Cost and Cost-Effectiveness of Donor Human Milk to Prevent Necrotizing Enterocolitis: Systematic Review

Abstract

Background:

Necrotizing enterocolitis (NEC) is a costly gastrointestinal disorder that mainly affects preterm and low-birth-weight infants and can lead to considerable morbidity and mortality. Mother's own milk is protective against NEC but is not always available. In such cases, donor human milk has also been shown to be protective (although to a lesser extent) compared with formula milk, but it is more expensive. This systematic review aimed at evaluating the cost of donor milk, the cost of treating NEC, and the cost-effectiveness of exclusive donor milk versus formula milk feeding to reduce the short-term health and treatment costs of NEC.

Materials and Methods:

We systematically searched five relevant databases to find studies with verifiable costs or charges of donor milk and/or treatment of NEC and any economic evaluations comparing exclusive donor milk with exclusive formula milk feeding. All search results were double screened.

Results:

Seven studies with verifiable donor milk costs and 17 with verifiable NEC treatment costs were included. The types of cost or charge included varied considerably across studies, so quantitative synthesis was not attempted. Estimates of the incremental length of stay associated with NEC were ∼18 days for medical NEC and 50 days for surgical NEC. Two studies claimed to report economic evaluations but did not do so in practice.

Conclusions:

It is likely that donor milk provides short-term cost savings by reducing the incidence of NEC. Future studies should provide more details on cost components included and a full economic evaluation, including long-term outcomes, should be undertaken.

Introduction

Globally, ∼1.1 m preterm (<37 weeks gestation) infants die each year, with the infant mortality rate for preterm infants at least three times that for term infants.¹ One condition mainly affecting preterm or low-birth-weight (<2,500 g) infants is necrotizing enterocolitis (NEC).² The incidence of NEC in developed countries can be as high as 13% among low-birth-weight infants and/or those born before 33 weeks.² NEC is usually treated medically with bowel rest and systemic antibiotics, but more severe cases are treated surgically via peritoneal drainage and laparotomy.^2,3 A systematic review suggested an overall mortality rate of around 20%,⁴ with up to fivefold higher mortality among those requiring surgery.⁵

Many agree that breast milk is a natural prophylactic for NEC^6–8 and mother's own breast milk is clearly the optimal choice for all infants for this, plus many other reasons.⁹ However, mothers with preterm babies may not be able to produce enough milk, may be too ill to breastfeed, or may die in childbirth.¹⁰ In this situation, the World Health Organization and others recommend using donor human breast milk (donor milk), which confers a number of advantages over the alternative of formula milk.^9,11–13 A recent systematic review⁷ reports that donor milk reduces the risk of NEC by almost two-thirds compared with formula. However, not all preterm babies who do not receive their mother's own milk receive donor milk instead.^14,15 Donor milk is more expensive than formula and although it confers various health benefits, it is not known whether its use is cost-effective.^16–18

We, therefore, undertook a systematic review to answer the following questions:

1. What is the cost of donor milk?

2. What is the initial cost to the health service of treating medical and surgical NEC?

3. Is exclusive donor milk feeding cost-saving and/or cost-effective compared with exclusive formula milk feeding when considering its impact on the short-term cost and health outcomes associated with NEC?

Materials and Methods

The protocol was registered on PROSPERO (reference number CRD42016042581).

Eligibility criteria

• Participants: All infants

• Intervention: Exclusive donor human breast milk

• Comparison (for economic evaluation studies): Exclusive formula milk

• Outcomes: Cost of donor milk, short-term health service cost of treating NEC (medical and surgical), cost-effectiveness of donor milk in terms of the short-term health and treatment costs of NEC. We did not specify a definition of NEC. Short term is defined as the time from birth to initial postnatal discharge. Costs data had to be “verifiable,” i.e., specified the data source. We sought costs data but included charges (to the patient or other payer) if costs were not reported.

• Study design: Any study reporting costs of donor milk and/or NEC, or any form of economic evaluation. Protocols, opinion pieces/editorials, and abstracts were excluded.

• Language and date: Due to resource constraints, only English language studies were included. We only sought studies published since January 1996 to recognize expected changes in cost structures over time.

• Publication status: Gray literature was sought.

Search methods

We carried out two search strategies: The first was to identify costs of donor milk, and the second was to identify costs of treating NEC. Both search strategies were also used to identify studies, including any form of economic evaluation. For both searches, the following databases were used: CINAHL, Cochrane Central Register of Controlled Trials (CENTRAL), EMBASE, Medline, and Google Scholar. The last searches were undertaken on October 11, 2016, for donor milk and on October 25, 2016, for NEC costs. A.B. conducted the searches.

