. Author manuscript; available in PMC: 2020 Jul 2.

Published in final edited form as: J Am Board Fam Med. 2019 May-Jun;32(3):318–328. doi: 10.3122/jabfm.2019.03.180275

Table 2:

Fecal immunochemical test (FIT) completion across study arms, intention-to-treat analysis using electronic health record data obtained six-months following randomization

	FIT completion rates						Adjusted difference in proportions^*
	Automated (A)		Live (B)		Automated plus Live (C)		B vs. A	C vs. A	C vs. B
	N	% completed FIT (95% CI)	N	% completed FIT (95% CI)	N	% completed FIT (95% CI)	Adjusted difference (95% CI)	Adjusted difference (95% CI)	Adjusted difference (95% CI)
Intention-to-treat	589	26.0 (22.4, 29.5)	589	32.3 (28.5, 36.0)	589	35.7 (31.8, 39.5)	6.3 (1.1, 11.4)	9.7 (4.4, 14.9)	3.4 (− 2.0, 8.8)
Per-protocol ^†	348	29.3 (24.5, 34.1)	487	36.3 (32.1, 40.6)	326	41.4 (36.0, 46.8)	6.9 (0.6, 13.4)	12.1 (4.9, 19.2)	5.1 (−1.8, 11.9)

Adjusted for clinic

^†

Per-protocol: Automated: successful delivery of text message and at least 1 automated phone call; Live: successful delivery of at least 1 live phone call (left message or spoke to patient); Automated plus live: successful delivery of text messages, at least 1 automated phone call, and at least 1 live phone call (left message or spoke to patient).