Subhash Pokhrel, PhD, MSc
Expected Key Learning Outcomes
The economic considerations for mothers
The impact of breastfeeding on health systems
Analysis of how far breastfeeding promotion and support interventions offer good value for money
Effective ways to invest in breastfeeding promotion
Outline of a business case made for breastfeeding promotion and support in the absence of robust economic evaluation
Human milk has several implications. Depending on what perspective one chooses, the use of alternatives to human milk for feeding infants has attracted enormous debate in the past. This chapter surveys contemporary literature around the economic value of breastfeeding and presents an example analysis to show how a business case for breastfeeding support could be made.
What economic value may human milk have? This question has featured in both academic and political debates for a long time. For some, human milk is protective against certain disease conditions and therefore it can provide substantial economic benefits. Breastfeeding is beneficial not only to the health and wellbeing of a child and their mother, but it also generates substantial cost savings to the national health services. Health services would have to treat fewer infant, childhood, and maternal diseases with increasing breastfeeding prevalence , . In addition, some authors argue that women who choose to breastfeed actually produce and supply breast milk and therefore contribute significantly to the national economy , . When costs of implementing breastfeeding support policies are considered, society is more likely to get a positive return on investment (ROI) from breastfeeding .
The other side of the argument positions breastfeeding as a costly enterprise to women because, if they chose to breastfeed their babies, they would be required to incur substantial private costs to enable milk expression . Like formula feeding, breastfeeding is also associated with private costs. In addition, breastfeeding may have implications for earnings and productivity of working women, potentially requiring a longer maternity leave, working part time, or missing opportunities for promotion , , . It also takes a substantial amount of a mother’s time to breastfeed her child . Therefore, both the private costs and the forgone opportunities women may experience by choosing to breastfeed can be considerable.
Underneath these individual benefit-cost arguments rests a question that has probably the most profound implications for any breastfeeding support policy. Can a health system ask women to initiate breastfeeding, and breastfeed for longer and exclusively, particularly when we, as a society, recognise that it is up to women themselves to make those explicit choices? What determines a woman’s decision to initiate (or cease) breastfeeding and how those factors relate to the thinking of a healthcare system appear to be central to this question . Therefore, it is important to consider whether breastfeeding is in fact an “economic” decision for women as well as for other stakeholders.
Breastfeeding is an economic decision but its nature varies according to the perspective taken. Working women may consider the consequences of breastfeeding (i.e. opportunities foregone and/ or monetary costs of breastfeeding relative to formula feeding) when deciding whether to breastfeed their babies, while employers and health systems may consider the need to support breastfeeding women through maternity pay and by creating baby-friendly workplaces and hospitals .
Maternal employment appears to be negatively associated with breastfeeding initiation and duration . This is particularly relevant since exclusively milk-feeding mothers would have to spend much more time every week on feeding their baby compared with other mothers . Understanding incentives or disincentives facing women that may influence their choices regarding initiation and duration of breastfeeding (any or exclusive) is therefore critically important.
Economic theories help us to understand what those incentives and disincentives might be, and how these may determine a women’s choice to breastfeed or formula feed and for how long. One such theory is that of individual net-benefit (utility) maximisation; in this case, individuals are assumed to make a choice (e.g., to initiate breastfeeding) that is perceived to benefit them and adhere to their decision until the benefits outweigh the costs . In this framework, any factor that is perceived as a barrier or disincentive by a mother, e.g., money, time, and negative feedback from friends or family, is a cost. Likewise, any factor that is perceived as a facilitator or incentive, e.g., money saved by not buying formula, better health of the child, bonding with their child, and access to breastfeeding support, is a benefit. The model also assumes that the incentives and disincentives to breastfeeding may change over time.
▶Fig. 11.1 depicts this notion of the decision making process proposed by Racine and colleagues . In this construct, the decision is an economic choice; the postpartum women weigh the benefits of breastfeeding against the costs of continuing or discontinuing breastfeeding. Some factors that are incentives for health systems (e.g., the health benefits of breastfeeding for infants/ children and for mothers) are also incentives for women. Although provision of breastfeeding support requires health systems to cover costs, this support is also an incentive for women encouraging them to choose to breastfeed.
Racine and colleagues implemented this model in a sample of 1,595 low-income families in the US, and found that the decision to discontinue breastfeeding was significantly associated with a number of disincentives: the Supplemental Nutrition Program for Infants, Women, and Children (WIC) participation at 2–4 months, mothers returning to work for 20–40h per week, mothers not attending a postpartum doctor’s visit, fathers not being in the home, a smoker in the household, no receipt of breastfeeding instruction at the paediatric office, the doctors not encouraging breastfeeding, and the mother experiencing depressive symptoms . The main implication of this finding is that any breastfeeding promotion programme will need to address disincentives associated with breastfeeding cessation. Understanding the economics of breastfeeding decisions is therefore helpful for policymakers.
