The quality of maternal nutrition and infant feeding counselling during antenatal care in South Asia

Abstract Antenatal care (ANC) provides a platform to counsel pregnant women on maternal nutrition and to prepare the mother to breastfeed. Recent reviews suggest that gaps in the coverage and quality of counselling during pregnancy may partly explain why services do not consistently translate to improved behavioural outcomes in South Asia. This scoping literature review collates evidence on the coverage and quality of counselling on maternal nutrition and infant feeding during ANC in five South Asian countries and the effectiveness of approaches to improve the quality of counselling. Coverage data were extracted from the most recent national surveys, and a scoping review of peer‐reviewed and grey literature (1990–2019) was conducted. Only Afghanistan and Pakistan have survey data on the coverage of counselling on both maternal nutrition and breastfeeding, nine studies described the quality of counselling and three studies assessed the effectiveness of interventions to improve the quality of services. This limited body of evidence suggests that inequalities in access to services, gaps in capacity building opportunities for frontline workers and the short duration and frequency of counselling contracts constrain quality, while the format, duration, frequency and content of health worker training, together with supportive supervision, are probable approaches to improve quality. Greater attention is needed to integrate indicators into monitoring and supervision mechanisms, periodic surveys and programme evaluations to assess the status of and track progress in improving quality and to build accountability for quality counselling, while research is needed to understand how best to assess and strengthen quality in specific settings.

women on healthy eating and keeping physically active, educate pregnant women in undernourished populations on increasing energy and protein intake and encourage women at risk of micronutrient deficiencies to take micronutrient supplements (WHO, 2016). In addition, the recommendations highlight the importance of the quality of care for a positive pregnancy experience. ANC also provides an opportunity to prepare the mother to initiate breastfeeding immediately after delivery and exclusively breastfeed for 6 months (WHO, 2018).
There has been a rapid expansion in the number of countries that offer nutrition counselling to pregnant women as a standard component of ANC (UNICEF, 2019). These services are valued in South Asia where maternal undernutrition is one of the main reasons underlying the region's high prevalence of child wasting (15%) and stunting (34%) (Torlesse & Aguayo, 2018;UNICEF, WHO, and World Bank Group, 2019). An estimated one in two women are anaemic, one in 10 South Asia women have a low stature (<145 cm), one in five have a low body mass index (<18.5 kg m 2 ) and one in four infants are born with a low birthweight (<2.5 kg) (Goudet, Murira, Torlesse, Hatchard, & Busch-Hallen, 2018;Stevens et al., 2013;UNICEF & WHO, 2019).
South Asia is performing relatively better than other regions on breastfeeding, but even so, 58% of newborns are not breastfeed immediately after birth and 41% of infants aged less than 6 months are not exclusively breastfed (UNICEF, 2016).
Recent reviews have suggested that gaps in the coverage and quality of nutrition counselling during pregnancy may partly explain why counselling services do not consistently translate to improved dietary intake, consumption of micronutrient supplements and infant feeding practices in South Asian countries (Benedict, Craig, Torlesse, & Stoltzfus, 2018;Goudet et al., 2018). To understand the magnitude of these gaps and how to address them, this study reviews data and literature from South Asia on the coverage and quality of counselling services on maternal nutrition and infant feeding during ANC and on the effectiveness of programmes, interventions and approaches to improve the quality of counselling.

| METHODS
This study examines evidence from the five largest South Asian countries (Afghanistan, Bangladesh, India, Nepal and Pakistan) on the (i) coverage of counselling on maternal nutrition and infant feeding practices given to pregnant women during ANC; (ii) quality of counselling on maternal nutrition and infant feeding practices given to pregnant women during ANC; and (iii) effectiveness of programmes, interventions and approaches to improve counselling on maternal nutrition and infant feeding during ANC. The maternal nutrition practices include meal frequency, quantity of food intake, consumption of nutritious/diverse foods, adherence to intake of iron, folic acid and multiple micronutrient supplements during pregnancy. The infant feeding practices include the early initiation of breastfeeding (EIBF) immediately after delivery, avoidance of prelacteal feeds and exclusive breastfeeding (EBF) for 6 months.

