Neonatal home care practices in rural Egypt during the first week of life

Authors


Corresponding Author Gary L. Darmstadt, Department of International Health E8153, Bloomberg School of Public Health, Johns Hopkins University, 615 North Wolfe Street, Baltimore, MD 21205, USA. Tel.: +1 (443) 287-3003; Fax: (410) 614-1419; E-mail: gdarmsta@jhsph.edu.

Summary

Objective  To provide information about home care practices for newborns in rural Egypt, in order to improve neonatal home care through preventive measures and prompt recognition of danger signs.

Method  Survey of newborn home care practices during the first week of life in 217 households in three rural Egyptian Governorates.

Results  Many practices met common neonatal care standards, particularly prompt initial breastfeeding, feeding of colostrum and continued breastfeeding, and most bathing practices. However, several practices could be modified to improve neonatal care and survival. Supplemental substances were given to 44% of newborns as pre-lacteal feeds, and to more than half during the first week. Nearly half (43%) of mothers reported that they did not wash their hands before neonatal care, and only 7% washed hands after diaper changes. Thermal control was not practiced, although mothers perceived 22% of newborns to be hypothermic.

Conclusions  The practices we observed, which are critical for newborn survival, could be improved with minor modifications. We provide a framework for communicating behaviour change and setting research priorities for improving neonatal health.

Abstract

Objectif  Procurer des informations sur les pratiques en matière de soins à domicile pour les nouveaux-nés en zone rurale en Egypte, afin d'améliorer les soins néonataux à domicile par des mesures préventives et susciter l'identification des signes de danger.

Méthode  Enquête sur les pratiques des soins à domicile pour les nouveaux-nés au cours de la première semaine de vie dans 217 ménages dans 3 Gouvernorats ruraux égyptiens.

Résultats  De nombreuses pratiques répondaient aux normes communes de soins néonataux, en particulier l'allaitement maternel initial, la nutrition par le colostrum et l'allaitement subséquent ainsi que la plupart des pratiques de bain. Cependant, plusieurs pratiques pourraient être modifiées afin d'améliorer les soins et la survie des nouveaux-nés. Des substances supplémentaires étaient données à 44% des nouveaux-nés comme aliments pré lactaire et à plus de la moitié d'eux au cours de la première semaine. Presque la moitié (43%) des mères ont rapporté ne pas se laver les mains avant les soins néonataux et seulement 7% lavaient leurs mains après le changement des couches. Le contrôle thermique n’était pas pratiqué bien que les mères aient constaté 22% de nouveaux-nés hypothermiques.

Conclusions  Les pratiques que nous avons observées, critiques pour la survie du nouveau-né, pourraient être améliorées par des modifications mineures. Nous fournissons ici un cadre pour de communication pour le changement de comportement et pour l’établissement de priorités de recherche afin d'améliorer la santé néonatale.

Abstract

Objetivo  Proveer información sobre los cuidados ofrecidos a recién nacidos en Egipto rural, con el fin de mejorar el cuidado neonatal en los hogares, mediante medidas preventivas y reconocimiento temprano de señales de peligro.

Método  Encuesta sobre los cuidados neonatales en casa durante la primera semana de vida en 217 hogares en 3 provincias rurales egipcios.

Resultados  Muchas prácticas tenían estándares comunes de cuidados neonatales, en particular el comienzo temprano de la lactancia materna, la alimentación con calostro y el mantener la lactancia materna, así como la mayoría de las prácticas relacionadas con el baño. Sin embargo, algunas prácticas podrían ser modificadas para mejorar los cuidados y la supervivencia neonatal. Se daban sustancias suplementarias a un 44% de los recién nacidos como alimentos pre-lácteos, y a más de la mitad durante la primera semana de vida. Casi la mitad de las madres (43%) reportaron que no se lavaban las manos antes de atender al recién nacido, y solo un 7% se lavaba las manos después de cambiar pañales. El control de la temperatura no se practicaba, aunque las madres percibían que un 22% de los recién nacidos estaban hipotérmicos.

Conclusiones  Las prácticas observadas, todas ellas críticas para la supervivencia de los neonatos, podrían mejorarse con modificaciones menores. Proveemos un marco para comunicar los cambios en comportamiento y fijar prioridades de investigación que mejoren la salud neonatal.

