Exclusive breastfeeding of infants up to six months of age provides the best nutrition and immunologic protection for the promotion of healthy growth and development in non-crisis and crisis settings. [1], [2] While numerous myths and cultural taboos surround the ability of mothers to breastfeed in the immediate aftermath of an emergency, breastfeeding is physiologically possible for most women. [1], [3] There are however, a minority of mothers in crisis settings who may have difficulty nursing their infants due to stress, fatigue, lack of privacy, dehydration, and inadequate diet. [1] Additionally, the immediate chaos and impact of the emergency may result in the death of mothers or separation of mothers and infants. Infants less than six months of age, who are not breastfed, need urgent identification and targeted skilled feeding support.

The priority approach to feeding these infants should be through relactation with their own mother or breastfeeding through a wet nurse. [1] If these options are not possible or when skilled staff, such as health providers or infant feeding counselors, indicate that breast milk substitutes (BMS) are necessary, this must be accompanied by training on hygiene, preparation and use of BMS. BMS in an emergency carries a substantial risk of malnutrition, morbidity and mortality, and should be a last resort option, used only when other safer options have been fully explored. In the majority of natural disasters and complex emergencies, BMS are donated and widely distributed with little monitoring or guidance to mothers. This often results in poor preparation using contaminated water, thereby significantly increasing infants’ risk of morbidity and mortality. [4] In emergency situations where there are large number of non-breastfed infants, the substantial risks associated with BMS distribution must be weighed against the risk malnutrition among infants consuming commercial products not intended for use as a BMS.

Haiti presented a complex set of challenges for addressing infant and young child feeding (IYCF) in the aftermath of the 2010 earthquake, including sub-optimal feeding practices prior to the earthquake as demonstrated by an exclusive breastfeeding prevalence of 21.7% in Port-au-Prince. [5] Mixed feeding, whereby mothers give breast milk and other replacement feeds, such as diluted sweetened condensed milk, was pervasive. [5] Furthermore, cultural taboos related to breastfeeding dissuaded many mothers from appropriate IYCF practices. [1], [3] In Haiti, beliefs that “strong emotions or bad news either spoil mother’s milk or stop its production; and that a mother who spends time under the sun cannot breastfeed because her milk becomes hot and may cause diarrhea to the child were particularly relevant in the post-earthquake displaced persons tents”. [6] Haiti also had the highest HIV prevalence rate (3.8%) in the Caribbean region, which served as an additional perceived barrier to breastfeeding. [5] Another complexity of the situation was that 20% of Haitian infants/children were classified as orphaned or vulnerable prior to the earthquake. [5]

Based on the large-scale displacement into makeshift shelters with significant issues of crowding; poor sanitary conditions, including limited excreta disposal; limited cooking facilities; the cessation of pre-existing weak public health services; and an increased risk for the spread of communicable diseases; the humanitarian community, in collaboration with the Ministère de la Santé Publique et de la Population (MSPP), identified the need to address IYCF, including the provision and management of BMS as a part of the nutrition response. Specifically, Ready to Use Infant Formula (RUIF) was identified as the safest option. RUIF is a commercially produced (Codex Standard 72–1981) hygienic, pre-mixed liquid infant formula that is consumed directly from the container or a cup, avoiding the use of water, bottles and teats. RUIF was already in use on an individual basis in select hospitals for the care of premature infants.

The National Nutrition Cluster (NNC) in Haiti, chaired by MSPP and UNICEF, was responsible for the procurement, management and administration of RUIF. The NNC non-governmental (NGO) partners established Points de Conseil en Nutrition pour Bébé (PCNB) or ‘baby tents’ in most tent cities throughout Port-au-Prince, Leogane and Jacmel. The baby tents provided a safe place for mothers to breastfeed and caregivers to receive IYCF counseling. The intent was to provide RUIF in some of these tents through strict criteria to targeted infants. [7]

A retrospective record review was conducted from 14 to 27 July, 2010 to obtain data on infants receiving RUIF in baby tents. Selection criteria for baby tents included: 1) the baby tent was operational in April 2010 in Port-au–Prince, 2) the baby tent was enrolling infants at the time of sample selection (June 2010), and 3) the baby tent planned to continue services through December 2010. A standardized data collection form was created based on data collected across baby tents and NGOs and included: basic demographics, admission criteria, primary caretaker, feeding practices, and admission and follow-up anthropometrics. Age of the infant was collected from the caregivers; the majority was able to provide a date of birth for the infant. Food intake was assessed through a 24 hour recall for each infant. Based on the specific NGO program, beneficiaries received RUIF either daily or weekly; therefore the records for each infant receiving RUIF differ in terms of the number of visits and anthropometric measurements.

Thirty baby tents administered by six NGOs were included in the evaluation. Trained enumerators extracted data from 590 individual infant records of infants admitted between April 1 and June 29, 2010. There were 16,946 individual visits recorded among these infants during this time period. Of the 590 infant admitted, 493 (83.6%) were included in the analysis. Infants were excluded for the following reasons: missing data (sex, date of birth, admission weight), implausible weight-for-age at admission, age was outside of the admission criteria (i.e greater than 12 months of age), enrollement start date was prior to April as defined by the protocol, or if the infant only attended the baby tent one time. The majority of exclusions (44.3%) came from infants only visiting the baby tent one time. Sixty-eight percent of infants included in the analysis were between 0 and 5 months of age at admission with a median age of 4 months. There were similar numbers of boys and girls, 249 (50.5%) and 244 (49.5%), respectively.