| Conceptual framework
We developed a conceptual framework that describes factors associated with the quality of counselling services at the levels of the health facility, community, household and individual ( Figure S1). Based on existing literature that describes the features of ANC that influence quality of counselling (Bitton et al., 2017;Donabedian, 1980;Hulton, Matthews, & Stones, 2000), we identified five key factors: accessibility to services, resource availability, environmental readiness, service provider readiness and interactions between service provider and client. A set of indicators, derived from service provider guidelines, training manuals and job aides on counselling (Alive & Thrive, 2013;Gavin et al., 2014;IYCN, 2012;UNICEF, 2011;WHO, 2012WHO, , 2015, were defined for each of the five factors (Table 1).

Key messages
• Despite significant investments in developing the capacity of frontline workers to provide counselling on maternal nutrition and infant feeding during antenatal care (ANC), there are evidence gaps in the quality of counselling and effectiveness of approaches to improve quality in South Asia.
• Inequalities in access to services, gaps in capacity building opportunities for frontline workers and the short duration and frequency of counselling contacts constrain the quality of counselling.
• The format, duration, frequency and content of health worker training and supportive supervision are probable approaches to improve quality.
• Greater attention is needed to integrate indicators of quality counselling into monitoring and supervision mechanisms, periodic surveys and programme evaluations, while research is needed to understand how best to assess and strengthen quality in specific settings.

| Study framework
A scoping methodology framework proposed by Arksey and O'Malley (2005) was used. It involved a five-stage review process: (i) identification of the research objectives, (ii) identification of relevant studies, (iii) section of studies, (iv) mapping of the data and (v) organization, summary and reporting of the results.

| Literature search
A PICOT search strategy, including the elements of population, intervention, comparison, outcomes and time, was used. PubMed was searched to identify research articles published between January 1990 and May 2019. Search terms were applied with various Boolean operators for country location, target group, ANC and counselling, maternal nutrition and infant feeding and date of publication ( Figure S2). In addition, grey literature published between January 1990 and August 2019 on programme evaluations was sourced from experts and organizational websites, including the WHO Library Database, Human Resources for Health Global Resource Center and the International Initiative for Impact Evaluation.

| Study selection
Eligibility criteria included the study location (Afghanistan, Bangladesh, India, Nepal andPakistan), year published (1990-2019), at least one measured indicator of the quality of counselling, study design, English language and full-text availability. Duplicate citations were removed, and titles and abstracts were screened to identify relevant studies. Two researchers independently screened the full texts of T A B L E 1 Indicators of the quality of counselling on maternal nutrition and infant feeding during antenatal care (E) Service provider and client interactions Quality of interactions between client and health worker in the delivery of counselling services E1: Heath worker establishes and maintains a trusting environment and builds rapport with client (e.g., greets client, uses verbal and non-verbal responses that show interest, speaks in a respectful/polite manner and takes time) E2: Health worker uses listening and learning skills to assess the client's needs and personalizes the discussions accordingly E3: Health worker provides information/advice on optimal practices and explains why they are important in a way that is easily understood and retained by the client E4: Health worker works interactively with the client to address concerns and questions, give practice support, build confidence and establish a plan to change behaviours E5: Health worker confirms the client's understanding E6: Health worker involves other family members (if present) E7: Client satisfaction with counselling services Abbreviation: ANC, antenatal care. all potentially relevant articles to assess eligibility. Any discrepancies between reviewers were resolved through discussion and consultation with a third researcher.