Introduction

Neonatal mortality — death in the first 28 days of life — accounts for approximately two-thirds of all infant deaths worldwide and nearly 40% of deaths of children under 5 (Black & Kelley 1999; Saving Newborn Lives 2001; Lawn et al. 2005). Three quarters of neonatal deaths occur in the first week, highlighting the need for early care. Approximately 99% of the four million annual neonatal deaths occur in developing countries, approximately half at home, often outside of the formal health system (Black & Kelley 1999; Lawn et al. 2005). Available estimates suggest that 36% of newborn deaths are due to serious infections (e.g. tetanus, pneumonia, septicaemia and diarrhoea), 28% are due to complications of pre-maturity, 23% are from birth asphyxia and another 7% are attributed to major congenital anomalies (Lawn et al. 2005). Approximately two-thirds of all neonatal deaths worldwide are estimated to be preventable based on efficacy and effectiveness of available interventions (Bhutta et al. 2005; Darmstadt et al. 2005a).

In Egypt, Demographic and Health Survey (DHS) data (El-Zanaty & Macro International 2006) suggest that neonatal deaths comprise half of all infant deaths. In Minya and Qaliubia, Egypt, a survey of 7000 households revealed that 76% of neonatal deaths occurred during the first week of life (Stanton & Langsten 1998). Most neonatal deaths occurred at home, where more than half of all births in rural Egypt occur, often attended by a daya (traditional birth attendant) or relative. Care was sought for only half who died, and for only 25% who died during the first week of life. Moreover, most mothers never recognized signs of illness or recognized symptoms less than 1 day prior to death. By the time neonatal illness is recognized, morbidity may progress to death so rapidly that treatment is ineffective. In a World Health Organization (WHO) community-based study of neonatal sepsis, 50% of deaths occurred within 2 day of presentation to the health facility (Mulholland 1998).

While an increasing proportion of births in rural Egypt occur in facilities, a woman's family — especially her mother — typically provides support and shares traditional knowledge before, during and after birth (El-Nemer et al. 2006). A tradition of sequestering the vulnerable newborn and mother in the home to protect them means that 92% of births at home receive no post-natal care from qualified medical personnel, although many receive home visits from a daya (El-Zanaty & Macro International 2006). Women tend to resume their normal household responsibilities within 3 days after birth, often with help from female relatives (Talaat et al. 2004). While most homes have electricity and access to an improved latrine, few rural households have clean water (El-Zanaty & Macro International 2006).

Improved neonatal home care may substantially improve survival. Evidence suggests that essential newborn care practices, for example related to feeding, hygiene/cord care, thermal control, bathing/skin care and recognition of danger signs are clearly associated with major causes of neonatal mortality, particularly serious neonatal infections (Table 1). Moreover, several studies using a variety of strategies to promote improved home care practices for newborns, post-natal contact with a health care provider and increased care seeking for illness have demonstrated significant reductions in neonatal mortality (Bartlett et al. 1991; Bang et al. 1993, 1999; Manandhar et al. 2004; Darmstadt et al. 2005c; Jokhio et al. 2005).

Table 1.   Targeted newborn home care practices that impact mortality and/or morbidity (Sources: WHO 2003; Beck et al. 2004; Bhutta et al. 2005; Darmstadt et al. 2005a)
Suboptimal newborn home care practices identifiedAssociated mortality and morbidity outcomesOptimal (evidence-based) newborn practices to be promoted
Feeding practices
 Discarding colostrumNeonatal infections (lack of passive antibody transfer)Feeding colostrum
 Delayed initiation of breastfeeding (> 1 hr after birth)Neonatal hypoglycaemiaImmediate breastfeeding (< 1 h after birth)
Neonatal hypothermia
Neonatal infections
 Giving pre-lacteal feedsNeonatal infections (risk of contaminated fluids or foods)Avoiding pre-lacteal feeds; immediate exclusive breastfeeding
 Giving mixed/supplemental non-breastmilk feedsNeonatal infections including diarrhoea, respiratory illness, and transmission of HIV (lack of passive antibody transfer, risk of contaminated fluids or foods)Exclusive breastfeeding on demand for the first 6 months.
Eye care
 Not providing eye care OR Instilling non-antibiotic drops in newborn's eye(s)Ophthalmia neonatorumPrompt administration of eye drops or ointment (silver nitrate 1%, tetracycline 1%, or povidone-iodine 2.5% <1 h after birth)
Thermal control
 Not attending to newborn during 3rd and 4th stages of labourNeonatal hypothermiaImmediate drying and wrapping of the newborn
Birth asphyxia
 Delay > 1 h after birth in initial mother-baby contactNeonatal hypothermiaImmediate (<1 h after birth) skin-to-skin mother-baby contact
Neonatal hypoglycemia
Neonatal infections
Bathing practices
 Immediate bathingNeonatal hypothermiaDelayed bathing (>6 h after birth) with mild non-antimicrobial soap & warm water; not bathing low birthweight babies
Neonatal infections (compromised skin barrier function)
Respiratory distress and destabilized vital signs
 Frequent bathingNeonatal skin infections (removal of vernix, compromised skin barrier function)Removal of blood or meconium with warm water and, if excessive soiling, mild non-antimicrobial soap
Neonatal hypothermia
 Lack of caregiver handwashing before/after handling baby or handwashing with water aloneNeonatal infectionsConsistent caregiver handwashing with soap and water before handling baby and after diaper changes
Umbilical cord care
 No umbilical cord care or applying non-antiseptic substances to cordUmbilical cord colonization and infectionApplying either nothing or chlorhexidine to umbilical cord stump (washing with soap and water only if soiled)
Neonatal sepsis
Neonatal tetanus
 Covering umbilical stump with diaperUmbilical cord colonization and infectionFolding back diaper to facilitate stump drying, which reduces vulnerability to stump infection.
Neonatal sepsis
Neonatal tetanus
Skin care
 Not providing skin care or Using toxic skin care productsNeonatal skin infections (compromised skin barrier function)Application of non-toxic topical emollient (e.g. sunflower seed oil) via massage
Skin inflammation or wounds
Neonatal sepsis

