Table 3

Summary of findings for effect of breastfeeding promotion interventions on ‘no breastfeeding’ rates.

Quality Assessment

Summary of Findings


Directness

No of events


No of Studies

Design

Limitations

Consistency

Generalizability to population of interest

Generalizability to intervention of interest

Intervention

Control

Relative Risk (95% CI)

Comments


Rate of no breastfeeding at day 1: low outcome-specific quality


38

21 RCTs, 17 QE [11,13,14,17,33,36,37,44,49,59,63,64,68-71,74,78,79,89-91,96,100,103,107,108,110,111,113,117,120-122,124-127]

Variable follow up periods used in studies. Recall criteria variable across studies (past 24 hr, past week or previous month).

Most studies suggest benefit. Significant heterogeneity

10 of 38 studies were conducted in developing countries

Pooled results for different types of interventions

48026

39843

0.68 [0.54, 0.87]

Random effects meta-analysis due to significant heterogeneity

Most studies used individual counseling and most were facility and community-based.

Effect of benefit refers to decrease in numbers not breastfeeding.


Rate of no breastfeeding at <1 month: low outcome-specific quality


33

21 RCTs, 12 QE [10,18,19,38,45,52,53,61,63,66,68,71-74,76,88,90,91,93,95,98,99,101-103,106,113,117,118,120,124,128]

Variable follow up periods used in studies. Recall criteria variable across studies (past 24 hr, past week or previous month).

10 of 33 studies suggest benefit. Significant heterogeneity

4 of 33 studies were conducted in developing countries

Pooled results for different types of interventions

770

1018

0.70 [0.62, 0.80]

Random effects meta-analysis due to significant heterogeneity.

Most studies used individual counseling.


Rate of no breastfeeding at 1-5 months: low outcome-specific quality


73

41 RCTs, 32 QE [10,12,13,17-20,35,36,38,40-42,44,45,47,49,51,52,54,58,59,61-64,67-69,71-74,76,78-83,86,90-93,95-99,101,103-106,108-110,112-121,123,124,126,129,130]

Variable follow up periods used in studies. Recall criteria variable across studies (past 24 hr, past week or previous month).

25 of 73 studies suggest benefit. Significant heterogeneity

16 of 73 studies were conducted in developing countries

Pooled results for different types of interventions

15473

17578

0.82 [0.77, 0.89]

Random effects meta-analysis due to significant heterogeneity

Most studies used individual counseling.


Haroon et al. BMC Public Health 2013 13(Suppl 3):S20   doi:10.1186/1471-2458-13-S3-S20

Open Data