Table 2. Effect of study-level characteristics on medication adherence (n = 27).

Variable	Subgroups	No. of studies	No. of participants	SMD	CI	p-value for interaction*
Age	Children (<18 years)	4	467	1.28	0.36 to 2.2	0.37
	Adults (≥18 years)	23	2097	0.85	0.62 to 1.02
Healthcare setting	Primary / ambulatory care /outpatient	20	2023	0.85	0.59 to 1.10	0.32
	hospital setting	7	561	1.14	0.63 to 1.64
Complexity of the intervention	EAM only	3	235	0.24	-0.35 to 0.84	0.01
	EAM + EAM reminder	7	829	1.51	0.81 to 2.22
	EAM + health professional input	13	1288	0.73	0.54 to 0.93
	EAM + EAM reminder + health professional input	4	232	1.05	0.77 to 1.33
Method of adherence feedback a	Direct-to-patient	6	655	0.96	0.37 to 1.54	0.15
	Healthcare provider facilitated	12	1062	0.79	0.54 to 1.03
	Both	7	665	1.42	0.82 to 2.01
Timing of adherence feedback	Real time (immediate)	9	1048	1.03	0.66 to 1.40	0.53
	Delayed	18	1536	0.88	0.57 to 1.19
Study duration	Short (6 months)	11	856	0.70	0.40 to 0.99	0.09
	Long (over 6 months)	16	1728	1.08	0.74 to 1.43
Blinding b	Blinded to adherence monitoring function	7	741	1.06	0.59 to 1.53	0.09
	No blinding	10	982	0.62	0.46 to 0.79

SMD, standardised mean difference; CI, confidence intervals for effect size; EAM, Electronic medication monitor

^aTwo studies did not provide adherence feedback to participants and were not included in the subgroup analysis.

^bIt was not clear in 10 studies whether participants were blinded to adherence monitoring. These were excluded from the subgroup analysis.