Given this evidence, a crucial strategy to reduce neonatal mortality in Egypt and other developing countries is to improve delivery and early post-partum home care of newborns, emphasizing preventive measures and early danger sign recognition and management. However, no comprehensive data exist on Egyptian caregivers’ practices in the home during the first week of life. Five studies conducted in Egypt have surveyed health care provider practices and/or determined factors contributing to maternal mortality, but none focused on neonatal care or advice given to mothers (SPAAC 1989, 1998; Child Survival Project 1994; Langsten 1998; UNICEF 1999). Anthropological studies reference several potentially harmful traditional provider practices, such as administering pre-lacteal feeds including intinctions of herbs to protect babies from colic and diarrhoea (Tekce 1982; Haddad 1999), and application of lead-containing kohl to the newborn's eyes or umbilical stump to ward off ‘evil eye’ and prevent infections (Kennedy 1967; Inhorn 1994).

Recognizing this research gap, the Healthy Mother/Healthy Child Results Package (HM/HCRP) of John Snow Incorporated (JSI) aimed to reduce neonatal mortality by 15% in target Governorates by improving prevention and management of newborn illnesses, including home care. The Maternal and Child Health (MCH) Care Services of the HM/HCRP aimed to help communities prevent neonatal sepsis; prevent and/or manage neonatal hypothermia; manage birth asphyxia; promote early and exclusive breastfeeding; and recognize and treat newborn illness at home. Prior to introducing a home-based package of essential neonatal care practices in the programme area, we identified home care practices for Egyptian neonates during the first week of life in Luxor, Aswan and Fayoum Governorates, where the maternal and child health programme of the HM/HCRP was being implemented. Study findings were intended to shape behaviour change communication strategies to improve home care practices and local knowledge about neonatal care. Objectives of the study were to identify: (1) providers of home care; (2) routine neonatal care practices, including feeding, hygiene/cord care, thermal care, bathing/skin care, and recognition of danger signs and (3) suggested modifications to care practices to improve newborn health and survival. Local and global implications of these findings are discussed.

Methods

Study site and population

The study sites were rural areas in the Governorates of Aswan (Upper Egypt), Luxor (Upper Egypt) and Fayoum (Lower Egypt), where the HM/HCRP of JSI was being implemented. Rural areas in Egypt have a much less developed health and public infrastructure than urban Egypt and lower values on indicators such as employment, household durable goods, educational attainment and health status (El-Zanaty & Macro International 2006). Rural Upper Egyptian governorates have lower household wealth, health status and educational indices than rural Lower Egyptian governorates (El-Zanaty & Macro International 2006); accordingly, Aswan and Luxor had poorer access to care and less developed infrastructure than Fayoum (Table 2). Perinatal mortality is comparable in rural areas in both Upper and Lower Egypt (∼37/1000 births), but is 33% higher in rural than in urban areas (El-Zanaty & Macro International 2006). All three study regions were comparable in terms of religion and language, but Aswan had a more diverse population including resettled ethnic Nubians.

Table 2.   Comparison of Fayoum, Aswan, and Luxor governorates
 Lower EgyptUpper Egypt
FayoumAswanLuxor
  1. Source: Egypt DHS 2005 (El-Zanaty & Macro International 2006).

  2. †Specific figures unavailable for individual governorates; data presented here are figures for rural Lower Egypt and rural Upper Egypt as applicable.