Although the Net-Benefit (utility) Maximisation model is a valuable way to identify determinants of breastfeeding decisions that women may take (i.e. initiate, continue or discontinue; any or exclusive breastfeeding), the decision itself is complex. The decision not to breastfeed infants is also the decision to formula feed (i.e. use infant food or breast-milk substitutes). Infant food is often allocated by markets. If we were to rely on markets to allocate resources efficiently, consumers (i.e. postpartum women) would have to make informed (rational) choices. These choices require that postpartum women themselves are responsible for the full costs and benefits of their infant food purchase decisions. Much cultural knowledge of health risks of formula feeding is based on inaccurate or biased information and this, coupled with commercial vested interests, may not enable women to make an informed (rational) decision . It is known that not choosing to breastfeed leads to a decrement in infant and maternal health, thereby costing health systems millions of dollars , . Those health systems costs are usually borne by taxpayers (as in the British National Health Service [NHS]) or others (e.g., social/private insurance) and not by women who make consumption decisions (purchase of breast-milk substitutes), a phenomenon known as externalities (an attribute of market failure). This is particularly important as the extent to which the women who choose to use breast-milk substitutes are willing to bear this cost is less understood. In this instance, the market price of artificial infant food becomes much lower than its true economic costs to women who want to purchase it, making breastfeeding a less attractive option .
Another linked issue around the use of markets to allocate infant food efficiently is that of agency. In the case of infant feeding, one could argue that the infants are the actual consumers and not their mothers. Mothers make decisions on behalf of their infants – a classic principal-agent problem in economics . While agents (mothers) make decisions on behalf of their principals (infants), it is likely that agents are acting in their best interests rather than in the best interests of their principals. It is argued elsewhere that given the difficulty to accommodate the needs of the breastfeeding mother in the context of institutional frameworks, it is likely that the interests of the mother and the infant may not always align .
Whether to breastfeed is thus a complex decision that postpartum women have to make by weighing the incentives (benefits) and disincentives (costs) of breastfeeding relative to that of formula feeding. Breastfeeding is not a binary choice; it is rather a set of complex choices around initiation, duration, and exclusivity. What women decide to do on infant feeding may have far reaching implications beyond their families.
One of the economic disincentives (costs) associated with women’s infant feeding decisions is private costs . Despite breast milk being considered as the best form of nourishment for infants and usually in sufficient supply for the first few months of life, it is not free for women who choose to breastfeed. There are private costs associated with breastfeeding. Two types of private costs are prevalent: monetary costs and time costs.
In a study conducted in Liverpool, England 149 women between the age of 18 and 43 were asked to report the purchases associated with their infant feeding practices (mean age of infants: 13 weeks) . The study identified a number of equipment items needed to enable women to breastfeed. This included nursing bras, nightshirts, breast pads, antiseptic nipple spray, breast cream, breast shells, nipple shields, breast pump, breast-milk storage bottles, breast-milk freezer bags, steriliser, and support pillow. Two separate models (high costs and low costs) were used to estimate the average costs of purchasing the equipment. A set of breastfeeding equipment was purchased for £34.60 per week (high-cost model) or £2.40 per week (low-cost model). Likewise, formula feeding mothers had bought bottles, teats, steam steriliser, formula, bottle warmer, bottle carrier, powder dispenser and bottle/teat brushes. A set of formula-feeding equipment including the food was purchased for £31.43 per week (high cost model) or £6.30 per week (low-cost model).
On average, breastfeeding cost women £11.58 per week compared with £9.60 per week for formula-feeding (2002–2003 prices). However, the study found that women, particularly the first time mothers, in both groups, ‘spent money on items that were not needed or used only once or twice’ . Higher spending was characterised by education, socio-economic status and age. Although women included in the study spent more per week on breastfeeding compared with formula feeding, provision of better support (information) could have led the women to avoid purchasing items that were unnecessary or to go for cheaper alternatives where available.
Depending on the healthcare context, there may be other forms of private monetary costs associated with infant feeding. Frick and colleagues identified food for the mother herself and medical care use for herself or her child (in non-NHS/insurance settings) as potential private costs required to enable mothers to breastfeed .
