The aging population has promoted the development of demand in the long-term care (LTC) insurance for the elderly, but the assessment and process management of LTC are not yet perfect, which can easily lead to losses in the medical insurance fund. Firstly, OLAP analysis is used to evaluate data, extract key influencing factors, and then Bayesian algorithm is used for probability modeling to quantify the impact of various factors on system reliability. A LTC insurance anomaly warning model is constructed and verified through analysis. It is found that the evaluation institution has problems such as low homogenization, poor communication and coordination, deviation from the mean, and relevant policy recommendations are proposed.