Development indicators
 Access to improved sanitation†92%80%80%
 Households with dirt or sand floor†7%35%35%
 Households owning any mechanized transport†5.3%3.9%3.9%
 Literacy†56%38%38%
 Ever attended school (females)†69%55%55%
 Overall infrastructureModerately poorPoorPoor
 Links to tourist economyNoneSomeSome
 Population densityModerateLowLow
Health indicators
 Full EPI vaccination coverage at 12 months†91%86%86%
 Ethnic majorityPredominantly ArabArab and NubianPredominantly Arab
 Access to careModerate (proximity to Cairo, availability of health facilities)Poor (low ratio of providers to population, poor transport, and high distances to facilities)Poor (low ratio of providers to population, poor transport, and high distances to facilities)
 Received any antenatal care†74%51%51%
 Received regular antenatal care (≥ 4 visits)†62%37%37%
 Availability of skilled care at birthFacilities and homesFacilities onlyFacilities only
 Traditional birth attendants (TBAs)More commonFewFew

These particular programme sites were selected by JSI based on availability of trained community workers (raedat) who were assigned to local households. The entire programme area was used in this study. The raedat monitored all births and deaths among the households in the programme area. All recently delivered women who resided in the entire area served by the raedat, and had delivered a living or dead newborn ≤ 7 days before the interview, were asked about home care provided to the newborn since birth. All 217 women meeting these study criteria were identified and interviewed by 70 raedat in February 2000. Permission for the survey was granted by the funding agency (United States Agency for International Development, USAID), the programme implementers (JSI), and the Egyptian government.

Questionnaire and interviewers

An initial questionnaire was devised in English by the investigators and reviewed by two independent consultants, the Director of Health of the Fayoum Governorate, and his coworkers. A relevant subset of questions was selected from the comprehensive questionnaire, adapted and translated into colloquial Arabic for field administration. Face and content validity of the questionnaire were achieved through internal validating questions, back-translating the translated survey, pilot testing in the field and reviewing and adapting the survey with raedat during training to assure they knew how to ask the questions and that the questions conveyed the intended meaning. The questionnaire was revised as recommended by the pilot testers and raedat. To minimize recall bias with events specific to the first week of life, mothers were interviewed ≤ 7 day after delivery. In addition to recording mothers’ responses, the raedat recorded answers of other household members who had been present during delivery and the early neonatal period.

Training of raedat

The raedat, some of whom were nurses, had at least some high school education; all had prior experience in conducting home-based interviews. A 1-day training was conducted on-site on study objectives, identification and recruitment of eligible women and study relevance of each question. Each raedat then identified all eligible women in their area, obtained consent to conduct the interview and re-visited identified households within a week of delivery to administer the questionnaire. The majority of raedat interviewed 2 or 3 women, although occasionally, up to 5 women were interviewed by the same raedat.

Data collection and management

Senior raedat in Fayoum supervised data collection, and the Egyptian Assistant Director for MCH Care Services monitored quality control. In Luxor and Aswan, peer-to-peer review of the questionnaire by the raedat ensured data quality. Pairs of supervisors selected during training sessions facilitated prompt case review. To aid identification and for peer verification of data, the mother's name, interviewer, interview date and address were available on each questionnaire, but were kept confidential after initial review for data quality. None of the mothers refused the interview; each took approximately 1–2 h to administer. Collected data were manually validated using validating items in the questionnaire prior to analysis; data presented below are from valid items. Percentages were based on the total number who responded to each question. Where sample size permitted, chi-squared tests of proportions were conducted to determine significance of regional differences.

Results

Study population

Most of the study participants (77%, 166/217) were aged 20–34 years. Only 8% (18/217) were < 20, and 15% (33/217) were ≥ 35 years. Sixty percent of women were from Fayoum (130/217), 24% from Luxor and 16% from Aswan.

Care providers

A primary caregiver was defined as the person responsible for the majority of tasks associated with providing for the newborn during the first week, including feeding, diaper and clothing changes, bathing, provision of warmth, skin care and umbilical cord care. This was usually the mother, but the mother's mother also occasionally served as the primary caregiver, particularly in poorer governorates of Lower Egypt. Rarely, other family members were identified as primary caregivers. The mother, mother's mother, and daya were equally likely to give specialized care (such as eye care), but specialized care was less common in the poorer governorates of Lower Egypt.

