Association rule mining of care targets from hospitalized dementia patients from a medical center in Taiwan
Wei-Fen Shih, Che‐Wei Lin, Wen-Fu Wang, Hsin‐Hung Wu · Journal of Statistics and Management Systems · 2018
Caring for people with dementia is a huge task, and patients with behavioral and psychological symptoms of dementia such as wandering, resistance to care, agitation, and aggression would increase workloads for staffs to take care of patients. It is critically important to provide a comprehensive care for hospitalized patients with dementia. This study uses apriori algorithm to analyze hospitalized patients with dementia in a medical center in Taiwan. Fifty eight patients with the detailed information from the self-built database are the objects. Gender, age, type of dementia, number of days in hospital, and hospital medical expenses are antecedents in apriori algorithm, while fifteen care targets are both antecedents and consequents. With the minimum support of 40%, minimum confidence of 95%, and lift greater than one, twenty one rules are generated. Eight refined generalized rules are further summarized. Seven out of eight rules depict the correlation between or among care targets, while one generalized rule is related to male patients with dementia. The hospital management can bundle those care targets to provide better healthcare for patients with dementia in this case hospital and pay much attention to male patients with dementia for particular care targets in practice.