The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed in this review [35]. A PRISMA checklist and search terms are available in Multimedia Appendix 1.

A total of 8 electronic databases were searched to identify potential studies: Scopus, Web of Science, the Cochrane Database of Systematic Reviews, ScienceDirect, Google Scholar, the Association for Computing Machinery, SpringerLink, and Institute of Electrical and Electronics Engineers Xplore. The registers of the following trials were also searched to identify any existing relevant trials: Cochrane Central Register of Controlled Trials, WHO International Clinical Trials Registry Platform, International Clinical Trials Registry Platform, and ClinicalTrials.gov.

Peer-reviewed studies, including quantitative and quality research, mixed methods, descriptive studies, randomized controlled trials, and quasi-experimental design trials with or without blinding, were included in the review. The research methodology was not limited in any way. Any papers examining interventions employing e-technologies that required internet access and aimed at addressing any breastfeeding outcome were included. Considering that SMS messaging can also be used without internet access, studies that used SMS messaging have been excluded from the review. The inclusion and exclusion criteria are shown in Textboxes 1 and 2.

Participants were healthy pregnant or postnatal women, either primiparous or multiparous, who intended to breastfeed. In addition, studies targeting both parents as study participants were also included if the intervention targeting women was described separately. Women of all ages, ethnicity, occupation, and any socioeconomic status were included.

The search was conducted on November 2018, whereas the screening stages occurred between January and March 2019. The following main key terms were used in the search strategy: nursing, breast-feed, “breast-feeding,” “breast feeding,” “breast milk” and lactation; and computers, telehealth, computer-mediated and mobile application. A detailed example of the search strategy tailored for Scopus is outlined in Textbox 3. This search strategy was adapted for each of the other databases (Multimedia Appendix 2).

The retrieved studies for the search were transferred to Elsevier’s Mendeley Desktop Reference Manager by the first author (AA) to download papers and remove duplicates. The first author screened all titles and abstracts of the papers, and the full texts were reviewed to identify eligible studies. Then, 3 authors (AA, SM, and KW) discussed the suitability of the eligible studies for the final review. Studies that investigated maternal depression, HIV/Hepatitis C virus, smoking, diabetes, alcohol, overweight/obesity, fertility, prematurity, or cesarean section were excluded from the review. A table of the excluded studies with reported reasons is included in Multimedia Appendix 5. A diagram of the systematic review and data extraction is presented in Figure 1.

A self-adaptive narrative synthesis approach with thematic analysis [36] was used to categorize the purpose of internet-based interventions employing e-technologies and the strategies that were used in those interventions. Meta-analysis could not be processed because of the substantial heterogeneity existing among the studies. However, breastfeeding outcomes are presented in a narrative form.

The search identified 5201 papers. After excluding duplicates, 2599 papers were screened for eligibility and 615 remained (Figure 1). After examination of the full texts, 17 studies fully met the inclusion criteria. The quality of the studies was appraised by 2 researchers (AA and SM), using the mixed methods appraisal tool [37,38]. After the quality appraisal, 1 study was excluded [39], and a total of 16 studies remained (Figure 1). Any disagreement on quality checks was resolved through discussion between the 3 authors (AA, SM, and KW). Tables showing the quality appraisal process are presented as additional files in Multimedia Appendix 4.

The included studies involved 4018 participants from 8 different countries: the United States (n=5), Australia (n=3), Finland (n=3), Taiwan (n=1), Canada (n=1), Thailand (n=1), the Philippines (n=1), and Ireland (n=1). The interventions in the included studies were delivered to women prenatally or postnatally, and all reported one or more breastfeeding outcomes (Table 1). Only 3 studies reported using a theoretical framework to inform the design of the intervention: the sustained breastfeeding framework based on predicting and changing behavior theory [40], Bandura’s [41] social cognitive theory [42], and Bandura’s [43] self-efficacy theory [44]. The description of the included studies is provided in Multimedia Appendix 3.

Characteristics of the internet-based e-technology interventions in the 16 included studies were categorized by the mode of e-technology delivery, purpose of the intervention, and key strategies that were used in those interventions.

Breastfeeding intention, knowledge, exclusivity, and duration were the most reported outcome measures in the studies (Table 4).

Of 11 web-based breastfeeding interventions that had a combination of education and support focus, 8 demonstrated improvements in breastfeeding outcomes [42,44,47-49,51,52,54].