Our broad search terms (as title or keyword) were as follows:

• Donor milk: We used a combination of the following: (Human milk or donor milk or donor human milk or donor breast milk or milk bank*) AND (cost or economic* or value or budget or fee or saving or income or price or expense).

• NEC: We used a combination of the following: (NEC or necrotizing enterocolitis or necrotizing enterocolitis) AND (cost or economic* or value or budget or fee or saving or income or price or expense).

The full search strategy can be found in Supplementary Data (Supplementary Data are available online at www.liebertpub.com/bfm). For both searches, the references of any relevant reviews or studies meeting the eligibility criteria were checked to identify any further studies/sources.

Study selection

The search results were de-duplicated by using EndNote v7. Both authors independently screened all of the titles and abstracts against the eligibility criteria. All studies that either author thought should be included were added to a full text list. These studies were then read independently in full, and each author made a decision on whether or not each study should be included. Discrepancies were discussed until agreement was reached.

Data collection process and data items

Data extraction was initially undertaken by A.B. and checked by C.T. A data extraction template was designed in Excel for each of the donor milk and NEC treatment costs research questions, including the following details:

• Donor milk: cost, units, year of costing, country/currency, cost components included (e.g., transport, payments to donors), and whether cost or charge.

• NEC treatment: cost by type (medical/surgical), year of costing, country/currency, cost components included (e.g., indirect costs such as “hotel” or overhead costs and physician fees), whether data were collected from a primary research study (i.e., with individual infants) or from a secondary source (i.e., generalized costings), sample size for primary studies, infants included/excluded (e.g., by birth weight, early mortality), definition of NEC, cost sources, whether cost or charge, whether costs/charges were adjusted for infant characteristics (in primary studies), and whether total cost of treatment or incremental (compared with an infant without NEC). We also extracted data on the length of stay associated with NEC.

Costs as opposed to charges, and incremental rather than total values were extracted if both were provided, although total (nonincremental) values were also extracted where these provided more detailed results (e.g., a breakdown by type of NEC or standard deviations). Values adjusted for infant characteristics were reported where available. For NEC treatment costs, the mean and standard deviation costs/charges were reported where provided. Exceptions (e.g., where a study only reported median values) were noted. For the economic evaluation studies, we planned to summarize information on sources of costs and effectiveness data and the methods used in calculating cost savings or cost-effectiveness, together with study results and details of any sensitivity analysis.

Strategy for data synthesis

All costs were inflated to 2015 local currency units and converted to 2015 U.S. Dollars at purchasing power parity by using the World Bank Consumer Price Index and exchange rates.¹⁹ For each estimate of the cost of donor milk, we calculated the cost per 100 mL as well as an indicative cost per infant in two scenarios. The first scenario was use with an infant with birth weight 500–1,250 g fed only with donor milk for the initial hospital stay, based on a total consumption of 11 L as estimated from the data presented in Cristofalo et al.'s recent randomized controlled trial.²⁰ The second scenario was use of donor milk as an adjunct to mother's own milk where available for an average infant <33 weeks or <1500 g admitted to neonatal intensive care, by using the mean consumption of 2.1 L per infant from Carroll and Herrman's study.²¹ The volume of donor milk per infant varied considerably in this study (range 3–9,271 mL where used; 28% received none) and was dependent on the method of feeding at discharge. However, we use the overall mean (including infants who did not receive any donor milk) because it is not always possible to know an infant's method of feeding at discharge at the time that donor milk would be initiated (and the ethics of rationing use to only those mothers definitely intending or able to breastfeed are also questionable). We aimed at synthesizing data on donor milk and NEC treatment costs by calculating mean costs across studies where there were sufficient comparable studies available, by using appropriate sub-groups (e.g., by country) if necessary. Where comparable data were not available, quantitative synthesis did not occur and a descriptive analysis was carried out. The results of the economic evaluation studies were summarized individually.

Results

Study selection

The numbers of studies screened, assessed for eligibility, and included and excluded in the review can be seen in Figure 1.

FIG. 1.

Study selection flow diagram. DM, donor milk; NEC, necrotizing enterocolitis.

