This article introduces a dynamical stratified blocked randomized algorithm when the static stratified blocked randomized algorithm is not suitable to resolve the problem caused by uncertain stratified factors and levels during calculation of the sample size at the stage of clinical trial design. Prior to the start of the clinical trial, a list of random numbers is created by blocked randomized algorithm. During field implementation, a block of random numbers is dynamically allocated to a stratify level. Thereafter, a subject is randomized into different groups in that block. The study of Chinese medicine for gastric cancer anemia and blood hypercoagulability is used as an example to illustrate how to design and implement dynamic stratified block randomized algorithm. The results show that the dynamic stratified block randomized algorithm is more flexible and adaptable than the static stratified blocked randomized algorithm. However, its application is more complex and requires higher standards in clinical trials.
ObjectivesTo investigate risk factors for unplanned readmission in ischemic stroke patients within 31 days by using random forest algorithm.MethodsThe record of readmission patients with ischemic stroke within 31 days from 24 hospitals in Beijing between between 2015 and 2016 were collected. Patients were divided into two groups according to the occurrence of readmission within 31 days or not. Chi-squared or Mann-Whitney U test was used to select variables into the random forest algorithm. The precision coefficient and the Gini coefficient were used to comprehensively assess the importance of all variables, and select the more important variables and use the margind effect to assess relative risk of different levels.ResultsA total of 3 473 patients were included, among them 960 (27.64%) were readmitted within 31 days after stroke hospitalization. Based on the result of random forest, the most important variables affecting the risk of unplanned readmission within 31 days included the length of hospital stay, age, medical expense payment, rank of hospital, and occupation. When hospitalization was within 1 month, 10-day-hospitalization-stay patients had the lowest risk of rehospitalization; the younger the patients was, the higher the risk of readmission was. For ranks of hospital, patients from tertiary hospital had higher risk than secondary hospital. Furthermore, patients whose medical expenses were paid by free medical service and whose occupations were managers or staffs had higher risk of readmission within 31 days.ConclusionsThe unplanned readmission risk within 31 days of discharged ischemic stroke patients was connected not only with disease, but also with personal social and economic factors. Thus, more attention should be paid to both the medical process and the personal and family factors of stroke patients.