An explanatory randomised controlled trial of a nurse-led, consultation-based intervention to support patients with adherence to taking glucose lowering medication for type 2 diabetes

Table 3 Outcomes twelve weeks after randomisation to intervention or standard care

	Intervention arm		Standard care arm		Intervention effect (95% C.I)	P-value
	Baseline	Final visit	Baseline	Final visit
Primary outcome
Days correct dose taken (SD) (%)*	-	77.4 (26.3)^a		69.0 (30.8)^b	8.4 (0.2, 16.7) ^c	0.044
Secondary outcomes
SF12 Physical (SD) ^†	43.3 (11.2) ^r	44.6 (11.1) ^c	45.5 (10.3) ^r	46.3 (9.0) ^l	-0.7 (-2.7, 1.4) ^s	0.52
SF12 Mental (SD) ^†	48.0 (10.6)	49.5 (10.4) ^c	50.0 (9.63) ^t	52.6 (8.8)^l	-1.6 (-3.9, 0.6) ^s	0.15
Diabetes treatment satisfaction (SD)	-	30.6 (5.4) ^c	-	31.3 (4.6) ^f	-0.7 (-2.2, 0.7) ^u	0.32
HbA_1c % (SD) ^†	8.34 (1.26) ^k	8.34 (1.24) ^k	8.29 (1.23) ^l	8.21 (1.32) ^f	0.06 (-0.19, 0.32) ^m	0.64
HbA_1c mmol/l (SD)	67.7 (13.8)	67.7 (13.6)	67.1(13.4)	66.2 (14.4)	0.7 (- 2.1, 3.5)
Medication-related
MARS Self report adherence (SD) ^†	23.6 (2.3) ^d	23.6 (2.6) ^e	23.6 (2.8) ^f	24.1 (1.6) ^g	-0.4 (-1.0, 0.2) ^h	0.20
Satisfaction with communication (median, IQR)**		4 (2,5)		4 (2,5)		0.13
% reporting hypoglycaemia (N)***		1.6% (2)		0,0 (0)		0.52*

Values are mean (SD) unless otherwise stated. Missing values: a = 13, b = 4, c = 17, d = 15, e = 27, f = 8, g = 12, h = 39, k = 16, l = 9, m = 24, r = 1, s = 26, t = 2, u = 25.
* - Estimate of intervention effect adjusted for baseline where available and 95% confidence interval derived from the non-parametric bootstrap percentile method with 50,000 replications. † = In the analysis of the final visit measure adjusting for a baseline that has occasional missing data, the missing indicator method is used. ** p value from Mann–Whitney test., *** p value from Fisher’s exact test. MARS Medication Adherence Report Scale.

ISSN: 2731-4553