Table 2.

Results of Meta-analysis Assessing the Effect of Experimental Treatment vs. Control Based on the Mean Difference in Change on Alveolar Bone Measurements and Utilizing a Random Effects Model

		Treatment Effect Test	Heterogeneity Test
Outcome	Mean Differencea	p Valueb	Q Valuec (df)	p Valued	I2, %e	Tauf
Buccolingual	1.89 (1.41, 2.36)	< .001	12.28 (5)	.031	59.3	0.44
Midbuccal	2.07 (1.03, 3.12)	< .001	48.83 (5)	<.001	89.8	1.22
Midlingual	1.18 (0.17, 2.19)	.022	35.77 (4)	<.001	88.8	1.08
Mesial	0.48 (0.18, 0.79)	.002	5.33 (4)	.26	24.9	0.17
Distal	0.24 (–0.05, 0.53)	.102	2.92 (4)	.57	0.0	0.0

^aThe effect size estimated is the difference (treatment – control) in mean change in the specified alveolar bone measurement and represents the gain associated with treatment. Values in millimeters, mean (95% confidence interval).

^bSignificance probability associated with the test of a significant treatment effect.

^cTest statistic and degrees of freedom associated with the Cochran Q test of heterogeneity of treatment effect among studies.

^dSignificance probability associated with the Cochran Q test of heterogeneity of treatment effect among studies.

^eEstimated proportion of total variation across studies that is due to heterogeneity.

^fEstimate of the standard deviation of underlying effects across studies, reflecting variability among effect sizes.