Table 3 Using 5-fold cross validation, each woman in the dataset received a score from the model built to predict ovarian cancer. The women were ranked by score from lowest to highest and then partitioned into five quintiles. This table presents the number of women with and without ovarian cancer in each quintile along with the odds ratio (relative to the entire group) and the 95% confidence interval for the odds ratio
From: Genetic risk score for ovarian cancer based on chromosomal-scale length variation
Quintile | Number of women without ovarian cancer | Number of women with ovarian cancer | Total number of women | Odds ratio | 95% confidence interval |
|---|---|---|---|---|---|
1 | 925 | 3 | 928 | 0.03 | 0.01--0.09 |
2 | 925 | 3 | 928 | 0.03 | 0.01--0.09 |
3 | 901 | 27 | 928 | 0.30 | 0.21--0.45 |
4 | 842 | 86 | 928 | 1.04 | 0.82--1.33 |
5 | 632 | 295 | 927 | 4.76 | 4.01--5.65 |