PubMed Health. A service of the National Library of Medicine, National Institutes of Health.

Chung M, Ip S, Yu W, et al. Interventions in Primary Care to Promote Breastfeeding: A Systematic Review [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US); 2008 Oct. (Evidence Syntheses, No. 66.)

2Methods

This report will be used by the USPSTF to update its 2003 recommendation on counseling to promote breastfeeding. Tufts-NEMC EPC, the Center on Primary Care, Prevention and Clinical Partnerships at AHRQ, and the USPSTF jointly developed an analytic framework and a set of study inclusion/exclusion criteria that are suitable to meet the USPSTF objectives. In addition, we utilized results from a recently completed AHRQ evidence report (Number 153)4 to answer two key questions.

Key Questions

1.

What are the effects of interventions to promote and support breastfeeding, in terms of short- and long-term child and maternal health outcomes?

2.

What are the effects of a) prenatal, b) peripartum, and c) postpartum breastfeeding interventions on breastfeeding initiation, duration, and exclusivity?

3.

Are there harms from interventions to promote and support breastfeeding?

4.

What are the benefits and harms for infants and children in terms of short-term outcomes, such as infectious diseases (including otitis media and diarrhea), development, and sudden infant death syndrome and infant mortality, and longer-term outcomes such as neoplastic diseases, autoimmune diseases (including type 1 diabetes), chronic diseases (including asthma, environmental allergies, type 2 diabetes, hypertension and hyperlipidemia), and obesity, compared among those who mostly breastfeed, mostly formula feed, and mixed feed; and how are these outcomes associated with duration of the type of feeding? Do the harms and benefits differ for any specific subpopulations based on socio-demographic factors?

5.

What are the benefits and harms on maternal health short-term outcomes, such as post-partum depression, anemia, and return to pre-pregnancy weight, and long-term outcomes, such as breast and ovarian cancer and osteoporosis, compared among breastfeeding, formula feeding, and mixed feeding, and how are these associated with duration of the type of feeding? Do the harms and benefits differ for any specific subpopulations based on socio-demographic factors?

The answers to key questions 4 and 5 have been reviewed in our previous report. A summary of the findings from that report is provided in Appendix G. This report focuses on key questions 1 to 3 concerning the effectiveness of primary care initiated interventions to promote and support breastfeeding in the prenatal, peripartum, and postpartum periods. We focused our review on studies conducted in developed countries. However, because of the widespread interest in the Baby Friendly Hospital Initiative1 (BFHI), randomized controlled trials (RCTs) of BFHI conducted in developing countries have also been included. Furthermore, because of the nature of the BFHI intervention, most of the observational studies on BFHI used a historical control for comparison. Thus, observational studies on BFHI using that study design have also been included.

Definitions used in this report

Definitions of “exclusive breastfeeding” varied widely in the literature. They ranged from “no supplement of any kind including water while breastfeeding” to “occasional formula is permissible while breastfeeding.” We elected to accept all definitions of “exclusive breastfeeding” as provided by the different study authors, but we qualified our findings by the details regarding those definitions.

Other categories (full, partial, mixed, non-specified) of breastfeeding besides exclusive breastfeeding are classified as “any” breastfeeding.

We have also defined the following categories of breastfeeding durations. Breastfeeding initiation is any breastfeeding at discharge or before 2 weeks post delivery; 1 to 3 months of breastfeeding is short-term; 4 to 5 months is intermediate-term; 6 to 8 months is long-term; and 9 months or more is prolonged. Breastfeeding shorter than 1 month was considered together with the “no breastfeeding” category. These categories of breastfeeding duration were arbitrary but defined a priori.

Types of breastfeeding intervention were classified broadly into three major categories: breastfeeding education, breastfeeding support (professional and/or lay support), and all other interventions. Many of the breastfeeding interventions contain more than one component.

a.

Formal/structured breastfeeding education directed at mothers and/or other family members:

Structured one-to-one or group education sessions or classes (e.g., curriculum or standard agenda)

b.

Breastfeeding support

(1)

Professional support:

  • System level support - BFHI; training of health professionals
  • Individual level support - one-to-one support during hospital stay or outpatient visits; social support (e.g., home visits or telephone support) from health professionals

(2)

Lay support: peer counseling; social support (e.g., home visits or telephone support) from peers

c.