The choice of infant feeding is also associated with time costs. In particular, ‘exclusive breastfeeding is time intensive, which is economically costly to women’ . In an Australian survey (2005–2006), 139 new mothers were asked to report their average weekly time spent on feeding (milk or solids), feeds preparation, and the total of the two. Mothers who were exclusively breastfeeding spent on average 7 hours extra per week on milk feeding their infants compared with other mothers. This difference was statistically significant and implied that premature weaning was probable ‘for women who are time-stressed, lack household help from family, or cannot afford paid help’ .
The time costs of breastfeeding have wider implications. As exclusive breastfeeding is associated with substantial time commitment, working women in particular may have to compromise on their earnings and productivity as choosing to breastfeed means choosing to take longer maternity leave or work part time and potentially miss promotion opportunities , , . For others, the time spent on breastfeeding could have other usage . The opportunities forgone by choosing to breastfeed may therefore be considerable. Breastfeeding promotion policies must therefore subsidise/share these costs through provision of various services, e.g., childcare, help with housework, prolonged maternity leave, and if mothers decide to return to work, the provision of breastfeeding breaks at the workplace .
As seen above, breastfeeding is an economic choice that women make. Therefore, it is important to support women to breastfeed for as long as they choose to. It appears that most women who stop breastfeeding don’t want to and often consider getting the help and support that would keep them breastfeeding for longer and exclusively . Supporting women who choose to breastfeed would therefore help align interests of the mother, the baby, and the health services. As breastfeeding (exclusive and/or longer duration) becomes more common as the result of this support, this will lead to wider economic benefits too , .
Central to any policy debate around breastfeeding should be the recognition that to breastfeed is an exclusive choice of a new mother. Any breastfeeding support policy therefore must acknowledge that new mothers who have chosen to breastfeed are well informed, well trained and well supported for however long they choose to breastfeed (exclusively or partially). It is possible that a breastfeeding promotion policy may help new mothers initiate breastfeeding; support thereafter enabling women to breastfeed for longer is what generates health and economic benefits to the mother, the baby, and the health services.
Having established the notion that supporting women who choose to breastfeed makes an economic sense, it is important to look at the evidence base to see what health benefits breastfeeding may offer to mothers and their babies. How would the positive health effects of breastfeeding translate into economic benefits both to national health systems and to wider society? At the micro level, do breastfeeding support interventions offer good value for money?
Breastfeeding has been found to be protective against a number of health conditions in infants and children. However, the strength of evidence varies by health conditions. For gastrointestinal infections, lower tract respiratory infections and acute otitis media in infants, and necrotising enterocolitis in pre-term babies, convincing evidence exist to suggest that breastfeeding prevents the incidence of those conditions . ▶Table 11.1 provides a summary of this evidence
A previous review of the benefits of breastfeeding identified evidence according to three categories: convincing (significant relationship established by systematic reviews/meta analyses), probable (association found in several studies but more evidence is needed), and possible (association found in few studies of less good quality) . A large number of disease conditions where breastfeeding could be protective were identified according to this evidence hierarchy (▶Fig. 11.2).
More recent systematic/evidence reviews have corroborated these findings , . As more studies are conducted, clearer pictures of the association between breastfeeding and these health outcomes will emerge. Of particular note are the three conditions (cognitive outcomes, sudden infant death syndrome, and childhood obesity) where studies increasingly indicate a negative correlation between breastfeeding and the incidence of these outcomes .
Mean value (95% CI)
Lower respiratory tract infection
   
Acute otitis media
Maternal breast cancer
CI = confidence interval, GP = general practitioner * Odds ratio (OR) or relative risk (RR) measuring how likely a disease condition listed above is in breastfeeding group compared to non-breastfeeding group. Mean values are reported in bold. A ratio of less than 1.0 indicates breastfeeding is protective. Both values less than 1.0 in the parenthesis provide confidence that the reported mean value of the risk measure was not observed by chance.
Thus, we see that existing evidence does support the notion that breastfeeding is beneficial to the health of infants and children. The disagreement, if any, is around the degree to which the association is likely to be causal for many of these conditions as, despite best efforts, individual studies might not have been able to fully remove the effect of existing confounders . Nevertheless, the state of the knowledge in this area should provide enough grounds for policy makers to develop evidence-based strategies in supporting women who have chosen to breastfeed to improve the health of infants and children.
Whilst the benefits of breastfeeding to infants and children are well established, the literature as to what extent breastfeeding may benefit women themselves is emerging. There is convincing evidence that breastfeeding and maternal breast cancer are negatively correlated. Breastfeeding for 18 + months over the lifetime of a woman (which may include breastfeeding more than one baby) is associated with significantly decreased risk of having breast cancer, compared to a woman who has never breastfed. A study conducted in the US has found that nearly 5,000 excess cases of breast cancer were associated with suboptimal breastfeeding durations . In the UK, optimal breastfeeding durations could have led to 865 fewer breast cancer cases for 313,000 first time mothers or a gain of 512 quality-adjusted life-years .