Cost of donor milk

Eight estimates of the cost of donor milk were reported across seven studies.^4,21–26 The cost per 100 mL of donor milk in each study can be seen in Table 1. Only one study reported the cost of production,⁴ which at $51/100 mL was actually considerably larger than any of the studies reporting charges (to the health service). The authors noted that improvements in how the milk banking service was organized could reduce this cost to a more comparable $21/100 mL.⁴ As most milk banks are not for profit, we would expect there to be little practical difference between costs and charges, with the latter ranging from $8/100²⁶ to $21/100 mL.²⁵ The lowest two charges were actually those from milk banking services that made a monetary payment to donors to compensate them for the time and effort required to pump milk.²⁶ Although the small number of studies from any one country or region warrants caution, donor milk appears to be particularly expensive in the United Kingdom compared with in the United States and Scandinavia. We calculated the mean cost for the four U.S. studies ($14 per 100 mL, with indicative costs of $1,500 per infant with birth weight 500–1,250 g fed only on donor milk and $286 per very-low-birth-weight infant), but note that the four estimates differed in terms of the cost components included. Due to large variation in costs between settings, we did not attempt any further quantitative synthesis.

Table 1.

Cost/Charge of Donor Milk per 100 mL and Indicative Cost per Infant (for Exclusive Use of Donor Milk and Use as an Adjunct to Mother's Own Milk) and Inclusions in Cost/Charge

Author	Country	Cost/100 mL	Cost per infant (BW 500–1,250 g), exclusive donor milk (11 L)	Cost per infant (BW <1,500 g), donor milk as adjunct (2.1 L)	Payment to donor?	Processing	Transport to user	Handling	Cost/charge (to user)
Jegier et al.²²	United States	14.41	1,585	303	N				Charge
Ganapathy et al.²³	United States	10.68	1,175	224	N	Y	Y	Y	Charge
Carroll et al.²¹	United States	14.24	1,566	299	N		N		Charge
Huertas²⁴	United States	15.21	1,673	319	N	Y			Charge
U.S. mean		13.64	1,500	286
Renfrew et al.⁴	United Kingdom	51.05	5,612	1,072	N				Cost
Pokhrel et al.²⁵	United Kingdom	21.30	2,343	447	N		4.81		Charge
Arnold²⁶	Denmark	9.87	1,086	207	2.68				Charge
Arnold²⁶	Sweden	8.04	884	169	1.67				Charge (out of county hospitals)

All values are reported as 2015 USD at PPP. Cost/charge components marked as Y for included (value where provided), N for not included, and blank if not reported.

Studies are ordered by country and then by date.

BW, birth weight; PPP, purchasing power parity.

NEC treatment costs and length of stay

A total of 17 studies reported one or more values for NEC treatment costs.^{3,4,6,23,25,27–38} Table 2 details the NEC treatment costs and length of stay in each study separated by whether costs/charges were provided and whether these were incremental to those for an infant without NEC or total costs for an infant with NEC. Further characteristics of each study, including the definition of NEC employed (if provided), can be found in Supplementary Table S1. Due to the variation between studies in infant birth weight, definition of NEC, type of NEC treatment including the specific surgical procedure, which cost components were included, and healthcare context, there is a very large overall range of results in Table 2 and quantitative synthesis of costs data was not believed to be appropriate. For example, the lowest value of $3,025 was for the incremental cost of medical NEC (no definition provided) excluding indirect, “hotel” and physician costs,³² whereas the highest value of $604,526 was for the total charge for small bowel surgery (using ICD-9 code 777.5).³⁷ Even between two seemingly comparable U.S. studies, reported one year apart, the incremental cost of medical NEC had a fivefold variation, from $14,511²⁹ to $77,948.²³ NEC treatment costs also appear variable within studies. Where means and standard deviations were reported, the coefficient of variation ranged from 0.20 (incremental costs for surgical NEC from Ganapathy et al.²³) to 0.98 (total charges for large bowel surgery from Zhang et al.³⁷), although values of 0.3 to 0.4 were typical.

Table 2.

Initial Hospitalization Costs/Charges of Necrotizing Enterocolitis and Inclusions in Cost/Charge (in Addition to Direct Treatment Costs)

