Pragmatic evaluation of methods for retrieving unpublished information on comparator interventions in a systematic review of smoking cessation trials

Figures & data

Figure 1. The frequencies of reporting of individual comparator contextual characteristics when examining published and unpublished data sources.

Note. Bars omit ‘Materials sent by study authors’ where these did not provide any additional information relevant to that contextual characteristic.

Table 1. Sources of comparator intervention BCTs in smoking cessation trials that provided a completed checklist and/or unpublished materials.

	N of BCTs	% of total BCTs
Source of information (N = 91)*
Published article	304	16.2
Published supplementary materials	173	9.2
Unpublished materials sent by authors	413	21.9
Checklist	986	52.6
Total	1882	100.0
Source of information (N = 40)**
Published article	154	15.1
Published supplementary materials	64	6.4
Unpublished materials sent by authors	334	33.9
Checklist	454	44.6
Total	1017	100

Note.

*For 91 studies, authors provided additional information either in the form of unpublished materials, checklist or both.

**For 40 studies, authors provided additional information both in the form of unpublished materials and the checklist.

Table 2. Number of BCTs identified in published and combined information sources for different types of providers and different types of comparator interventions.

		Published		Combined
	N	M	SD	M	SD
Provider	73
Nurses, physicians, pharmacists	28	4.28	5.5	13.35	12.43
Psychologists, counsellors	45	6.16	5.47	26.98	14.48
Type of comparator intervention	84
Brief support	24	5.42	5.98	13.17	11.36
Treatment-as-usual	40	4.45	5.02	19.88	15.26
Psychological support	20	7.90	6.34	35.05	8.92

Note. Provider information available for 73 of 91 trials (80.2%); Information on type of comparator intervention available for 84 of 91 trials (92.3%).

Table 3. Simple and interaction effects results found for the differences in the N of BCTs for different types of provider (research question 4a) and types of comparator (research question 4 b) in published and combined information sources.

Effect	MS	Df	F-statistic	p value*	η^2p
Provider (published)	62.31	1	2.07	0.15
Provider (combined)	3277.62	1	17.42	< 0.01
Provider x Source	1218.05	1	16.42	< 0.01	0.18
Type of comparator (published)	79.64	2	2.51	0.09
Type of comparator (combined)	2721.96	2	16.25	< 0.01
Type of comparator x Source	1033.28	2	15.91	< 0.01	0.28

Note. Provider information available for 73 of 91 trials (80.2%); information on type of comparator intervention available for 84 of 91 trials (92.3%).