 145 patients with HNC. Functioning was qualified using the extended ICF checklist. Multiple imputation was used to handle missing data. The least absolute shrinkage and selection operator for generalized linear models was used to identify conditional associations between the ICF categories. Bootstrap aggregating was used to enhance the accuracy and validity of model selection.ResultsThe resulting graph shows largely meaningful associations between the ICF categories. One central point could be visualized consisting of a circular path of d330 Speaking, d350 Conversation, b510 Ingestion functions, s320 Structure of mouth, and b310 Voice functions. Another important structure in the graph were the bow-shaped associations beginning with d335 Producing nonverbal messages to b130 Energy and drive functions.ConclusionGraphical modeling can be used to describe associations between different areas of functioning in HNC patients. They found associations can be the basis for improved rehabilitation and gives a deeper understanding of functioning in HNC patients.