Author	Country	Birthweight	Total cost/charge	Medical cost/charge	Surgical cost/charge	Indirect costs (“hotel”/overheads) included?	LOS	Physician fees included?	Primary (P)/secondary (S) data
Total (nonincremental) costs/LOS
Weimer²⁷	United States	All			178,758	Y	39 days	N	S
Bartick et al.²⁸	United States	1,500–2,499 g		92,834		Y			S
Johnson et al.²⁹	United States	<1,500 g	111,298 ± 46,217			Y excluding overheads	Mean 87 days	N	P
Johnson et al.⁶	United States	<1,500 g, <35 weeks gestation	185,980 ± 104,079	157,252 ± 64,614	323,875 ± 151,991	Y	Mean 85 days		P
Stey et al.³Peritoneal drainage	United States	All			280,987 ± 97,787	Y	Median 51 days		P
Stey et al.³Peritoneal drainage and laparotomy	United States	All			405,257 ± 133,903	Y	Median 104 days		P
Stey et al.³Laparotomy	United States	All			347,994 ± 108,844	Y	Median 91 days		P
Smith et al.³⁰	Australia	1,500–1,999 g	30,733			Y			S
Struijs et al.³¹Early closure	Netherlands	All			134,980 (Median)	Y	Median 100 days	Y	P
Struijs et al.³¹Late closure	Netherlands	All			140,209 (Median)	Y	Median 96 days	Y	P
Incremental costs/LOS
Wight³²	United States	All		3,025		N		N	S
Ganapathy et al.²³	United States	≤28 weeks		77,948 ± 28,389	208,594 ± 40,846		Medical mean 12 days; surgical mean 43 days	N	P
Johnson et al.²⁹	United States	<1,500 g	17,056	14,511	24,665	Y excluding overheads	Difference in means 16 days	N	P
Johnson et al.²⁹	United States	<750 g	20,942			Y excluding overheads		N	P
Johnson et al.²⁹	United States	750–999 g	16,707			Y excluding overheads		N	P
Johnson et al.²⁹	United States	1,000–1,249 g	11,890			Y excluding overheads		N	P
Johnson et al.²⁹	United States	1,250–1,499 g	9,627			Y excluding overheads		N	P
Johnson et al.⁶	United States	<1,500 g, <35 weeks gestation	47,144			Y	Difference in means 15 days		P
Colaizy et al.³³	United States	<1,500 g		24,023	35,699	Y	17 days	Y	S
Renfrew et al.⁴	United Kingdom	500–999 g		49,518	96,877	Y	Medical 32 days; surgical 66 days	Y	S
Renfrew et al.⁴	United Kingdom	1,000–1,749 g		29,379	57,425	Y	Medical 19 days; surgical 39 days	Y	S
Renfrew et al.⁴	United Kingdom	1,750–2,500 g		44,579	72,625	Y	Medical 35 days; surgical 55 days	Y	S
Pokhrel et al.^25a	United Kingdom	All		26,797	29,151	Y	27 days	Y	S
Mahon et al.^34a	United Kingdom	All		25,158	27,711	Y	27 days	Y	S
Drane³⁵	Australia	1,500–1,999 g		14,469	19,113	N		Y	S
Drane³⁵	Australia	<1,500 g		25,548	49,904	N		Y	S
Total (nonincremental) charges/LOS
Abdullah et al.³⁶	United States	All	233,524	193,851	350,519		Medical median 36 days; surgical median 62 days		P
Zhang et al.³⁷All surgical	United States	All			507,811 ± 457,732		Mean 72 days		P
Zhang et al.³⁷Small bowel	United States	All			604,526 ± 487,942		Mean 87 days		P
Zhang et al.³⁷Large bowel	United States	All			463,267 ± 453,054		Mean 65 days		P
Zhang et al.³⁷Small and large bowel	United States	All			531,170 ± 423,001		Mean 79 days		P
Incremental charges/LOS
Bisquera et al.³⁸	United States	<1,500 g		121,015	305,739		Difference in meansMedical 22 days; surgical 60 days	N	P

All costs/charges are reported as 2015 USD at PPP, and means are reported unless specified. Cost/charge components marked as Y for included, N for not included, and blank if not reported. Costs/charges are reported as means (±standard deviation)/reference values from secondary sources unless stated.

Pokhrel et al. and Mahon et al. use the same underlying sources (UK Department of Health Reference Costs and NHS Hospital Episode Statistics) and approach to calculating costs, but the data are based on different years. Studies are ordered by country and then by date.

LOS, length of stay.

Length of stay data are more comparable between studies. Three studies reported estimated median total (nonincremental) length of stay for surgical NEC based on primary data collection,^3,31,36 which ranged from 51 to 104 days. One further study reported an estimate of the mean of this parameter of 72 days.³⁷ Based on the results of Bisquera et al.³⁸ and Ganapathy et al.,²³ the incremental length of stay due to surgical NEC (compared with that for an infant without NEC) is likely to be around 50 days. Fewer studies reported primary data for medical NEC, with one estimate of total median length of stay of 36 days,³⁶ of which around half are likely to be the incremental length of stay associated with NEC.^23,38

