The effect of rigorous study design in the research of autologous bone marrow-derived mononuclear cell transfer in patients with acute myocardial infarction

Table 5 Univariate and multivariate meta-regression analyses of BMSC treatment effects

Study characteristics	Univariate analysis		Multivariate analysis^c
	Mean difference (95% CI)	P	Mean difference (95% CI)	P
Bone marrow aspiration^a	−3.27 (−3.89 to −2.65)	<0.0001	−6.62 (−9.65 to −3.59)	0.0003
Catheterization^b	−0.54 (−3.57 to 2.50)	0.688	−5.55 (−9.35 to −1.76)	0.0071
Effects of study quality
Randomization	1.23 (−2.62 to 5.08)	0.474	−2.62 (−6.16 to 0.92)	0.135
Allocation concealment	0.43 (−3.54 to 4.39)	0.807	−0.05 (−4.69 to 4.59)	0.982
Effects of study characteristics
Cell dose	2.23 (−0.10 to 4.55)	0.059	1.45 (−3.42 to 6.32)	0.535
Injection time	2.10 (−0.15 to 4.35)	0.065	1.46 (−3.48 to 6.39)	0.538
Baseline LVEF	1.69 (−0.45 to 3.84)	0.114	3.40 (−1.66 to 8.46)	0.172

^aBone marrow aspiration in the control group. ^bCatheterization in the control group. ^cAdjusted by randomization, allocation concealment, cell dose, injection time, and baseline LVEF.

ISSN: 1757-6512