Infants were admitted for varying reasons. The most frequent reason was the infant was orphaned (41%), followed by HIV-infected mother (15%) (Table 1). Among orphans, 18.3% were single orphans where the mother had died, 81.7% were double orphans. A smaller proportion of children were reported to be abandoned (8.3%), with the majority (83%) reported to be abandoned by both parents. Based on the admission data, 49.3% of infants receiving RUIF were orphaned or abandoned and could be considered extremely vulnerable. The majority (96%) of infants with HIV-infected mother were admitted to one NGO program that focused on HIV care, which had been established and was providing BMS prior to the earthquake. However, only 14% of these infants were born before the earthquake and therefore admission of the majority of infants based on this criterion was contrary to the guidance provided in the baby tent guidelines as they were not artificially fed prior to the earthquake. A second NGO was focused on orphaned and abandoned infants; 83.4% of their admissions were based on this criterion and only 1 mother was reported to be a caregiver in this program. While re-lactation was a goal of the PCNB program, only 1% of mothers were actively attempting re-lactation.

Forty-eight (9.7%) of infants in the program were erroneously admitted because they weighed less than 3 kg at younger than 6 months of age and, according to program guidelines, should have been referred for inpatient treatment of acute malnutrition. None of these infants were referred for treatment; four defaulted.

Mothers were the most frequent (38%) caregiver present at the time of the infants’ admissions, followed by aunts and grandmothers (26% and 13%, respectively). Of note, 11% of caregivers were male: 10% fathers and 1% uncles (Table 2).

Forty-five infants, all less than 6 months of age, only received breast milk prior to admission; however, 86.7% of them were admitted within 1 week of their birth into the NGO program specific for infants with HIV-infected mothers. The majority of infants under 6 months of age (87%, 292/337) were receiving foods in addition to or in replacement of breast milk at the time of admission. All children aged 6 months and older were consuming food other than breast milk. Across age groups, the use of powdered infant formula and powdered milk (not intended for infant feeding) was quite high (Table 3).

WAZ were calculated as the primary anthropometric indicator of nutritional status. At admission, the mean WAZ of all infants was −1.41 (SD 1.82). Younger infants had a significantly lower WAZ upon admission compared with older infants, −1.80 versus −1.23 (p = 0.001), respectively. Among orphans, the mean WAZ at admission was −1.37 (SD 1.82). There was a significant difference in the mean WAZ at admission between NGO’s (ANOVA, p<0.001), with one NGO having a mean WAZ of −2.31.

There was no significant difference in mean WAZ (pooled t test, t = 0.55, p = 0.58) nor in the prevalence of underweight (χ² = 0.26, p = 0.6; 50.9% and 49.1% were underweight) of orphans compared with non-orphans at admission, respectively.

Thirty-four percent of infants were underweight at enrollment; 19.5% of infants enrolled in the program were severely underweight. Of note, more infants were severely underweight (WAZ <−3) compared with moderately underweight (WAZ >−3 to <−2) (Table 4). Within one NGO program, 60% of admissions were underweight (WAZ <−2).

Seventy-four (15%) infants exited the program before the end of June; 27% of exits occurred among children who “aged” out of the programs at 12 months of age and 6.0% of exits were a result of death. Defaulting accounted for the majority of exits, 36 (49%). There was no further information on defaulters or deaths. Only two infants exited to breastfeeding programs through successful re-lactation programming.

The number of visits each infant made to the PCNB varied significantly, ranging from 2 to 102 visits (mean 30). Measurements were not taken at each visit; the average number of visits with a measurement was 14 (range 2–99). Older infants (6–12 months) made more visits to the baby tents (median 25.5) than younger infants (median 17) (pooled t-test p = 0.005).

Responding to infant feeding needs in emergency contexts continues to present a challenge. Since 2010, RUIF has not been used on a large scale as subsequent emergencies have not presented the same context as Haiti. The floods in Pakistan in 2010, the Horn of Africa crisis of 2011, the Sahel emergency of 2012 and the ongoing displacement of Syrians have all resulted in unsolicited donations and continue to highlight the need to control donations and minimize the inappropriate use of BMS in a more coordinated manner. While quantitative data, such as weight gain, may not be available as evidence that the baby tent program was successful, the experience in Haiti did provide knowledge and experience for IFE response in subsequent emergencies. Since 2010, IYCF counseling and support have become increasingly prominent components in emergencies, including in the Syrian crisis, where similar baby ‘caravans’ are being operated in refugee camps of Jordan and relactation is more successful. A scaled up interest in IFE has been noticed, including revisions of standard operating procedures for the handling of BMS, new assessment methodologies and development of appropriate and feasible indicators for assessment.

If the baby tent program with RUIF model is to be implemented as a component of IYCF emergency programs, specific actions must be taken by the IYCF in emergencies community. A generic protocol should be drafted and readily adapted to the context of emergencies which require IYCF interventions. This protocol should contain: admission and discharge criteria, standardized infant records forms and guidance on monitoring infant growth, minimal program monitoring and evaluation package, training package, appropriate numbers of staff, logistics components which address procurement, warehousing, transport and waste disposal issues. An IYCF toolkit encompassing these areas is underdevelopment. Training NGOs on the baby tent/RUIF model for IYCF in emergencies should be a mandatory part of preparedness activities. Special attention should be paid to NGOs with longstanding programs in emergency prone areas, and how short-term emergency IYCF programming can complement their programs. A stockpile of suitable and precise scales and length mats should be maintained for rapid deployment. Finally, it is important to remember that the use of BMS is only a small part of an overall IYCF program, albeit a very intensive one. It is easy to lose sight of the overall support for breastfeeding mothers given the added complexity and demands of the BMS component, but programs should still focus on the broader aspects of IYCF: support for breastfeeding mothers and appropriate complimentary feeding of older infants.