Economic evaluation of donor milk compared with formula milk

We initially identified three economic evaluations.^4,32,39 Although none of these met all of our inclusion criteria, it is important to consider why and to discuss other limitations of the two studies that claimed to provide economic evaluations of the use of donor milk. Wight³² presented a return on investment analysis for the use of donor milk for very-low-birth-weight infants (<1,500 g). Three outcomes were considered: NEC, sepsis, and overall length of stay (for any reason related to the use of human milk rather than formula). The analysis was based on local treatment costs of NEC and sepsis and daily “hotel” costs, but excluding physician fees. The unit cost of donor milk was provided by the local milk bank. Effectiveness data were taken from a single, nonrandomized study comparing the use of fortified mother's own milk (not donor milk) with formula,⁴⁰ which is problematic as donor milk is unlikely to be as effective as mother's own milk⁴ (although for ethical reasons there are no “head-to-head” trials of donor milk versus mother's own milk). In the economic analysis, infants could be fed with donor milk for either one or two months, although no rationale for these durations was provided and effectiveness was assumed to be equal regardless of the duration of use. The method of calculating total milk requirements (2,000 mL for one month or 7,100 mL for two months) was not described. Formula milk was assumed to be free of charge. The year of costing was not specified, and no sensitivity analysis was performed.

Arnold³⁹ claimed to present three models of the cost-effectiveness of banked donor milk for the prevention of NEC in preterm infants. The first model repeated Wight's results. The second model multiplied length of stay data from Schanler et al.'s study⁴⁰ by 1997 daily neonatal intensive care unit charges to give a length of stay-based charges saving. Donor milk unit costs were stated but could not be verified. Donor milk was assumed to be provided for 60 days (a rationale for this duration was not provided), although a clear description of the total volume of donor milk required was included. Formula milk was assumed to be free of charge. No sensitivity analysis was performed. The third model claimed to calculate the preventable costs of NEC to the U.S. state of Texas (i.e., comparing the cost of NEC in the scenario where all preterm infants were fed human milk [either mother's own milk or donor milk] with the current scenario where some infants were not fed with human milk but with formula instead). NEC costs were taken from Bisquera et al.'s study,³⁸ although this study actually provides charges and not costs. No consideration of the cost of donor or formula milk was included, so this model cannot be described as an economic evaluation. The reduction in the incidence of NEC from the use of donor milk was calculated by subtracting the 1.2% risk of NEC in infants fed donor milk in a small randomized UK study⁴¹ from the 10.1% risk of NEC for all infants (regardless of type of milk) in the United States.⁴² This is problematic because differences in study contexts reduce comparability and a failure to consider that the 10.1% risk for all infants is a weighted average of the risks for human and nonhuman milk and not a sum of the risks in these two feeding conditions.

The main intervention being evaluated in Renfrew et al.'s economic evaluation⁴ was the provision of a lactation consultant to help mothers breastfeed their own infants. In a secondary analysis, the use of donor milk was considered as an adjunct to mother's own milk, so the comparison was not between exclusive use of donor milk and exclusive use of formula.

Discussion

We found seven studies including verifiable costs or charges of donor milk. Four of these were conducted in the United States, and the mean cost of donor milk across these studies was approximately $14 per 100 mL. Based on existing estimates of donor milk use,^20,21 this would equate to costs of just under $300 per very low-birth-weight infant and of around $1,500 for an infant with birth weight 500–1,250 g fed entirely on donor milk. We found 17 studies including initial costs to the health service/charges for the treatment of NEC; however, there was insufficient comparability between the estimates provided to enable quantitative synthesis. Many studies did not even report what cost components were or they were not included in their estimates. In particular, studies reporting the cost of donor milk did not include the fixed costs of setting up a milk bank (meaning the costs reported here are underestimates) or which specific tests on donors and their milk were undertaken. In addition, studies only reporting charges are particularly difficult to synthesize as cost-to-charge ratios vary between hospitals. Even within primary studies evaluating the cost of treating NEC, there was considerable variation between costs for individual infants, suggesting that costs are also affected by a range of other factors/comorbidities. Future costing studies should clearly specify which costs are included, if possible by using a “bottom-up” approach to costing.

Estimates of (excess) length of stay are easier to compare than NEC treatment costs and enable local bed-day costs to be applied by others wanting to estimate the financial impact of NEC on their own institution. The results of primary studies suggested that medical NEC would add around 18 days to an infant's length of stay and surgical NEC would add around 50 days.

