ABSTRACT
AIM: Previous studies have suggested that big stick design (BSD) method can only be used in clinical trials of two treatments with equal proportion, which has good statistical performance and has become the recommended choice of randomized methods. This study expands BSD method, so that it can be applied to three groups, and provides more randomized methods for clinical trials. METHODS: On the basis of BSD method used in two treatments with equal proportion, the derivation conditional allocation probability of BSD method used in three treatments with equal proportion was carried out. BSD method was compared with simple randomization (SR) method, permuted block design (PBD) method and block urn design (BUD) method by Monte-Carlo simulation in balance and randomness. RESULTS: In terms of balance, PBD method was the best, followed by BUD method, BSD method, and SR method was the worst. In terms of randomness, SR method was the best, followed by BSD method, BUD method and PBD method. The comprehensive performance showed that BSD method was better than BUD method, PBD method and SR method. CONCLUSION: The expanded BSD method used in three treatments with equal proportion has good comprehensive performance, and it can be the recommended randomization method for clinical trials of three treatments with equal proportion.
ABSTRACT
Objective To determine the level of soluble B7-H1 (sB7-H1) in serum of patients with colorectal cancer ( CRC) , and to investigate its clinical application value in CRC .Methods 152 cases of CRC, 57 cases of benign colorectal diseases and 59 healthy subjects were enrolled .ELISA was used to determine the sB7-H1 level in serum.The levels of carcinoembryonic antigen (CEA), carbohydrate antigen 19-9 (CA19-9) and carbohydrate antigen 72-4 (CA72-4) were determined by electrochemiluminescence. The level of carbohydrate antigen 50 (CA-50) was determined by chemiluminescence.ROC curve was used to evaluate the diagnostic efficiency of sB 7-H1 alone or combined with other tumor markers in CRC .It also analyzed the correlation between serum sB 7-H1 level and some clinicopathological characteristics including tumor location , depth of invasion , lymph node metastasis status , distant metastasis status and tumor stages.Results (1)The differences of sB7-H1 level among CRC group, colorectal benign disease group and healthy control group ( compared with each other respectively ) were statistically significant ( P<0.001 ) . Levels of CEA, CA19-9, CA72-4 and CA-50 were significantly different between CRC group and benign colorectal disease group ( P<0.05 ) , while there was no significant difference between benign colorectal disease group and healthy control group (P>0.05).(2)The diagnosis ability of sB7-H1, CEA, CA19-9,CA72-4, and CA-50 were determined by the area under curve (AUC), which was 0.730, 0.772, 0.639, 0.663 and 0.635, respectively.The combination of sB7-H1 and CEA showed the best effect , and the AUC reached 0.831.(3)There was no significant correlation between the levels of sB 7-H1 and CEA, or CA19-9, CA72-4, CA-50.(4) Immunohistochemistry result showed that the positive rate of CRC group was 52％, while that of benign disease group was 6.3％.The difference of the positive rate between two groups was statistically significant (P<0.01).And there was a significant correlation between the positive rate of sB 7-H1 level in serum and tumor tissues of CRC patients ( P<0.05 ) .( 5 ) The serum sB7-H1 level in CRC patients was related to lymph node metastasis status (P<0.05).But there was no relationship between sB7-H1 level and other clinicopathological characteristics ( including tumor location , depth of invasion , distant metastasis status and tumor stage ) (P>0.05).Conclusion Serum sB7-H1 has a high clinical application value and could serve as a new tumor marker in CRC .