Apart from maternal breast cancer, there seems to be a lack of good quality evidence on the link between breastfeeding and other maternal outcomes. Breastfeeding is probably associated with ovarian cancer and rheumatoid arthritis, and possibly with several other health outcomes, e.g., maternal depression, endometrial cancer, osteoporosis and bone fracture . A relatively recent review finds studies that support the link between breastfeeding and Type 2 diabetes, breastfeeding and hypertension, and breastfeeding and coronary heart disease . Putting this evidence into perspective, optimal breastfeeding in the US could have averted an additional 8,500 myocardial infarction cases and an extra of over 36,000 hypertension cases .
It is important here to note that absence of good quality evidence does not necessarily imply that there is no association between breastfeeding and the above conditions. While we wait for more methodologically sound studies in the future to corroborate whether the probable and possible links are definitive ones, the current state of the knowledge appears to be enough for policy makers to develop and implement breastfeeding support strategies to improve maternal health outcomes.
How would reductions in the incidences of the above health conditions, as more women initiate, continue and exclusively breastfeed their infants, translate to benefits to national health systems? A systematic review of the evidence in this area found that increased breastfeeding rates were associated with potential cost-savings to the national health systems across a range of countries . Although the studies included in this review reported the impact of optimal breastfeeding differently (▶Table 11.2), the conclusion was robust: There is an economic case for breastfeeding support.
More economic studies have evolved since 2012 when the above review was published. ▶Table 11.3 summarises a cross-section of new studies (where more than one country is included) showing the economic benefits of optimal breastfeeding. Although each study has included different outcomes, employed different assumptions to model the cost savings, and used slightly different underlying methods, all studies highlight the economic loss currently observed due to suboptimal breastfeeding. In other words, if breastfeeding rates were increased at a level deemed appropriate or realistic in countries where breastfeeding rates are low, this would generate substantial cost savings to the respective national health system in each of these countries.
Reported economic impact
Ball and Wright 1999 
Excess costs of US $331 per not-breastfed infant for a year
Barton et al. 2001 
Mean difference of US $3,366 between breastfed and non-breastfed infants during neonatal unit stay for that year
Buchner et al. 2007 
A saving of Euro 250 per newborn per year on best case scenario of 100% breastfeeding for 6 months or more
Cattaneo et al. 2006 
Mean difference of Euro 160 per infant per year
Wight 2001 
Mean difference of US $200 per infant in the first 6- months of life; extra cost of US $9,669 per infant for not using human milk in a neonatal unit or a savings of US $11 per US $1 spent on human milk
Bartick and Reinhold 2010 
US $3.35 billion savings in treatment costs and US $13 billion including the value of premature deaths, at 90% breastfeeding rates
Drane 1997 
Australian $9 million in treatment costs and Australian $11.5 million including special education costs, at 80% breastfeeding rates
Riordan 1997 
Between US $1.2 and 1.3 billion in treatment costs attributable to formula feeding
Weimer 2001 
US $3.6 billion savings including the value of premature deaths, at 75% breastfeeding rate
Smith et al. 2002 
Australian $1.5 million in treating four diseases – gastrointestinal infections, respiratory illnesses, eczema, and NEC – in children aged 0–4 in Australian Capital Territory alone
Hoey and Ware 1997 
$200 per infant savings compared to bottle feeding
Reported economic impact
Pokhrel et al. 2015  Renfrew et al. 2012 
UK (covering 4 home countries)
Rollins et al. 2016 
Walters et al. 2016 
7 South East Asian countries
One of the methodological issues in the analysis of economic impact of suboptimal breastfeeding has been the uncertainty around the estimated impact. Most studies have relied on point estimates, but it is a well-known fact that several assumptions would have to be made in order to model such an impact under any employable method. These assumptions may in turn introduce uncertainty around the predicted impact. Some researchers ,  have looked at this important methodological issue. In the US, the most recent estimates put direct medical costs of suboptimal breastfeeding at $2.6 billion (95% confidence interval: $2.3, $2.9 billion), 79% of which are maternal . Having taken into account the uncertainty around various assumptions, their findings do not alter what has long been shown to be the case — that there are substantial costs associated with suboptimal breastfeeding to the national health services. Rather, the uncertainty analyses have provided decision makers with reasons to be confident about these findings and develop and implement breastfeeding support policies and strategies.