For instance, Huang et al [52] evaluated breastfeeding knowledge in a quasi-experimental study (n=120) where there was a significant difference in breastfeeding knowledge level between the intervention and control groups in the posttest results (P<.001), with no differences in the pretest knowledge level. In addition, the exclusive breastfeeding rates were statistically higher at 3 to 5 days and at 2, 4, and 6 weeks in the intervention group (n=60) compared with the control group (n=60; 48.3%, 45%, 31.7%, and 26.7% versus 38.3%, 20%, 20%, and 20%, respectively) [52].

Exclusive breastfeeding during hospital stay has been shown to improve with web-based interventions [44,51,54]. Hannula et al [51] evaluated the effect of a web-based breastfeeding educational support system on exclusive breastfeeding rates in 705 women in Finland. The intervention group (n=431) had access to a website that offered articles, pictures, videos, and an educational game. During hospital stay, the intervention group breastfed exclusively more often than the control group (71% versus 58%; P<.001) and likewise on discharge (76% versus 66%; P=.01). In addition, 2 other quasi-experimental studies conducted in Finland with a convenience sample of 1300 [44,54] also indicated higher exclusive breastfeeding rates in a group of women who used web-based interventions compared with the control group during their stay in the hospital. Salonen et al [44] reported that exclusive breastfeeding was more common in the intervention group (51.0% versus 27.4%; P<.001) and similar findings were also reported by Salonen et al [54] (50.1% versus 30.4%; P<.001).

Providing web-based breastfeeding intervention with an interactive and asynchronous online discussion board was found useful in motivating intervention mothers to continue breastfeeding for longer periods of time [49]. For instance, in a longitudinal cohort study among regional Western Australian women, using online support and discussion forums was associated with higher exclusive breastfeeding rates at 6 months among women who lived in remote areas (n=10; 5.9% versus 0.6%; P=.01) [49].

Access to a website with information on breastfeeding has also been offered to 127 women in a recent study in Ireland along with sending emails up to 6 months postpartum and providing access to a breastfeeding helpline and support group [47]. The study demonstrated a positive impact on breastfeeding duration, as participants from both urban and rural areas acknowledged an increase in breastfeeding duration because of participation in the study, 42.2% and 86.7%, respectively. However, urban women (42% at 6 weeks and 53.3% at 3 months) reported no impact on the length of time they breastfed from participation in the study, compared with rural women (26.7% at 6 weeks and 13.3% at 3 months). This study found that providing additional support postnatally, including exclusive access to the educational website, was more preferred by women in rural areas compared with those in urban areas, with 46.7% and 22%, respectively [47].

Geoghegan-Morphet et al [48,58] designed a web-based breastfeeding support clinic called the Maternal Virtual Infant Nutrition Support clinic to offer evidence-based breastfeeding education combined with peer and professional support. The resource had 4 aspects: (1) text, graphics, and video breastfeeding educational resources; (2) an interactive discussion forum for participants, which is supervised and facilitated by an international board–certified lactation consultant; (3) data collection capacity; and (4) a web-based infant journal for text and photo entries. The study found that 60.8% of mothers were exclusively breastfeeding at 6 months postpartum in the intervention group compared with 28.2% of mothers, similar to statistics from the Canadian 2010-2013 data for Ontario mothers [58].

A US-based pilot study [50] designed the HealthyMoms intervention aimed to educate women about breastfeeding using a game-based learning platform (3D Gamelab) on the web. A total of 3 quests on the 3D Gamelab platform were completed by the study participants and focused on three main topics: deciding about breastfeeding, feeding your baby, and getting support. Each one of the quests focused on a specific breastfeeding topic using web-based education activities, such as watching a video, reading a short introduction, or adding posts and responses to the information. Breastfeeding self-efficacy was measured with the 14-item breastfeeding self-efficacy scale—short form (BSES-SF), and breastfeeding intention was assessed using a 1-item measure with 4 categories (eg, just breastfeed/no formula, just formula/no breastfeeding, both, or unsure). The majority of mothers (68%) reported a high intention for exclusive breastfeeding; however, a one-way analysis of variance test found no significant differences among the groups in mean BSES-SF scores before (P=.26) or after (P=.68) the intervention [50].

Furthermore, 2 studies evaluated the impact of web-based breastfeeding monitoring systems among breastfeeding mothers [42,46]. Of the systems, one [42] found no significant differences in breastfeeding outcomes between the women in the intervention group (n=49) and the control group (n=57) upon discharge (P=.71). However, the women in the intervention group had better exclusive breastfeeding rates (63%, 63%, and 55%, respectively) at 1, 2, and 3 months, respectively, compared with the control group (40%, 19%, and 19%, respectively). Participants (n=26) in a study by Ahmed and Ouzzani [46] also received postdischarge breastfeeding support through web-based breastfeeding monitoring systems and reported that the system helped them to minimize breastfeeding problems [46].

