User's Manual

PillCam Capsule Endoscopy
20 Accuracy of the Device—COLON 2
Using the 6 mm or greater sized polyp threshold, the positive and negative percent agreements of CCE with
OC were, respectively, 68.8% (132/192, 95% CI 61.7-75.2%) and 81.3% (413/508, 95% CI 77.6-84.6%).
Using the 10 mm or greater sized polyp threshold, the positive and negative percent agreements were
64.9% (50/77, 95% CI 53.2-75.5%) and 92.9% (579/623, 95% CI 90.6-94.8%).
Following the completion of a colonoscopy when the polyp detected on CCE was not identified, the results
of the CCE evaluation were unblinded and the colonoscopy was repeated in a second attempt to identify the
polyp identified on CCE. The unblinded colonoscopy results for the >=6 mm threshold showed a 1% and
3% increase in the positive and negative percent agreement values, respectively. The unblinded
colonoscopy results for the >=10 mm threshold showed a 1% increase in the positive and negative percent
agreements.
i ii
Notes:
i. Five of ninety-five patients who per protocol should have undergone a second OC procedure were determined not to require it by
the physician.
ii. The reported positive and negative percent agreement of the unblinded analysis are potentially biased because OC was repeated
only on those patients with an initial CCE false positive finding and not also on those patients with an initial CCE true negative
finding based on the blinded OC evaluation.
Table 4: Observed Categorical Data, 10 mm Threshold
CCE
OC
negative positive
negative 579 27 606
positive 44 50 94
623 77 700
Table 5: Percent Agreement of CCE with OC
Percent
Agreement
Polyp Size Threshold
≥ 6 mm ≥ 10 mm
Positive
Percent
Agreement
(95% CI)
a
68.8% (132/
192)
95% CI
(61.7%,
75.2%)
64.9% (50/
77)
95% CI
(53.2%,
75.5%)
Negative
Percent
Agreement
(95% CI)
a
81.3% (413/
508)
95% CI
(77.6%,
84.6%)
92.9% (579/
623)
95% CI
(90.6%,
94.8%)
a. 95% CI based on Clopper Pearson