Fecal Immunochemical Test Sensitivity for All Cases According to Regions, Seasons, Months and Temperature Levels and by Location of Tumors, Kaiser Permanente–Northern California (KPNC) 2007 to 2013
| All Cases | Right-Sided Cancers | Left-Sided Cancers | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Tests, n | Cases | Positive | Adjusted sensitivity, %* | Cases | Positive | Adjusted sensitivity, %* | Cases | Positive | Adjusted sensitivity, %* | |
| Overall | 1,141,162 | 1837 | 1425 | 75.8 (73.8 to 77.8) | 700 | 525 | 72.6 (69.2 to 75.9) | 1124 | 893 | 78.3 (75.9 to 80.7) |
| Temperature levels | ||||||||||
| 1 (Lowest) | 60,197 | 131 | 103 | 79.6 (77.8 to 81.4) | 39 | 29 | 76.9 (73.8 to 80.0) | 91 | 74 | 81.6 (79.4 to 83.8) |
| 2 | 322,864 | 465 | 361 | 76.2 (74.3 to 78.2) | 186 | 135 | 73.0 (70.0 to 76.4) | 275 | 223 | 78.4 (76.0 to 81.0) |
| 3 | 160,516 | 260 | 203 | 76.0 (74.0 to 78.0) | 107 | 83 | 72.8 (69.5 to 76.2) | 150 | 119 | 78.2 (76.0 to 80.7) |
| 4 | 528,129 | 845 | 657 | 75.2 (73.2 to 77.2) | 330 | 251 | 72.0 (68.6 to 75.4) | 510 | 403 | 77.5 (75.0 to 80.0) |
| 5 (Highest) | 69,456 | 136 | 101 | 75.7 (73.6 to 77.8) | 38 | 27 | 72.6 (69.2 to 76.1) | 98 | 74 | 77.9 (75.4 to 80.4) |
| Seasons | ||||||||||
| Winter | 239,873 | 348 | 269 | 76.4 (74.4 to 78.3)† | 132 | 95 | 73.2 (69.9 to 76.5) | 215 | 173 | 78.6 (76.2 to 81.0) |
| Spring | 393,119 | 554 | 427 | 74.9 (73.0 to 77.0) | 235 | 176 | 71.6 (68.2 to 75.0) | 315 | 249 | 77.3 (75.8 to 79.8) |
| Summer | 340,998 | 593 | 460 | 75.0 (73.0 to 77.0) | 210 | 159 | 71.8 (68.3 to 75.2) | 379 | 299 | 77.3 (74.8 to 80.0) |
| Autumn | 167,172 | 342 | 269 | 78.6 (76.8 to 80.5) | 123 | 95 | 75.7 (72.6 to 78.9) | 215 | 172 | 80.6 (78.5 to 83.0) |
| Test result month | ||||||||||
| December/January | 87,825 | 161 | 127 | 79.0 (77.0 to 80.7)† | 54 | 38 | 75.9 (72.8 to 79.1) | 106 | 89 | 80.9 (78.7 to 83.1) |
| February/March | 220,792 | 294 | 229 | 75.7 (73.7 to 77.7) | 118 | 86 | 72.4 (69.0 to 75.8) | 174 | 141 | 78.0 (75.5 to 80.4) |
| April/November | 175,195 | 283 | 218 | 76.4 (74.4 to 78.3) | 107 | 81 | 73.2 (70.0 to 76.5) | 173 | 135 | 78.6 (76.2 to 81.0) |
| May/October | 204,814 | 333 | 259 | 76.2 (74.3 to 78.2) | 148 | 117 | 73.3 (70.0 to 76.6) | 182 | 141 | 78.5 (76.1 to 81.0) |
| June/July | 249,656 | 380 | 299 | 74.5 (72.5 to 76.6) | 138 | 106 | 71.1 (67.8 to 74.7) | 239 | 191 | 76.9 (74.3 to 79.4) |
| August/September | 202,880 | 386 | 293 | 75.6 (73.6 to 77.6) | 135 | 97 | 72.4 (69.0 to 75.8) | 250 | 196 | 78.0 (75.5 to 80.3) |
Ambient temperature levels were defined using the mean and 1.25 times the standard deviation (SD) of maximum temperatures by year aggregated over weekly intervals, >mean + 1.25 SD, >0.25mean, >means-0.25 SD, >means-1.25 SD and <means-1.25 SD (reference), in that order. Thus, the highest and lowest temperature levels were 1.25 times the level above or below the mean, respectively.
↵* The P-values for the differences in sensitivities in pair-wise comparison using Bonferonni correction were all < .01 relative to this group.
↵† This model was adjusted for age, race/ethnicity, and sex. All models included a variable on colorectal cancer diagnosis and the sensitivity was derived using a predictive margins conditioned on cancer diagnosis.