A total of 3 studies reported on the usability of mobile apps without examining the effectiveness of the app on breastfeeding outcomes [55-57]. One Australian study evaluated the usability of a smartphone breastfeeding app, Breastfeeding Solutions, among rural Australian breastfeeding women [57]. The app was an interactive guide to resolve breastfeeding problems, provide functions for searching for problem solutions, and deliver timely information for mothers. The results of the study demonstrated its usability by a longer duration of breastfeeding compared with general statistics from southwest Victoria [57].

A total of 2 other studies with small sample sizes evaluated the usability and usefulness of the mobile phone apps MoomMae [56] and Milktrack [55] designed to provide support to breastfeeding women. The MoomMae [56] intervention aims to support mothers in feeling more comfortable breastfeeding in public and efficiently track their feeding and pumping logs.

A total of 2 studies used computer kiosk interventions [40,45]. The main purpose of using computer kiosks was to provide breastfeeding education and support by incorporating several educational modules such as basics of breastfeeding, benefits of breastfeeding, and coping with breastfeeding [40,45]. A pilot evaluation study [45] showed significantly greater breastfeeding knowledge in the intervention group (n=7) after they interacted with the prenatal module (P<.05) in comparison with the control group (n=8). This study showed significantly greater intention to exclusively breastfeed after women interacted with the prenatal module (P<.05) and improvement in breastfeeding confidence compared with the control group [45]. Similarly, Joshi et al [40] reported a significant improvement in breastfeeding knowledge scores (P=.03) only at week 6 of the follow-up between the control (mean 23.2, SD 3.7) and intervention (mean 25.3, SD 2.6) groups. However, no significant differences were perceived in the average change in knowledge scores between the control and intervention groups at any other follow-up time points. This study also reported a gradual increase in the breastfeeding self-efficacy scores until week 6, followed by a decrease in self-efficacy scores at 3 months (P=.46) and 6 months (P=.54). Moreover, the intervention group reported significantly higher intention to breastfeed (P=.049), and the results indicated a significant improvement in breastfeeding intention scores over a 6-month period with all study participants (P<.05).

This study presents a review of internet-based breastfeeding interventions employing e-technologies; investigates their purpose, mode of delivery, and key strategies; and systematically describes their effectiveness on breastfeeding outcomes. The main purpose of the reviewed studies focused on education and support in 3 modes of delivery: web-based interventions, mobile apps, and computer kiosks. Among the 5 key strategies, personalization and web-based discussion were the most common strategies used in the interventions. This review provides evidence that internet-based interventions employing e-technologies that provide a combination of early interactive antenatal breastfeeding education with postnatal web-based discussion support can improve breastfeeding outcomes in hospital stay [44,47,48,51,52,54] and exclusive breastfeeding rates up to 6 months [49]. These results are supported by similar face-to-face interventional studies where there is a combination of antenatal education and postnatal support [59-61].

Interventions with personalized feedback and tailored information to mothers through their support systems were found to be motivational and led to positive breastfeeding outcomes. These effective strategies point toward a need for informed, highly interactive, and tailored-designed breastfeeding interventions employing e-technologies. Consistent with other effective clinical interventions, conducting multimodal, multiphased, and interactive interventions are successful in breastfeeding practices [60,62,63]. Web-based support could be offered informally, such as peer communication through online discussion forums, or formally, such as web-based lactation consultation. The interactivity, connectivity, and two-way communication provided between mothers and lactation consultants in breastfeeding interventions may expand the opportunities for educating and engaging mothers and thus improve breastfeeding outcomes. Providing access to timely and qualified lactation consultants in similar interventions has been found to be particularly attractive to mothers experiencing some breastfeeding complications [64]. In a Cochrane review of support interventions, the findings demonstrated that women who receive any form of support are less likely to stop exclusive breastfeeding before 5 months postpartum [65].

In this review, interventions that provided monitoring and breastfeeding tracking appeared to be least effective in improving breastfeeding outcomes, which could be because of a lack of meaningful educational and supportive interactions for different breastfeeding challenges. It is important to note that women in the included studies had higher prenatal breastfeeding intention rates, which could be a confounding factor in the effectiveness of the interventions [40,45,50]. Breastfeeding women are very likely to seek out extensive evidence-based information on breastfeeding information from reliable sources. The acquired breastfeeding knowledge is presumed to have positive influences on their feeding decisions [66].