If breastfeeding improves ‘the quality of life for women through the reduction in incidence of breast cancer and for children through reducing acute and chronic diseases’ , it is reasonable to expect that this benefit may translate to wider societal impact. A healthier population with a better quality of life may be economically and socially more productive. More research is needed to explore this wider societal impact of improved breastfeeding rates.
Some evidence does exist to support the notion that the economic benefits of breastfeeding are wider. In a recent study, Rollins and colleagues found the optimal breastfeeding scenario in which every infant was breastfed until at least 6 months of age could have avoided $302 billion (0.49% of gross national income) globally in economic losses from cognitive deficits . Most of this economic cost was in high income countries ($231.4 billion or 0.53% of gross national income compared to $70.9 billion or 0.39% of gross national income in low- and middle-income countries). Another recent estimate by Walters and colleagues for seven South East Asian countries also shows that the economic loss due to cognitive deficit in suboptimal breastfeeding populations is high (US$1.63 billion)  corroborating similar estimates by Renfrew and colleagues for the UK  (▶Table 11.3).
Premature death is another wider outcome of suboptimal breastfeeding. The most recent estimate from the US shows that 3,340 premature deaths (between 1,886 and 4,785) could have been averted through optimal breastfeeding . Importantly, 78% of these deaths were maternal (986 due to myocardial infarction, 838 due to breast cancer, and 473 due to diabetes). Of the 721 excess deaths in children, sudden infant death syndrome claimed 492 lives and necrotising enterocolitis 190 lives.
Whether one should put a monetary value on premature deaths to reflect this aspect of societal impact is a contentious issue. Bartick and colleagues have used a method to assign a monetary value against their estimated premature deaths due to suboptimal breastfeeding . They find that the total cost of premature deaths was $14.2 billion (between $8.8 and $19.6 billion) and that the costs were evenly distributed between maternal and child population.
It has been argued that women who choose to breastfeed actually produce and supply breast milk and therefore contribute significantly to the national economy . Current human milk production levels exceed $3 billion annually in Australia and potentially $110 billion a year in the US, but premature weaning means nearly two thirds of this value may have been lost . Smith therefore argues that ‘failure to account for mothers’ milk production in GDP and other economic data has important consequences for public policy’ . The other side of the coin is the negative impact of breastfeeding. Breastfeeding is associated negatively with labour market outcomes, particularly for working women’s own earnings, work productivity and promotion prospects , , , . The opportunities forgone by choosing to spend longer time breastfeeding may be considerable for some women , particularly when emerging evidence does not support the perception that their breastfed babies may have better future-earning prospects than non-breastfed children . It is important to consider these negative implications of breastfeeding. However, the current state of the knowledge strongly implies that, on balance, optimal breastfeeding could potentially lead to substantially more societal benefits than societal costs.
The health, economic and societal benefits described above are “potential”. In other words, if we as a society were able to increase current breastfeeding rates (initiation, duration, and exclusivity) to an optimal level (e.g., all babies exclusively breastfed for 4 months), this increase in breastfeeding prevalence would generate those benefits. However, increasing breastfeeding rates would require health services to implement breastfeeding promotion and support interventions that are effective. Implementing effective breastfeeding promotion and support interventions will require upfront investment and the size of this investment may be considerable. Then the question is: Would we still get the substantial benefits described above after we have taken into account the cost of implementing the interventions? In other words, do breastfeeding promotion and support interventions provide “value for money”? This is exactly the sort of question decision makers often ask because, in their role as public health investor, they will need to justify whether the benefits described above outweigh the costs required to implement breastfeeding promotion and support programmes.
What does the evidence say? A summary of a cross section of published studies that looked at the cost-effectiveness of breastfeeding interventions is provided in (▶Table 11.4). Rice and colleagues find that enhanced contact with specially trained staff who provide education, support and a care plan for mothers is a cost-saving (cheaper and more effective) intervention, compared with usual care . Likewise, proactive telephone support where a feeding support team calls women daily for one week following hospital discharge offers some promise to be a cost-effective intervention, compared with reactive telephone support where women have to call the feeding team for any breastfeeding support . However, breastfeeding groups (BIG) for pregnant and breastfeeding women in a deprived area with weekly group meetings facilitated by a health professional is not cost-effective, as this intervention is unlikely to increase breastfeeding rates among women but costs similar compared to usual care (home visits) .
Study and Context
Rice et al. 2010 , UK hospital (neonatal unit)
Enhanced contact with specially trained staff. Staff provided education, support and a care plan for mothers.