| Synthesis of data
For studies that examine the quality of counselling on maternal nutrition and infant feeding during ANC, data were synthesized by source and country; study setting, design and sample; maternal nutrition and infant feeding practices examined; and reported results for each indicator of the counselling quality. For studies that examine the effectiveness of programmes, interventions and approaches to improve to improve counselling on maternal nutrition and infant feeding during ANC, data were synthesized by source and country; study design, subjects, intervention group and comparison group; maternal nutrition and infant feeding practices examined; and reported results for each indicator of counselling quality. Effectiveness to improve counselling was defined as measured improvements in any of the indicators of the quality of counselling in the intervention group compared with the comparison group.

| Ethical considerations
Ethical approval was not required for this study as it used data from existing reports and articles.

| RESULTS
After excluding duplicates, 924 unique articles were screened, and 12 studies were selected ( Figure S3), including 10 peer-reviewed articles and two grey literature reports. Findings are reported by evidence on the quality of counselling on maternal nutrition and infant feeding (nine studies) and effectiveness to improve the quality of counselling (three studies).

| Coverage of counselling services during ANC
Data extracted from national survey reports on the coverage of maternal nutrition and infant feeding counselling during ANC are provided in Table 2 together with data on the coverage of four or more ANC visits during pregnancy. All five countries have survey data on ANC coverage; however, only two countries have data on the coverage of counselling on maternal dietary intake (Afghanistan and Pakistan), and three countries have data on the coverage of breastfeeding counselling (Afghanistan, India and Pakistan).
Country comparisons of counselling coverage are only possible for Afghanistan and Pakistan, which have data on the same indicators.
The proportion of women who received counselling on dietary intake was almost 50% higher in Pakistan than Afghanistan (70% vs. 48%) whereas counselling on EBF was more than three times higher in Pakistan than Afghanistan (54% vs. 16%). A possible explanation for these differences is that Pakistan has a stronger network of health workers to provide counselling to women during pregnancy, including its network of community-based lady health workers (LHWs). In both Afghanistan and Pakistan, the coverage of counselling on EBF is lower than maternal dietary intake (16% vs. 48% in Afghanistan and 54% vs. 70% in Pakistan). This indicates that the primary health care services in both countries are giving relatively greater attention to promoting a healthy diet during pregnancy than to preparing pregnant women to exclusively breastfeed.
The coverage of four or more ANC visits during pregnancy was similar to the coverage of counselling on EBF in Afghanistan (18% vs. 16%) and Pakistan (51% vs. 52%). However, there were large differences in the coverage of four or more ANC visits and counselling on dietary intake in Afghanistan (18% vs. 48%) and Pakistan (51% vs. 70%) and in the coverage of four or more ANC visits and breastfeeding counselling in India (51% vs. 80%). This suggests that these countries are utilizing other service delivery platforms in addition to ANC services to provide counselling on dietary intake or breastfeeding to women during pregnancy, such as community-based workers.
3.2 | Quality of counselling on maternal nutrition and infant feeding

| Characteristics of included studies
The nine studies included four that were conducted in India    (Huda et al., 2018) and Nepal (McPherson et al., 2010), and no eligible studies were conducted in Afghanistan. All studies examined both maternal nutrition and infant feeding practices except Dhandapany et al. (2008), which reported on infant feeding only. The study designs included four cross-sectional studies (Avula et al., 2015;Dhandapany et al., 2008;Mahar et al., 2012;Singh et al., 2012), one prospective cohort study (Pricilla et al., 2017), two qualitative studies (Dykes et al., 2012;McPherson et al., 2010) and two mixed methods studies (Huda et al., 2018;Majrooh et al., 2014). The studies collectively described four out of five factors of quality and 13 out of 24 indicators; no studies examined indicators of environmental readiness. A summary of the included studies is provided in Table 3, and Table S1 lists the indicators on quality of counselling for each of the included studies.