It is, however, clear that the incremental costs of treating infants with NEC are large, so any potential preventative measures are worthy of consideration. We were unable to include any economic evaluations of exclusive donor milk versus formula milk in our review, even if we extended the outcomes included beyond NEC. Both Wight³² and Arnold³⁹ suggested that exclusive donor milk could be cost-saving and would, therefore, dominate the use of formula milk for preterm infants. However, neither study is of sufficient methodological quality to provide convincing evidence of the cost-effectiveness of donor milk. Our protocol did not include the use of the results from our review of costs in a “back of the envelope” economic evaluation that could be subject to similar concerns regarding methodological quality. Therefore, although we believe it is likely that the use of donor milk would be at least cost-effective, and possibly also cost-saving compared with formula milk, we agree with others^16–18 that a full economic evaluation is warranted. Such an evaluation could consider both the use of exclusive donor milk versus exclusive formula milk, different durations of donor milk feeding (with differential effectiveness) and/or the use of donor milk as a complement or “top-up” to mother's own milk while breastfeeding is being established. It should include all the major short- and long-term health (and associated cost) consequences related to the use of donor rather than formula milk (not just NEC) and potentially issues related to the acceptability of donor milk to parents.^18,43,44 Although our results provide a starting point, any economic evaluation will need to clearly justify the costs of donor milk and NEC used as parameter inputs and, given the wide range of costs in the studies included in this review, it should include a range of sensitivity analyses to determine the robustness of the results to different costs of donor milk and NEC treatment. It may also be helpful to contact milk banks directly for costs data as part of an economic evaluation.

Our review is not without its limitations. On the cost of donor milk, we did not compare donor milk practices and organizations, health systems, or breastfeeding cultures in each of the countries in which the included studies were undertaken and it may be that the large variation in costs is partially explained by such differences, which may even exist within countries. In the United Kingdom, for example, England tends to be served by local milk banks (n = 14 nationwide) whereas one bank covers the whole of Scotland.⁴⁵ On the cost of NEC, we did not specifically search for studies assessing length of stay that did not also include costs so we may have missed some primary studies. On economic evaluations, our inclusion criteria meant that only studies comparing the use of exclusive donor with formula milk were included. The use of exclusive donor milk is rare in practice (but may occur, for example, where a mother dies in childbirth or is HIV positive and cannot safely breastfeed). This restriction was intended to ensure a consistent comparator so that we could synthesize results across studies but may mean that we have excluded studies that would have provided an insight into the cost-effectiveness of donor milk. We did not compare donor milk with mother's own milk because the latter would clearly dominate in an economic evaluation, being both cheaper and at least as effective. Using the idea of extended dominance, if exclusive donor milk was found to be cost-effective compared with formula, then using donor milk as an adjunct to mother's own milk would be even more so.

We focused on the incidence and short-term costs of NEC as the only outcome associated with the use of donor milk. It may also be the case that donor milk reduces the severity of NEC, although to claim this we currently need to generalize from the results of a study using mother's own milk.⁴⁶ There are many other potential benefits of human milk such as reduced neurodevelopmental complications, neonatal infections, and improved cardiovascular health,^47–50 as well as potential risks if safety standards are not maintained¹⁷ and a possible link with slower weight gain.⁷ These additional outcomes would have implications for both the costs and health outcomes associated with the use of donor milk. Spill-over effects should also be considered. One concern is that the provision of donor milk “crowds out” mother's own milk and this reduces breastfeeding rates. However, a recent systematic review suggests that this is not the case in practice.⁵¹ The studies identified were all carried out in developed countries, which may limit the international generalizability of the findings and publication bias may also be a problem, particularly for any economic evaluations that did not find positive evidence of the use of donor milk. Although we used Google Scholar in an attempt to find gray literature, we were unable to evaluate the extent of publication bias. We may also have missed studies published in languages other than English. In an attempt to maximize the credibility of our findings, we excluded studies where costs could not be “verified,” that is, traced back to their original source. However, no other quality appraisal of costing studies was undertaken. We also excluded a study where donor milk was purchased on the open market rather than from an official milk bank.⁵²

Conclusion

It is likely that the use of donor milk is cost-effective. To strengthen the evidence base, there is a need for a comprehensive economic evaluation of the use of donor milk, focusing on providing evidence in contexts where the use of donor milk is not currently standard practice. Such work should carefully describe how the costs of donor milk and of its implications for healthcare have been generated, given the variability in the estimates that we have identified between contexts, and, for some outcomes, also within contexts.

Acknowledgments

C.T. is funded by the NIHR CLAHRC West Midlands initiative. This article presents independent research, and the views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR, or the Department of Health.

Footnotes

Disclosure Statement

No competing financial interests exist.

References

World Health Organization. Born Too Soon: The Global Action Report on Preterm Birth. Geneva: World Health Organization, 2012.

Niño

, Sodhi

, Hackam

. Necrotizing enterocolitis: New insights into pathogenesis and mechanisms. Nat Rev Gastroenterol Hepatol, 2016; 13:590–600.

Stey

, Barnert

, Tseng

, et al. Outcomes and costs of surgical treatments of necrotizing enterocolitis. Pediatrics, 2015; 135:e1190–e1197.

