Table 5.

		Quality Assessment					Summary of Findings
					Generalisability		Intervention		Control
No of studies (ref)	Intervention	Design	Limitations	Consistency	To population of interest	To intervention of interest	No. Events	No. births	No. events	No. births	Effect size (95% CI)
Neonatal Mortality (Tetanus Deaths):
1 (18)	CBK and education	Before and after	Low quality		Masai population*	Yes	0	1984	415	5716	RR = 0.01 (0.001 - 0.09)
1 (80)	CBK, TT plus multiple interventions	Before and after	Multiple interventions		India. Lady health worker delivered	Multiple interventions	0	1951	2	1958	OR = 0.20 (0.01 - 4.18)
Neonatal Mortality (Sepsis Deaths):
No studies identified
Neonatal Mortality (All Cause):
1 (81)	CBK plus multiple interventions	cRCT	Multiple interventions		Pakistan TBA delivered	Multiple interventions	340	10092	439	19432	aOR 0.71 (0.62 - 0.83)
1 (18)	CBK and education	Before and after	Low quality		Masai population*	Yes	99	1984	1984	5716	RR = 0.17 (0.13 - 0.23)
1 (80)	CBK, TT plus multiple interventions	Before and after	Multiple interventions		India lady health worker delivered	Multiple interventions	35	1951	45	1958	OR 0.78 (0.50 - 1.21)
Neonatal Sepsis incidence:
1 (84)	CBK and demonstration	Before and after	Observational		Papua New Guinea**	Yes	1	67	8	64	OR = 0.11 (0.01 - 0.84)
Cord infection/oomphalitis:
3 (56, 58, 59)	CBK use	Adopters vs non-adopters	Observational	Heterogeneous	Egypt, Tanzania, Nepal	Yes					aOR 0.08-0.45
1 (11)	Use of individual items in CBK	Adopters vs non-adopters	Observational		Nepal	Yes					soap aRR = 0.49 (0.43-0.56)^

*Specific cultural practices and defined neonatal death as death occuring in first 6 weeks of life **Specific cultural practices

^ no effect of other components on multivariable analysis TT = tetanus toxoid vaccination