Design of Personalized Follow-up Policies Considering Patient Unpunctuality
Suitable follow-up policies are critical for avoiding complications and controlling medical service costs.Follow-up policies are designed based on operations research methods in the existing literature,such as Markov models and mathematical programming,which lack consideration of patients'unpunctual follow-up behavior,and the resulting policies cannot meet the needs of different patients.The Cox-Frailty model was employed to construct the risk function of adverse outcomes and stratify patients into several risk groups.The timing deviations between each follow-up that a policy scheduled and the one that a patient actually has were modeled by a set of independent and identically distributed random variables.At last,the medical service cost during the planning horizon was computed using the combined methods of virtual age and stochastic process,which was then minimized by a mixed integer nonlinear programming model.Based on the proposed model,the optimal frequency and time interval of follow-ups considering unpunctuality for pediatric T1DM patients with different risks during 1 year were derived and compared with their punctual counterparts,which formed the basis for hospitals to dynamically adjust policies.
万方数据
维普