Normal staff contact. Staff were not specifically trained to support breastfeeding mothers
Hoddinott et al. 2012 , Scotland (postnatal ward)
Proactive: Feeding support team called women daily for one week following hospital discharge. Whether to receive calls and with what frequency in the second week was chosen by women.
Reactive: Women could telephone the feeding team anytime over the two weeks following hospital discharge
Hoddinott et al. 2009 , UK, primary care
Breastfeeding groups (BIG) for pregnant and breastfeeding women in a deprived area. Included weekly group meetings facilitated by a health professional.
* A cross section of studies included in the UK National Health Service Economic Evaluation Database (NHS EED) (www.crd.york.ac.uk).
It seems that good quality economic evaluations in this area are sparse. Paucity of good quality studies evaluating cost-effectiveness of breastfeeding interventions does not necessarily mean that breastfeeding interventions do not provide good value for money; this simply shows current lack of good quality evidence in this area. However, many interventions aimed at promoting and/ or supporting breastfeeding are found to be effective. In their scrutiny of breastfeeding or feeding with breast milk interventions for infants admitted to neonatal units, Renfrew and colleagues found a number of interventions to be effective despite limitations in the evidence base . These interventions include: kangaroo skin-to-skin contact, peer support, simultaneous breast milk pumping, multidisciplinary staff training and the Baby Friendly accreditation of the associated maternity hospital .
More scrutiny elsewhere finds counselling (peers or health personnel), Baby-Friendly Hospital support, and community mobilisation approaches do improve breastfeeding prevalence, but higher impact can be achieved via running the interventions concurrently in a combination of health system, home, and community settings . Likewise, another systematic review finds that breastfeeding education/support generally increases the rates of exclusive breastfeeding and decreases no breastfeeding at birth, 4 weeks, and 60 weeks, but combined individual and group counselling seems more effective than individual or group counselling alone . Importantly, breastfeeding education/support interventions and peer support interventions in low- and middle-income countries have greater impact than those in high-income countries , .
Thus, breastfeeding support interventions exist in different guises and may include peer support, support units/teams, antenatal education, counselling, staff training, or school education but, whatever form it takes, providing breastfeeding support means scarce resources are utilised . As ‘many of these interventions inter-relate, it is unlikely that specific clinical interventions will be effective if used alone’ . More cost-effectiveness studies are therefore needed to help policy makers decide whether the overall benefits of these effective interventions, preferably implemented as a package, are worth their costs.
So far, we have seen that good quality evidence exists to demonstrate the scale of potential health, economic, and societal benefits that optimal breastfeeding (i.e. increasing breastfeeding initiation, duration and exclusivity) may generate, even after considering any negative impact of breastfeeding. However, it is unlikely that these potential benefits can be reaped without putting significant, upfront investments in breastfeeding promotion and/or support interventions. Also, many breastfeeding interventions are found to be effective in increasing breastfeeding initiation, duration, and exclusivity with varying degree of impact for low/ middle-income and high-income countries. However, we do not seem to have sufficient good quality economic studies evaluating the cost effectiveness of those interventions. In the absence of many good quality economic evaluation studies in this area, how can business cases be made for breastfeeding promotion and/or support? In particular, can the implementation costs of those effective breastfeeding interventions be justified?
Building a business case often means we provide a single metric that tells us the extent to which current investment will generate an economic return within a defined time horizon. Expected rates of return (RR) from any money invested in an economic activity (e.g., provision of breastfeeding support) would help us decide whether to undertake that investment. Thus, RR can help us compare investment priorities; the portfolio with higher RRs is prioritised over the ones with lower RRs . Variants of RRs exist in public health; the most common being a benefit-cost ratio . Public health investors can use information such as the benefit-cost ratio to make their case for investment or disinvestment explicitly. In the field of tobacco control, for example, it has been shown that every £1 invested in Stop Smoking Services in England leads to a return of £2.82 after 10 years . The National Institute for Health and Care Excellence (NICE) in England has developed a number of decision support tools, known as the return on investment (ROI) tools, to help public health investors build their business cases . A number of other decision support tools exist, including the one to estimate the value for money of social marketing campaigns to support breastfeeding, but a more comprehensive ROI tool for breastfeeding support interventions may be developed in the future .
Until there is a comprehensive tool exclusively developed to estimate the ROI of breastfeeding promotion and/or support interventions, we have to rely on published data to build business cases for breastfeeding promotion/support. One such example is presented below based on published data available for the UK setting .
