Table 2

Statistical comparisons of five endometrial lesions
Parameter a vs. Parameter b Positives (Parameter a) Positives (Parameter b) χ2value* p value*
Simple hyperplasia vs. atypical complex hyperplasia 8/27(30%) 20/41(49%) 2.465 0.116
Simple hyperplasia vs. endometrioid carcinoma 8/27(30%) 54/103(52%) 4.457 0.035
Simple hyperplasia vs. endometrial serous carcinoma 8/27(30%) 8/21(38%) 0.381 0.537
Simple hyperplasia vs. clear cell carcinoma 8/27(30%) 6/9(67%) 3.50*** 0.111**
Atypical complex hyperplasia vs. endometrioid carcinoma 20/41(49%) 54/103(52%) 0.156 0.693
Atypical complex hyperplasia vs. endometrial serous carcinoma 20/41(49%) 8/21(38%) 0.640 0.424
Atypical complex hyperplasia vs. clear cell carcinoma 20/41(49%) 6/9(67%) 4.32*** 0.467**
Endometrioid carcinoma vs. endometrial serous carcinoma 54/103(52%) 8/21(38%) 1.433 0.231
Endometrioid carcinoma vs. clear cell carcinoma 54/103(52%) 6/9(67%) 4.18*** 0.500**
Endometrial serous carcinoma vs. clear cell carcinoma 8/21(38%) 6/9(67%) 4.20*** 0.236

*: We took partitioning Chi-squares when individual comparisons were performed, in order to avoid amplifying the probability to commit type I mistake, p value should be reset as p’ = 1-m√(1-p), “m” represented group numbers. In this case, p’ = 1-5√(1-0.05) = 0.005, so we considered p < 0.005 as statistically significant.

**: When expected cell value was 5 or less, we used Fisher’s exact test.

***: The expected cell values were 3.50, 4.32, 4.18 and 4.20 for simple hyperplasia vs. clear cell carcinoma, atypical complex hyperplasia vs. clear cell carcinoma, endometrioid carcinoma vs. clear cell carcinoma and endometrial serous carcinoma vs. clear cell carcinoma, respectively.

Liang et al.

Liang et al. Diagnostic Pathology 2013 8:138   doi:10.1186/1746-1596-8-138

Open Data