Renfrew

, Craig

, Dyson

, et al. Breastfeeding promotion for infants in neonatal units: A systematic review and economic analysis. Health Technol Assess, 2009; 13:1–146.

Neu

, Walker

. Necrotizing enterocolitis. N Engl J Med, 2011; 364:255–264.

Johnson

, Patel

, Bigger

, et al. Cost savings of human milk as a strategy to reduce the incidence of necrotizing enterocolitis in very low birth weight infants. Neonatology, 2015; 107:271–276.

Quigley

, McGuire

Formula versus donor breast milk for feeding preterm or low birth weight infants. Cochrane Database Syst Rev, 2014:CD002971.

, Brewer-Asling

, Magnus

. A case study on the economic impact of optimal breastfeeding. Matern Child Health J, 2013; 17:9–13.

World Health Organization. Guidelines on Optimal Feeding of Low Birth-Weight Infants in Low- and Middle-Income Countries. Geneva: World Health Organization, 2011.

10.

Murguia-Peniche

, Kirsten

. Meeting the challenge of providing neonatal nutritional care to very or extremely low birth weight infants in low-resource settings. World Rev Nutr Diet, 2014; 110:278–296.

11.

Arslanoglu

, Corpeleijn

, Moro

, et al. Donor human milk for preterm infants: Current evidence and research directions. J Pediatr Gastroenterol Nutr, 2013; 57:535–542.

12.

Eidelman

, Schanler

, Johnston

, et al. Breastfeeding and the use of human milk. Pediatrics, 2012; 129:e827–e841.

13.

UNICEF. Baby friendly initiative. 1995; Available at https://www.unicef.org.uk/babyfriendly/what-is-baby-friendly (accessed September 11, 2016).

14.

Zipitis

, Ward

, Bajaj

. Use of donor breast milk in neonatal units in the UK. Arch Dis Child Fetal Neonatal Ed, 2015; 100:F279–F281.

15.

Hagadorn

, Brownell

, Lussier

, et al. Variability of criteria for pasteurized donor human milk use a survey of US neonatal intensive care unit medical directors. JPEN J Parenter Enteral Nutr, 2016; 40:326–333.

16.

James Lind Alliance. Preterm birth top 10. 2014. Available at www.jla.nihr.ac.uk/priority-setting-partnerships/preterm-birth/top-10-priorities (accessed December 1, 2016).

17.

NICE. CG93 Donor Milk Banks: Service Operation. London: NICE, 2010.

18.

British Association of Perinatal Medicine. The use of donor human expressed breast milk in newborn infants: A framework for practice. 2016. Available at www.bapm.org/publications/DEBM%20framework%20July%202016.pdf (accessed 01/12/2017).

19.

World Bank. World development indicators. 2016. Available at http://data.worldbank.org/data-catalog/world-development-indicators (accessed December 1, 2016).

20.

Cristofalo

, Schanler

, Blanco

, et al. Randomized trial of exclusive human milk versus preterm formula diets in extremely premature infants. J Pediatr, 2013; 163:1592–1595.e1591.

21.

Carroll

, Herrmann

. The cost of using donor human milk in the NICU to achieve exclusively human milk feeding through 32 weeks postmenstrual age. Breastfeed Med, 2013; 8:286–290.

22.

Jegier

, Meier

, Engstrom

, et al. The initial maternal cost of providing 100 mL of human milk for very low birth weight infants in the neonatal intensive care unit. Breastfeed Med, 2010; 5:71–77.

23.

Ganapathy

, Hay

, Kim

. Costs of necrotizing enterocolitis and cost-effectiveness of exclusively human milk-based products in feeding extremely premature infants. Breastfeed Med, 2012; 7:29–37.

24.

Huertas

KP.

A Sunrise Review: Mandated Healthcare Coverage for Banked Human Milk. Kent, WA: Washington State Department of Health, 2015.

25.

Pokhrel

, Quigley

, Fox-Rushby

, et al. Potential economic impacts from improving breastfeeding rates in the UK. Arch Dis Child, 2015; 100:334–340.

26.

Arnold

LD.

Donor milk banking in Scandinavia. J Hum Lact, 1999; 15:55–60.

27.

Weimer

JP.

The Economic Benefits of Breastfeeding: A Review and Analysis. Washington, DC: US Department of Agriculture, 2001.

28.

Bartick

, Reinhold

. The burden of suboptimal breastfeeding in the United States: A pediatric cost analysis. Pediatrics, 2010; 125:e1048–e1056.

29.

Johnson

, Patel

, Jegier

, et al. Cost of morbidities in very low birth weight infants. J Pediatr, 2013; 162:243–249.e241.