In terms of the mode of delivery, web-based platforms were the dominant and more effective mode of delivering breastfeeding internet-based interventions employing e-technologies. Several studies used web-based interventions, and the majority of them reported an increase in exclusive breastfeeding after the intervention and overall positive outcomes in their results [42,44,47,49,51,52,54]. This finding may be explained by the ability of web-based interventions to sustain their effects in terms of providing early education and continued support to participants during the intervention from a variety of care providers (eg, peers and lactation consultants) in a range of settings (eg, hospital and home). Another reason could be related to the reliability and credibility of the governmental web-based platforms [67]. Women need to make informed decisions regarding the health and well-being of their babies. Having a trusted resource would empower women to make their own choices in breastfeeding. However, many studies have reported the lack of a regulatory system to assist end users in identifying the best available web-based e-technologies [68-70].

Although the delivery of breastfeeding interventions through mobile apps is encouraging because of their relative simplicity and continuous availability [71], the mobile apps in this review focused on supporting women to access milk banks or express their breastmilk among small groups of women. Furthermore, the studies mainly aimed at assessing the usability of the apps rather than examining their effectiveness without any theoretical framework. Although a significant absence in reporting a theoretical framework was observed in the majority of the included studies, none of the interventions with mobile apps reported using any type of behavior change theoretical framework. Using theory to inform intervention development or evaluation can play a key role in breastfeeding interventions as a strong predictor of breastfeeding behavior [72].

Finally, computer kiosk interventions seemed to be effective in enhancing women’s knowledge and short-term confidence level. Creating an interactive platform coupled with continuing, accessible support can aid mothers in having specific information to suit their personal needs [45,73]. The use of conversational computers could provide affordability and portability, as observed in other health-related interventions [74,75]. However, for long-term effects, there is a need to incorporate personalized approaches and professional support into interventions to create a practical response to the personal needs of mothers. In addition, sustainability remains an issue because these types of interventions could be outdated [76], and findings cannot be generalized to other settings.

A lack of detailed reporting of rigorous intervention development, evaluation, and the content of the implementation intervention was also observed, which makes it difficult to determine the impact of interventions on breastfeeding outcomes [77]. Future studies could design intervention implementation and evaluation guided by a theory toward a comprehensive intervention development process and for a better chance of effective breastfeeding intervention.

This systematic review has as a key strength that it followed the PRISMA guidelines [35], an established methodology for ensuring transparent reporting of systematic reviews. However, there are several limitations to this study; for example, it was difficult to generalize the findings observed from several studies because of their small sample size, different timing of studies, and lack of clarity on start and end points. The substantial heterogeneity of the intervention’s outcome measures and lack of clear definitions of breastfeeding outcomes in all the included studies added more complexity and prevented meta-analysis.

Another limitation of the study is that most of the included interventions were delivered and published in developed and high-income countries with good quality health care systems [78], which may not be generalizable to women in developing countries or those with diverse socioeconomic status and cultural background [29]. Setting up breastfeeding intervention in regions with inadequate health care providers and resources should be prioritized to improve breastfeeding practices [79]. Furthermore, the approaches of the interventions were focused on changing the behavior of individuals and lacked some key aspects related to cultural differences that may influence the breastfeeding practices of women and could impede them from achieving the desired outcome. This is particularly essential as the breastfeeding decisions women make differ according to whether their own culture is supportive of breastfeeding or not and on the cultural context in which they live. Thus, findings need further investigation regarding their transferability. Applicability needs to be considered when delivering breastfeeding interventions to women in different social milieu.

The findings of the study demonstrate that internet-based e-technologies are transforming the access and delivery of breastfeeding interventions and that they have a considerable potential to assist breastfeeding mothers when seeking support and advice about breastfeeding. The results show that web-based interventions that provide a combination of education and ongoing support are the best models of interventions employing e-technologies to support long-term breastfeeding outcomes. In addition, the review reveals that the two dominant effective strategies are personalization, and online discussion forums form credible sources. Further studies need to explore the usability and effectiveness of interventions employing e-technologies that have theory-based systems designs that could incorporate encouragement and discussion opportunities from credible social and professional sources. The sociocultural needs of women also need to be integrated into these technologies to provide culturally tailored breastfeeding support.