Table 3.

Carbohydrate counting accuracy at baseline and follow-up, by meals

			Qualifiers	CHOa estimate accuracy, gb
	Meal	Average CHO, g	Group	Baseline	Follow-up	p- valuec	p- valued
Day 1	breakfast	52.3	control	−8.1	−6.2	0.39	0.91
			intervention	−12.3	−10.7	0.50
	lunch	83.4	control	−14.9	−15.5	0.86	0.29
			intervention	−18.5	−13.8	0.19
	dinner	48.9	control	9.2	7.9	0.70	0.036
			intervention	1.8	10.7	0.012
	snack	66.0	control	7.7	6.3	0.65	0.32
			intervention	2.2	5.1	0.35
Day 2	breakfast	76.6	control	−2.4	−0.3	0.54	0.92
			intervention	−9.2	−6.5	0.46
	lunch	92.0	control	10.2	10.3	0.99	0.71
			intervention	−2.4	0.2	0.59
	dinner	162.0	control	−10.5	−10.7	0.97	0.82
			intervention	−18.2	−16.5	0.78
	snack	20.5	control	8.3	3.0	0.014	0.62
			intervention	8.1	4.3	0.085
Day 3	breakfast	63.4	control	25.7	21.5	0.35	0.60
			intervention	23.3	22.4	0.85
	lunch	54.0	control	12.6	8.5	0.15	0.67
			intervention	10.6	8.2	0.43
	dinner	52.4	control	15.6	18.3	0.52	0.29
			intervention	4.3	13.5	0.040
	snack	71.8	control	7.4	7.3	0.97	0.51
			intervention	1.3	4.1	0.38

^aCHO = carbohydrate;

^baccuracy was defined as a difference between actual and estimated CHO;

^csignificance of change between baseline and follow-up accuracy within the group, adjusted for age, gender,and diabetes duration;

^dsignificance of the overall intervention effect.