| Accessibility
Accessibility is defined as geographic, financial, sociocultural and temporal access to counselling services and the timeliness, frequency and duration of counselling. Three studies examined one or more indicators of accessibility (Huda et al., 2018;Mahar et al., 2012;Singh et al., 2012). A study in India found that pregnant women were more likely to receive advice on maternal nutrition and breastfeeding if they belonged to wealthier households and utilized higher level facilities (Singh et al., 2012), whereas in Pakistan, coverage of counselling services was higher among clients who accessed health care from private facilities than public facilities (Mahar et al., 2012), indicating socioeconomic barriers among poorer households. Mahar et al. (2012) reported that ANC counselling interactions in public and private hospitals in urban Pakistan lasted only 3 and 5 min, respectively. A pilot study in Bangladesh assessed the feasibility, acceptability and perceived appropriateness of mobile phone counselling of pregnant women and mothers on maternal nutrition and infant feeding (Huda et al., 2018). It found that 22% of women were not satisfied with the frequency of biweekly counselling and that two thirds of women missed at least one counselling call because of household responsibilities or difficulties in charging their mobile phones.

| Resource availability
Resource availability is defined as the sociocultural appropriateness of counselling messages and availability of hardware and skilled health workers. Four studies examined resource availability, including client satisfaction with the types of messages, and job aids and approaches used by health workers and reported positive feedback (Avula et al., 2015;Dykes et al., 2012;Huda et al., 2018;McPherson et al., 2010). In Nepal, pregnant women and other family members appreciated a pictorial booklet that was used by health workers to counsel women and found that the information, contents and length were appropriate and the messages understood (McPherson et al., 2010). In Pakistan, women were pleased to participate in cookery demonstrations that were organized by LHWs to inform and counsel women on dietary quality (Dykes et al., 2012). In Bangladesh, women reported that they were satisfied with direct counselling through mobile phones (Huda et al., 2018), although this was their subjective perception and does not mean that the messages or approaches used to convey the messages were appropriate. A study in India examined the knowledge of different cadres of frontline workers, Anganwadi workers (AWW), accredited social health activists (ASHA) and auxiliary nurse midwives (ANMs) (Avula et al., 2015). It found that the proportion of frontline workers that had accurate knowledge on EBF (99-100%) was greater than on aspects of maternal nutrition, including the importance of a diverse diet (56-66%), additional food intake during pregnancy (47-53%) and iron and folic acid (IFA) supplements during pregnancy (69-56%

| Provider readiness
Provider readiness is defined as appropriately skilled, motivated and

| Service provider and client interactions
All nine studies examined one or more indicators on the quality of interactions between service provider and client in the delivery of counselling services. Six studies reported on the proportion of clients who received information/advice on optimal practices from health workers; this proportion varied widely according to the subject of the counselling, study setting and the type of health facility.
Pregnant women were less likely to receive information or counselling on breastfeeding than on maternal nutrition in studies that made this comparison in Pakistan (Mahar et al., 2012;Majrooh et al., 2014) and urban India (Pricilla et al., 2017) but not in rural India (Singh et al., 2012). Studies in Pakistan found that the counselling coverage was greater in private health facilities than public facilities (Mahar et al., 2012) and higher in greater level facilities than lower level facilities (Pricilla et al., 2017). Other indicators of the interactions between service provider and client were T A B L E 3 Summary of studies examining the quality of counselling on maternal nutrition and infant feeding during antenatal care

T A B L E 3 (Continued)
Source and country Study setting, design and sample

Maternal nutrition and infant feeding practices examined Results
Sample: 275 women who were pregnant or had a birth in the last 6 months • Most women understood the content of the counselling and felt the information provided was very important and beneficial for both mother and child (E7). • Two thirds missed at least one counselling call because of household responsibilities or difficulties in charging the mobile phones (A5).