30.

Smith

, Thompson

, Ellwood

. Hospital system costs of artificial infant feeding: Estimates for the Australian Capital Territory. Aust N Z J Public Health, 2002; 26:543–551.

31.

Struijs

, Poley

, Meeussen

, et al. Late vs early ostomy closure for necrotizing enterocolitis: Analysis of adhesion formation, resource consumption, and costs. J Pediatr Surg, 2012; 47:658–664.

32.

Wight

NE.

Donor human milk for preterm infants. J Perinatol, 2001; 21:249–254.

33.

Colaizy

, Bartick

, Jegier

, et al. Impact of optimized breastfeeding on the costs of necrotizing enterocolitis in extremely low birthweight infants. J Pediatr, 2016; 175:100–105.e102.

34.

Mahon

, Claxton

, Wood

. Modelling the cost-effectiveness of human milk and breastfeeding in preterm infants in the United Kingdom. Health Econ Rev, 2016; 6:54.

35.

Drane

Breastfeeding and formula feeding: A preliminary economic analysis. Breastfeed Rev, 1997; 5:7.

36.

Abdullah

, Zhang

, Camp

, et al. Necrotizing enterocolitis in 20 822 infants: Analysis of medical and surgical treatments. Clin Pediatr, 2010; 49:166–171.

37.

Zhang

, Ortega

, Camp

, et al. Necrotizing enterocolitis requiring surgery: Outcomes by intestinal location of disease in 4371 infants. J Pediatr Surg, 2011; 46:1475–1481.

38.

Bisquera

, Cooper

, Berseth

. Impact of necrotizing enterocolitis on length of stay and hospital charges in very low birth weight infants. Pediatrics, 2002; 109:423–428.

39.

Arnold

LD.

The cost-effectiveness of using banked donor milk in the neonatal intensive care unit: Prevention of necrotizing enterocolitis. J Hum Lact, 2002; 18:172–177.

40.

Schanler

, Lau

, Hurst

, et al. Randomized trial of donor human milk versus preterm formula as substitutes for mothers' own milk in the feeding of extremely premature infants. Pediatrics, 2005; 116:400–406.

41.

Lucas

, Cole

. Breast milk and neonatal necrotising enterocolitis. Lancet, 1990; 336:1519–1523.

42.

Uauy

, Fanaroff

, Korones

, et al. Necrotizing enterocolitis in very low birth weight infants: Biodemographic and clinical correlates. J Pediatr, 1991; 119:630–638.

43.

Lording

RJ.

A review of human milk banking and public health policy in Australia. Breastfeed Rev, 2006; 14:21–30.

44.

Underwood

MA.

Human milk for the premature infant. Pediatr Clin North Am, 2013; 60:189–207.

45.

UK Association of Milk Banks. Find a milk bank. 2017; Available at www.ukamb.org/milk-banks (accessed February 21, 2016).

46.

Guthrie

, Gordon

, Thomas

, et al. Necrotizing enterocolitis among neonates in the United States. J Perinatol, 2003; 23:278–285.

47.

Hanson

, Korotkova

. The role of breastfeeding in prevention of neonatal infection. Semin Neonatol, 2002; 7:275–281.

48.

Gibertoni

, Corvaglia

, Vandini

, et al. Positive effect of human milk feeding during NICU hospitalization on 24 month neurodevelopment of very low birth weight infants: An Italian cohort study. PLoS One, 2015; 10:e0116552.

49.

Le Huerou-Luron

, Blat

, Boudry

. Breast- v. formula-feeding: Impacts on the digestive tract and immediate and long-term health effects. Nutr Res Rev, 2010; 23:23–36.

50.

Menon

, Williams

. Human milk for preterm infants: Why, what, when and how?. Arch Dis Child, 2013; 98:F559–F562.

51.

Williams

, Nair

, Simpson

, et al. Use of donor human milk and maternal breastfeeding rates a systematic review. J Hum Lact, 2016; 32:212–220.

52.

Geraghty

, McNamara

, Dillon

, et al. Buying human milk via the internet: Just a click away. Breastfeed Med, 2013; 8:474–478.

53.

Stoll

, Hansen

, Fanaroff

, et al. Late-onset sepsis in very low birth weight neonates: the experience of the NICHD Neonatal Research Network. Pediatrics, 2002; 110:285–291.

54.

Fanaroff

, Korones

, Wright

, et al. Incidence, presenting features, risk factors and significance of late onset septicemia in very low birth weight infants. The National Institute of Child Health and Human Development Neonatal Research Network. Pediatr Infect Dis J, 1998; 17:593–598.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.05 MB