mHealth communication to strengthen postnatal care in rural areas: a systematic review

The IMBP (Fig. 1) guided this review by identifying the determinants of behaviour change that might affect PNC uptake. The IMBP accounts for uptake of any behaviour in any given population [22]. It is a powerful model for predicting behaviour across a variety of fields, and has been used to boost or reduce certain behaviours with public health implications [23‐26]. According to the IMBP, only a small number of determinants need to be considered to predict, change, or reinforce a particular behaviour in any given population [22]. The model proposes that any given behaviour is most likely to occur if a person has a strong intention to perform the behaviour, if a person has the necessary skills and abilities required to perform the behaviour, and if no environmental constraints prevent behavioural performance [21]. Intentions are influenced by the attitude toward performing the behaviour, perceived norms concerning performing the behaviour, and the person’s self-efficacy with respect to performing the behaviour [21, 22]. Attitudes, perceived norms, and self-efficacy are global perceptions that represent a variety of specific beliefs about the particular behaviour [22]. They are all functions of specific beliefs that influence intentions in the model. In this review, the uptake of PNC was determined by highlighting the various outcomes aligned with the primary determinants of the behaviour in the model, i.e., intention, skills, and environmental constraints.

After a rapid appraisal, search words were generated from three major PICO elements (population, intervention, outcome) and a group of synonyms and related concepts was created and combined using OR and AND operators, to ensure comprehensive search (Table 1). The following databases were used on the EBSCOHost interface: Academic Search Complete, Africa-Wide Information (containing data from the National Research Foundation database of current and completed research projects), CAB Abstracts, CINAHL with Full Text, Communication & Mass Media Complete, ERIC, Health Source: Consumer Edition, Health Source: Nursing/Academic Edition, MasterFILE Premier, MEDLINE with Full Text (containing the same data as PubMed), PsycINFO, SocINDEX with Full Text and SPORTDiscus with Full Text. In addition, checking the reference lists of the studies that were included added three further studies. The authors were contacted in the case of two studies, to confirm the settings of the studies. Before the final analysis, the search was rerun, but no other studies were retrieved for inclusion.

The inclusion criteria that were applied included that the literature had to be written in English, or that studies written in other languages had to have abstracts written in English, mHealth communication had to be between healthcare personnel and mothers, or involved community members, mHealth communication was aimed at improving health outcomes of either a mother or child, mHealth communication was with mothers or community members from semi-urban and rural areas, and that literature had to have been published between 1 January 2008 and 31 August 2018. Exclusion criteria were that mHealth communication focused on peer support amongst mothers, mHealth communication was with mothers or community members whilst mothers were still hospitalised, literature involved evaluation of mHealth programmes and health systems, and literature represented research briefs, research news, commentaries, editorials or proposals. Three reviewers (FM, MR and AF) independently screened titles, abstracts, and full-text articles to decide whether an article was relevant. Any discrepancies among reviewers were resolved by obtaining consensus. Only studies published in peer-reviewed journals were included. The results of study screening and selection are illustrated in Fig. 2.

A standardised, pre-piloted form was used to extract data from the studies included, to assess study quality and to synthesise evidence. Extracted information included bibliographical details, methodology, outcomes and intervention, critical appraisal rating and study findings (see Additional file 1.).

Three reviewers conducted the critical appraisal of the study independently, to ensure the validity and to increase the reliability of the process. A critical appraisal checklist that was suitable for each research design of studies that were appraised was selected. For randomised, controlled trials and qualitative and cross-sectional studies, checklists developed by the Critical Appraisal Skills Programme (CASP) [30] were used to assess the quality of each study. The quality of non-randomised studies was examined using the checklist for quasi-experimental studies (non-randomised experimental studies) of the Joanna Briggs Institute [31]. Although the scores differed according to the research design used, all these checklist tools focused on the following aspects: appropriateness of methods used, clarity of focus, the recruitment process, the accuracy of measures used, data collection, presentation and analysis, clarity in the statement of the findings, and appropriateness of context. Higher scores indicated higher quality. A data extract sheet provides the quality assessment of all studies included. Grades were assigned to show the level of the evidence according to guidelines of the American Dietetic Association [28] (Additional file 1).

Seven types of mHealth communication were used to strengthen PNC in rural areas; these were phone calls, one-way messaging, interactive messaging, audiovisual material and videos, voice messages, and combined messaging and phone calls. Mobile phone calls were used by village health team members to consult healthcare professionals on matters regarding PNC [32, 33]. These calls were also used by healthcare practitioners (HCPs) to remind mothers to attend PNC clinics and to inform them of their laboratory results [7, 13, 34, 35]. One-way mobile phone messaging was used by HCPs to remind mothers of clinic appointments, communicate test results from the clinic and to give them information on PNC [7, 13, 34‐39]. Interactive mobile phone messaging between HCPs and mothers was used to discuss breastfeeding practices and remind mothers to attend PNC clinics [40]. Mobile phone audio visual application and educational videos was sent to mothers by HCPs to illustrate various PNC practices [37]. A combination of one-way mobile phone messaging and phone calls was used by HCPs to remind mothers about their clinic appointment dates [13, 35]. Voice messages were delivered to mothers by HCPs to inform them of neonatal care [7]. Among all these types of mHealth communication, one-way mobile phone messaging was commonly used to influence the three primary determinants of PNC uptake, i.e., intention, skills and environmental constraints.

Intentions, as described in the IMBP, are influenced by attitudes, perceived norms and self-efficacy, all of which are functions of specific beliefs described in the model. Though not all articles addressed the three determinants of intention, the themes emerging from the results that were found to change or maintain PNC outcomes targeting attitude, perceived norms, and self-efficacy, were as follows.

Though it was not reported by all studies that were reviewed, several skills for PNC were strengthened. mHealth interventions such as reminders to mothers on skills previously taught were used. In Uganda, mothers gained skills in cord care, thermal care and breastfeeding immediately after birth when health professionals explained the skills through mobile phone consultations [32, 33]. When text messages were sent to mothers in Guatemala, providing information regarding breastfeeding practices, mothers reported that they practiced exclusive breastfeeding successfully [38]. In Bangladesh, when text messages were used to give information on neonatal care, mothers demonstrated skills related to breastfeeding immediately after birth and bathed newborns after 3 days [7].

mHealth communication addressed possible environmental constraints, such as inaccessibility, unavailability and affordability, which act as barriers to the uptake of PNC services in rural areas, although not all the articles explored this aspect. For instance, mHealth communication addressed the inaccessibility of PNC in rural areas of Uganda, and helped mothers to access information in a timely fashion, and prompted maternal and newborn care [32, 33]. In rural areas of Ethiopia, mHealth communication was used to track mothers by notifying them of PNC services; this provided information on the availability of PNC services [13, 36]. In Kenya, mothers reported that receiving mHealth communication facilitated PNC services [40]. In India, text messages to mothers facilitated more significant contacts without requiring mothers to travel to a health facility, and this enhanced the availability of PNC services [37]. With the use of mobile phone messaging to track newborns due for vaccination, there was timely vaccination of children in rural areas of Bangladesh – the messaging facilitated PNC [39]. In Zambia, text messaging was used to encourage mothers to return to the clinic and to give test results, leading to early diagnosis and enrolment for care [35].