Before embarking on a business case for breastfeeding promotion/support programmes, it is important to consider what evidence exists on effective interventions and what national guidance and strategies are in place in a particular context. As discussed earlier, breastfeeding promotion and support programmes work more effectively when they are delivered concurrently in a combination of different settings. This idea has been reflected in several key documents, such as the UNICEF Baby Friendly Initiative  and the NICE guidance on maternal and child nutrition .
Building a business case for breastfeeding promotion and support is usually achieved in several steps:
Step 1: Define intervention Usually, the intervention is ‘a multifaceted programme of interventions across different settings, including staff training, peer support, and activities to raise awareness and overcome barriers to breastfeeding, ensuring peer supporters are part of a multidisciplinary team and receive appropriate training’ .
Step 2: Identify and cost intervention components Once the intervention is defined, the next step is to identify individual components of the multifaceted package of interventions and cost them. Renfrew and colleagues present one such example for the region of Lancashire, UK (▶Table 11.5).
Step 3: Estimate the consequence of implementing the intervention To estimate the consequence of implementing the intervention, it is important to make three key assumptions:
Firstly, decide how much improvement in currently observed breastfeeding rates this multifaceted intervention is likely to bring. In the Lancashire example, it was assumed that implementing the above intervention would improve exclusive breastfeeding rates at 6 months from the current 0.5% to 7% (lower estimate) and exclusive breastfeeding rates at 4 months from the current 7% to 65% (higher estimate) . Note that the target rates are the ones currently observed for 4 months and at birth, respectively. Therefore, the intervention was assumed to support women who were exclusively breastfeeding at birth to continue until 4 months (lower estimate) and those who were exclusively breastfeeding at 4 months to maintain that until 6 months (higher estimate).
Secondly, decide how many infants will benefit from this intervention. Usually, this is the number of newly born babies who will survive in the current year (in the Lancashire example, n = 13,785 infants).
Thirdly, select the relevant “potential cost-savings (mean)” estimate provided by Renfrew and colleagues and summarised here in and multiply that figure by the number of infants expected to benefit from the interventions. In the Lancashire example, £9.93 (higher estimate) multiplied by 13,785 infants equates to approximately £136,891 (higher estimate) as the potential cost savings from gastrointestinal illnesses (“approximately” because the figures reported in ▶Table 11.6 are slightly different from the results of this simple calculation due to rounding).
Cost one-off (2012 prices)
Costs recurring (2012 prices)
Total costs (2012 prices)
UNICEF Baby Friendly Initiative accreditation for maternity units
Assumed to be in the budget already*
UNICEF Baby Friendly Initiative accreditation of universities
Assumed to be in the budget already*
Peer support services (priority national recommendation)
Assumed to be in the budget already*
Neonatal networks training
Provision of donor milk
Support service to filter harmful advertising
Breastfeeding-welcome employers and public space*
Assumed to fall outside of the health sector
Support to formula-feeding mothers
Included in current services, assumed no additional costs
Assumed to fall outside of the health sector
* It costs maternity/community units approx. £16,000 and universities £4,000 to go Baby Friendly.
**See , Appendix p. 202–203, for detailed costing.
This process is repeated for all other health outcomes, i.e. lower respiratory tract infection and acute otitis media in infants (n = 13,785 infants in Lancashire) and necrotising enterocolitis in preterm babies (n = 1,383 neonatal admissions in Lancashire).
Step 4: Estimate return on investment Fourthly, estimate the benefit-cost ratio by dividing the potential cost savings (incremental benefits) by the incremental costs of implementing the intervention. In this case, assuming that the benefits of breastfeeding are limited to the savings in treatment costs of acute diseases in children only (i.e., ignore maternal breast cancer benefits as reported by Renfrew et al 2012 ):
Potential cost savings
Mean (lower estimate)
Mean (higher estimate)
Total (lower estimate)
Total (higher estimate)
Gastroenteritis in infants
Lower respiratory tract infection in infants
Necrotising enterocolitis in neonatal units
Total potential cost savings from acute diseases in children (annual, 2012 prices)
Benefit-cost ratio for the current year = B / C = £82,667 / £446,300 = 0.19 (lower estimate) or = £553,454 / £446,300 = 1.24 (higher estimate) Now, assume that the size of the birth cohort for the next year is similar to this year but the incremental cost of intervention is lower (£329,300) as the health system would have to pay just the recurring costs from second year onwards. Recalculate the benefit-cost ratio (discounting for the second year could be ignored as the effect is relatively small).
Benefit-cost ratio for the next year = B / C = £82,667 / £329,300 = 0.25 (lower estimate) or = £553,454 / £329,300 = 1.68 (higher estimate)
Step 5: Interpret results with caveats The final step involves interpreting the benefit-cost ratio and acknowledging that the interpretation comes with some caveats.