Other interventions:

  • Skin-to-skin care
  • Pacifier use
  • Motivational interview

Overall Approach

This report updates the previous systematic review10 conducted for USPSTF, we focused our effort on primary studies published since Guise's search date of September 2001. We searched from January 2001 onward to ensure that all eligible studies have been accounted for. We elected not to systematically examine systematic reviews because all the other systematic reviews published since 2001 had search dates no later than 2002 except for one recent Cochrane review on support for breastfeeding mothers with a search date of November of 2005.13 In consultation with AHRQ and the USPSTF, we decided to capitalize on this Cochrane systematic review by including the data from eight RCTs in developed countries reported in the review in our meta-analysis. However, we did not reassess the quality of these eight trials. To be as comprehensive as possible, we elected to summarize the primary studies not covered (either by exclusion or because they were not yet published at the time) in the Cochrane review.

Study Selection

We included RCTs and controlled but not randomized trials. For BFHI, we also included before and after experimental studies and prospective cohort studies with concurrent or historical controls. Any counseling or behavioral intervention initiated from a clinician's practice (office or hospital) to improve breastfeeding initiation, duration, or both were considered. Examples of interventions include physician counseling, peer-to-peer support, and hospital practices such as those outlined in the Baby-friendly Ten Steps. Only English language studies were included.

Inclusion criteria for the studies are as follow:

  • Study Design. RCTs, clustered RCTs, quasi-RCTs, controlled, non-randomized studies; for BFHI: we also included non-randomized before and after experimental studies and prospective cohort studies with concurrent or historical controls
  • Population. Healthy term or near-term infants (≥35 weeks gestation or ≥ 2,500 g) and their healthy mothers and members of the mother-child support system (such as partners, grandparents, or friends)
  • Intervention. Intervention must be primary care initiated, conducted, or referable (i.e., if the primary care does not offer that service, it should have the ability to refer the mother-infant pair or family to that service). Potential interventions include but are not limited to counseling, structured education, support, distribution of written materials, and adoption in whole or in part of Baby-friendly Ten Steps. Interventions may be stand alone or multi-component/multi-dimensional. They could be conducted by a variety of providers (lactation consultants, nurses, peer counselors, midwives, or physicians) in a variety of settings (hospital, home, clinic, or community) as long as they are linked with the health care system and the provision of primary care. Health care system interventions, such as staff training, were included (N.B., examples of non-primary care initiated interventions that were excluded from this review include mass-media campaigns, work site lactation programs, community interventions, and peer-to-peer support that do not interact with the health care system). For the purpose of this review, maternity services are considered primary care.
  • Comparator. Usual prenatal, peripartum, and/or postpartum care defined within each study for women in the control groups.
  • Outcomes. Studies must report rates of breastfeeding initiation, duration of breastfeeding, or exclusivity of breastfeeding. Maternal or infant health outcomes reported in these studies are also included. For adverse events associated with breastfeeding interventions, we reviewed both included and excluded studies conducted in developed countries.

Search Strategy

We searched Medline, the Cochrane Controlled Trials Registry, CINAHL, and Cochrane Database of Systematic Reviews for articles from September, 2001 to January, 2007 using the MeSH terms and keywords, such as “breastfeeding”, “breast milk feeding”, “breast milk”, “human milk”, “nursing”, “lactation”, “counseling”, and “health education”. We also examined the bibliographies in existing systematic reviews for additional studies.

Data Extraction

One reviewer initially screened abstracts for possible inclusion. This initial screening used very broad criteria to ensure that all potentially relevant abstracts were included (i.e., any human studies with any kind of interventions to promote or support breastfeeding were screened in). A second person reviewed all the potentially relevant abstracts using the formal study inclusion/exclusion criteria. The full papers of the eligible abstracts were retrieved and examined in detail. After full articles evaluation, data from qualified studies were abstracted (Appendix B). Items of interest extracted were: study setting, population, control, description of intervention (type, person, frequency, and duration), definitions of breastfeeding (initiation, exclusivity, and duration) outcomes, definitions of health outcomes (when provided) in both mothers and children, and methods of analyses. We categorized interventions as multidimensional (as in Baby-friendly Ten Steps), individual or group education, in-person or telephone support, professional support or counseling, peer support or counseling, and miscellaneous category (written materials, rooming-in, early maternal contact, commercial discharge packets, and others).

Quality and Applicability Assessment

Two reviewers independently assessed the quality of published systematic reviews and controlled studies using criteria developed by the USPSTF.14 Each paper was assigned a quality rating of “good”, “fair”, or “poor”. The criteria of quality assessment for primary studies included the randomization techniques, clear definitions of outcomes, or intention to-treat analysis for RCTs and consideration for potential confounders in cohort studies. A third reviewer reviewed those studies in which the quality rating was discordant between the first two reviewers. Final grades in those studies were reached via consensus. We have also assessed the applicability (or external validity) of the study population to the United States primary care setting by examining the specific study conditions and population/sample characteristics. The overall assessment is categorized as either wide or narrow applicability.

