Pragmatic randomised trial of a smartphone app (NRT2Quit) to improve effectiveness of nicotine replacement therapy in a quit attempt by improving medication adherence: results of a prematurely terminated study

Table 2 Cessation outcomes in NRT2Quit trial (smoking status in past 4 weeks assessed at 8-week follow-up)

	Total (n = 41)	Intervention (n = 16)	Control (n = 25)	P value*	OR (95% CI)	Bayes factor^a uniform (OR of 1 to 3)	Bayes factor^a half normal (OR = 1.2, 1.7, 2.5)
Primary cessation outcome (verified)	% (n)	% (n)	% (n)
Smoking status ITT¹
Not smoking	14.6 (6)	25.0 (4)	8.0 (2)	0.19	3.83 (0.61–24.02)	1.92	1.24, 1.70, 1.99
Assumed to be smoking	85.4 (35)	75.0 (12)	92.0 (23)		–	–	–
Secondary outcome (self-reported)		% (n)
Smoking status ITT
Not smoking	17.1 (7)	25.0 (4)	12.0 (3)	0.40	2.44 (0.47–12.78)	1.52	1.18, 1.41, 1.43
Assumed to be smoking	82.9 (34)	75.0 (12)	88.0 (22)		–	–	–
Smoking status ITT
Not smoking	17.1 (7)	25.0 (4)	12.0 (3)	0.12	–	–	–
Smoking <5 cigarettes	2.4 (1)	6.3 (1)	0.0 (0)		–	–	–
Smoking ≥5 cigarettes	31.7 (13)	12.5 (2)	44.0 (11)		–	–	–
Not contacted/assumed to be smoking	488. (20)	56.3 (9)	44.0 (11)		–	–	–

¹Two subjects self-reporting not smoking had salivary cotinine of more than 100 ng/mL and anabasine levels of 0.2 ng/mL. With a lower cutoff value for salivary anabasine level suggested recently by the processing lab (<0.2 ng/mL), 18.8% of intervention and 4.0% of control participants met criteria for biochemical verification (odds ratio (OR) = 5.54, 95% confidence interval (CI) 0.52–58.76).
*For 2 × 2 analysis, the P value reported is for Fisher’s exact test; otherwise, for Pearson chi-squared.
^aBayes factor < 1/3 suggests support for the null hypothesis, Bayes factor > 3 suggests support for the experimental hypothesis, and intermediate values suggest that the data are insensitive (Dienes et al., [21], Brown et al., [12]).
Abbreviation: ITT intention to treat

ISSN: 1745-6215