The above benefit-cost ratios suggest that a multifaceted evidence-based breastfeeding support intervention as defined above is likely to be cost effective. The most conservative estimate (i.e. benefits limited to acute childhood diseases with lower estimates) suggests that there would be a net loss (£1 investment gives a return of £0.19 this year) but if considering higher estimates of the same benefits, the intervention is good value for money (£1 investment gives a return of £1.24 this year). As incremental costs of implementing the intervention decline in the following years as a result of not having to pay for one-off costs, the returns will increase
Several caveats are worth mentioning here. The benefits of breastfeeding support interventions are wider than just the cost savings in the acute childhood disease area. As discussed previously, good quality evidence consistently implies that breastfeeding-support interventions have potential to increase breastfeeding rates and such an increase generates wider benefits. The most obvious is the value of benefits from the treatment of fewer maternal breast cancer cases, which in the case of Lancashire could range from £399,000 to £724,000 over the lifetime of each annual cohort of first-time mothers . Once this benefit, together with the benefits coming from savings incurred elsewhere (e.g., fewer cases of sudden infant death syndrome and childhood obesity, and better cognitive outcomes) is included in the above calculation, a significant proportion of the investment is likely to be offset by the returns, even in the short term.
Over time, as members of staff providing breastfeeding support become more skilled, resulting in reduced additional training and leadership costs, the investment required to implement the intervention will fall from the initial level. This will make the investment generate a more favourable ROI. It is important, however, to recognise that it may take several years before a return on investment is seen from a breastfeeding-support intervention , but the estimates presented above suggest that evidence-based multifaceted support intervention is likely to result in a positive net benefit within a much shorter period.
The use of alternatives to human milk for feeding infants has attracted enormous debate within health economic literature. The answer to the question – what economic value may human milk have – varies depending upon what perspective is taken. Convincing evidence exists to support the viewpoint that breastfeeding is protective against a number of illnesses, in particular gastrointestinal illnesses, lower tract respiratory infections and acute otitis media in infants, necrotising enterocolitis in pre-term babies, and breast cancer in mothers.
Choosing to breastfeed is a complex economic decision that women have to make; the complexity is further aggravated by three key attributes: initiation, duration, and exclusivity. Women often weigh the benefits (incentives) of choosing to breastfeed, how long to breastfeed, and whether to breastfeed exclusively, against the costs (disincentives). Breastfeeding is often associated with significant private costs, in both money and time. In addition, maternal employment is negatively associated with breastfeeding durations, and the impacts often extend to wider labour market outcomes, such as prolonged maternity leave and limited prospects for promotion and productivity.
The current breastfeeding rates are suboptimal and increasing breastfeeding prevalence in those women who choose to breastfeed could bring substantial benefits to women themselves, to their children, to the national health systems and to the wider society at large. Women and children could enjoy better quality of life through reduction in their risk for certain diseases. To the national healthcare system, having to treat fewer cases of certain health conditions in a breast/milk feeding population means substantive cost savings. To the wider society, benefits accrue over time due to fewer premature deaths, coupled with lower prevalence of cognitive impairment and childhood obesity, in the breast/milk feeding population.
Good quality evidence on both the impact of breastfeeding as well as the effectiveness of breastfeeding promotion/support interventions is evolving. The current state of the knowledge clearly implies that promotion of breastfeeding on all three aspects (initiation, duration, and exclusivity) could lead to much more societal benefits than societal costs. It is therefore important to make a business case for breastfeeding promotion and support.
Whilst the evidence on economic impact of suboptimal breastfeeding is robust, the cost-effectiveness evidence of breastfeeding support interventions is relatively sparse. More research is therefore needed in this area. One way to make the economic case for breastfeeding promotion/support is to combine published robust studies around “what works” for increasing breastfeeding rates with other studies around the impact of suboptimal breastfeeding. This approach allows estimation of a single metric (e.g., benefit-cost ratio) demonstrating what economic returns are gained from every $1 spent in providing breastfeeding-support interventions.
Women are more likely to choose to breastfeed if they feel there are more incentives (benefits) than disincentives (costs)
There is evidence of a negative economic impact of suboptimal breastfeeding indicating national health services could save millions of dollars every year if current rates of breastfeeding were to increase
Investment in breastfeeding programmes needs to be targeted and evidence is available to suggest an integrated package of several interventions delivered concurrently is more cost effective than each activity delivered on its own
A return on investment approach to evaluate a package of breastfeeding-support interventions may be helpful for decision makers until more good quality cost-effectiveness studies evolve