Data Synthesis

Rates of breastfeeding initiation, short-term, intermediate-term, long-term, and prolonged breastfeeding were calculated for both the intervention and control groups in each study. The exclusivity of breastfeeding was recorded and the same calculations were performed for the exclusive breastfeeding rates. Moreover, the differences in the average duration of any or exclusive breastfeeding by the end of the study between the comparison groups were also calculated when the data are available.

The decision to combine studies in a meta-analysis and the subsequent selection of statistical methods can be challenging. Ideally, studies should only be combined if they are sufficiently homogeneous (i.e., comparable interventions, groupings, study designs, outcome measurements). In addition, the meta-analysis must be executed paying close attention to underlying assumptions and their attendant limitations. In this review, meta-analyses were performed, for RCTs and non-randomized but controlled studies to examine the effect of interventions on breastfeeding initiation, duration, and exclusivity. Although the studies in our meta-analyses are similar in design, they are still different in many respects: different combinations of intervention components and background social support, different health care systems defining “usual” or “routine” care, different timing and intensities of the interventions, and diverse study populations. Therefore, we also performed various subgroup analyses to analyze the heterogeneity across studies.

Meta-analyses and Meta-regression

To avoid multiple counting of the same study and subsequent improper weighting, we selected data from the longest duration of breastfeeding within each breastfeeding category to ensure one study enters the analysis only once. For example, if a study reported data on both 1- and 3-month breastfeeding rates, only the 3-month breastfeeding rate was selected for the analyses. We also included data from one recent Cochrane systematic review13 that reported findings from RCTs conducted in developed countries. Data on breastfeeding initiation, duration, and exclusivity from those studies were abstracted from the review and incorporated into our meta-analyses. Breastfeeding data reported in the Cochrane systematic review were verified. We used the data reported in the original publications in instances of inconsistencies.

We used the DerSimonian and Laird's random effects model for all meta-analyses.15 We tested for heterogeneity using Cochran's Q and assessed its extent with I2, which evaluates the proportion of between study variability that is attributed to heterogeneity rather than chance.16, 17

Subgroup analyses were performed to examine the impacts of study quality, the effects of timing of intervention (prenatal, postpartum, or combined prenatal and postpartum), and different components of breastfeeding interventions on breastfeeding initiation, duration, and exclusivity.

A random-effect meta-regression18, 19 was performed to test the association between the effects of interventions and breastfeeding durations when at least six data points were available. A significant p-value indicated an increasing or decreasing trend for the effects of breastfeeding promotion, compared to the control, with an increasing or decreasing breastfeeding durations. We reported our results using rate ratios and 95% confidence intervals. Intercooled Stata 8.2 was used for the calculations and graphics.

1 Written breastfeeding policy that is routinely communicated to all health care staff; train all health care staff in skills necessary to implement this policy; inform all pregnant women about the benefits and management of breastfeeding; help mothers initiate breastfeeding within one half-hour of birth; show mothers how to breastfeed and maintain lactation, even if they should be separated from their infants; give newborn infants no food or drink other than breast milk, unless medically indicated; practice rooming in - that is, allow mothers and infants to remain together 24 hours a day; encourage breastfeeding on demand; give no artificial teats or pacifiers to breastfeeding infants; foster the establishment of breastfeeding support groups and refer mothers to them on discharge from the hospital or clinic. (http://www​.unicef.org​/programme/breastfeeding/baby.htm)

Footnotes

1 Written breastfeeding policy that is routinely communicated to all health care staff; train all health care staff in skills necessary to implement this policy; inform all pregnant women about the benefits and management of breastfeeding; help mothers initiate breastfeeding within one half-hour of birth; show mothers how to breastfeed and maintain lactation, even if they should be separated from their infants; give newborn infants no food or drink other than breast milk, unless medically indicated; practice rooming in - that is, allow mothers and infants to remain together 24 hours a day; encourage breastfeeding on demand; give no artificial teats or pacifiers to breastfeeding infants; foster the establishment of breastfeeding support groups and refer mothers to them on discharge from the hospital or clinic. (http://www​.unicef.org​/programme/breastfeeding/baby.htm)

Cover of Interventions in Primary Care to Promote Breastfeeding
Interventions in Primary Care to Promote Breastfeeding: A Systematic Review [Internet].
Evidence Syntheses, No. 66.
Chung M, Ip S, Yu W, et al.

PubMed Health Blog...

read all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...