| Effectiveness to improve the quality of counselling
The three included articles (Table 4) relate to studies conducted in India (Baqui et al., 2006) and Bangladesh (Nguyen et al., 2017(Nguyen et al., , 2018; no eligible studies were conducted in Afghanistan, Nepal or Pakistan.
The studies examined the impact of programme interventions on indicators of accessibility (Nguyen et al., 2018), resource availability (Baqui et al., 2006;Nguyen et al., 2018), provider readiness (Nguyen et al., 2018) and service provider and client interactions (Baqui et al., 2006;Nguyen et al., 2017Nguyen et al., , 2018; no studies reported on indicators of environmental readiness. Table S2 lists the indicators on quality of counselling for each of the included studies. In Bangladesh, a cluster randomized control trial evaluated the effect of providing nutrition-focused ANC compared with standard ANC on maternal dietary diversity, micronutrient supplement intake and early breastfeeding practices (Nguyen et al., 2017(Nguyen et al., , 2018 knowledge of maternal nutrition and infant feeding at endline (Nguyen et al., 2018). The increase between baseline and endline in the proportion of women receiving information was significantly greater in the intervention group than the comparison group for the receipt of information on eating at least five food groups and avoidance of prelacteal foods but not for eating additional food, taking IFA supplements or taking calcium supplements, EIBF or EBF (Nguyen et al., 2017). Overall, the increase between baseline and endline in the number of messages on maternal nutrition and breastfeeding received by women was greater in the intervention area than the comparison group (Nguyen et al., 2018). Differences in coverage of service delivery and counselling quality explained a large portion of the variation among villages for consumption of IFA and calcium supplements, whereas differences in the quality of counselling services explained 60% of the programme's impact on women's dietary diversity during pregnancy (Nguyen et al., 2018). We developed a conceptual framework and set of 24 indicators

| DISCUSSION
to characterize the quality of counselling across five key factors (accessibility to services, resource availability, environmental readiness, service provider readiness and interactions between service provider and client). We used this framework to gather evidence on the quality of counselling on maternal nutrition and breastfeeding and the We found only three studies that examined approaches to improve the quality of counselling during ANC, and all included a very small number of indicators (range one to five). The most comprehensive study (Nguyen et al., 2018)  These findings show that studies which examine counselling on maternal nutrition and infant feeding during ANC in South Asia rarely focus the quality of counselling and none do so comprehensively.
These evidence gaps need greater attention to ensure that poor quality counselling services do not continue unchecked and to guide the design of these services so that they are more impactful on maternal nutrition and infant feeding outcomes. Our framework of indicators on the quality of counselling may help to inform the development of metrics to measure the multidimensional features of the quality of counselling. However, more research is needed to identify indicators that reflect the most impactful aspects of quality in specific settings and to understand the pathways through which the quality of counselling on maternal nutrition and infant feeding can be strengthened. ANC programmes should adapt and integrate quality of counselling indicators into programme monitoring, supportive supervision mechanisms and programme evaluations. A challenge that remains for countries is to find low-cost, efficient and real-time methods to assess the quality of counselling. There is also need for advocacy and political will to address disparities in access to quality services to ensure equity in the provision of care.
In line with the scoping review methodology, our review of literature did not grade the quality of evidence. To provide the highest level of evidence on intervention effectiveness, we limited our eligibility criteria to include only randomized controlled trials and quasiexperimental studies as these designs provide the most precise estimates of the likely effects of an intervention with limited risk of bias. The scarcity of data and information on the coverage, quality and effectiveness of approaches to improve the quality of counselling during ANC is both a limitation and key finding of this study.

| CONCLUSION
ANC provides an important platform to inform and counsel pregnant women and their family members. Our review highlights the shortage of data and evidence on the coverage and quality of counselling on maternal nutrition and infant feeding offered to women during ANC in South Asia and effectiveness of programmes, interventions and approaches to improve quality. Greater attention is needed to integrate appropriate indicators into monitoring and supervision mechanisms, periodic surveys and programme evaluations in order to assess the status and track progress in improving quality, build accountability for quality counselling, and to conduct research to understand what works to improve the quality of counselling.