Feeding practices

Nearly half of mothers breastfed immediately, and another third initiated breastfeeding shortly thereafter; however, 8% of neonates did not breastfeed until the second day. Nearly all received colostrum, and pre-lacteals were commonly administered. One-third of women gave sugar water or a combination of anise and caraway as pre-lacteal feeds. Eighty percent of mothers reported that they exclusively breastfed, but some mothers administered supplemental feeds, commonly plain water or anise and/or caraway. Not breastfeeding was rare.

Thermal control

After delivery, the baby was rarely held in skin-to-skin contact with the mother (8%), and in approximately half of cases, more than 1 h elapsed before initial mother–baby contact, usually because of maternal problems. One-fifth of mothers thought their baby felt cold after birth; however, no newborn temperature data were available. There was very little variation between study sites in thermal care practices.

Bathing practices

Most babies were bathed during the first week. Usually, babies were bathed once during the first week, with either soap and water or water alone. Few neonates were bathed immediately, but most were bathed on the first or second day.

Hygienic practices

Handwashing was not routine. Despite better infrastructure, educational indicators, and access to care than in the other governorates, caregivers in Fayoum reported significantly fewer good hygiene practices (Table 3). Overall, fewer than 10% of mothers washed their hands after diaper changes or before feedings. Most babies’ clothes were changed routinely after baths, but most mothers also changed their baby's clothes independent of bathing, usually daily. All babies wore diapers, usually changed when soiled.

Table 3.   Key newborn care practices in the rural Egyptian programme area
 Fayoum (n = 130)Luxor (n = 53)Aswan (n = 34)Overall (n = 217)
  1. Respondent totals differ based on number responding to the question. Values within parentheses are expressed in percent.

  2. *Statistically significant difference between study areas (P < 0.001).

Care provider
 General newborn care
  Mother106/129 (82)39/64 (61)25/49 (51)170/242* (70)
  Mother's mother14/129 (11)17/64 (27)16/49 (33)47/242 (19)
  Mother-in-law6/129 (5)4/64 (6)2/49 (4)12/242 (5)
  Daya0/129 (0)1/64 (2)0/49 (0)1/242 (0.4)
  Other3/129 (2)3/64 (5)6/49 (12)12/242 (5)
 Specialized care (e.g. eye care)
  Mother1/18 (6)3/4 (75)2/3 (67)6/25 (24)
  Mother's mother6/18 (33)1/4 (25)0/3 (0)7/25 (28)
  Mother-in-law2/18 (11)0/4 (0)0/3 (0)4/25 (16)
  Daya6/18 (33)0/4 (0)0/3 (0)6/25 (24)
  Other3/18 (17)0/4 (0)1/3 (33)4/25 (16)
Feeding practices
 Breastfeeding practices
  Fed colostrum112/115 (97)46/51 (90)27/33 (82)185/199 (93)
  Breastfed within 1 h112/123 (91)38/52 (73)23/34 (68)173/209* (83)
  Breastfed within 12 h120/123 (98)50/52 (96)34/34 (100)204/209 (98)
  Exclusively breastfed first 7 days113/125 (90)37/52 (71)16/34 (47)166/211* (79)
  Administered pre-lacteal feed(s)49/129 (38)30/50 (60)10/33 (30)89/212 (42)
  Sugar water32/49 (65)10/30 (33)3/10 (30)45/89 (51)
  Anise & caraway4/49 (8)16/30 (53)3/10 (30)33/89 (37)
  Other13/49 (27)4/30 (13)4/10 (40)21/89 (24)
  Administered supplemental feeds42/130 (32)40/53 (75)29/34 (85)111/217* (51)
  Water17/42 (40)0/40 (0)0/29 (0)17/111 (15)
Sugar water4/42 (10)3/40 (8)2/29 (7)9/111 (8)
  Anise7/42 (17)18/40 (45)13/29 (45)38/111* (34)
  Caraway0/42 (0)15/40 (38)10/29 (34)25/111 (23)
  Formula0/42 (0)1/40 (3)14/29 (48)5/111 (5)
  Other14/42 (33)3/40 (8)0/29 (0)17/111 (15)
Thermal control
 Mother-baby skin-to-skin contact9/119 (8)5/52 (10)3/33 (9)17/204 (8)
 Dried/wrapped promptly116/118 (98)48/52 (92)34/34 (100)202/208 (97)
 Delay > 1 h prior to initial mother-baby contact30/130 (23)12/53 (23)12/34 (35)126/217 (58)
 Mother thought baby felt cold24/106 (23)9/50 (18)8/32 (25)41/208 (20)
Bathing practices
 Cleansing agent
  Soap & water19/47 (40)34/44 (77)10/23 (43)63/114* (55)
  Water alone28/47 (60)7/44 (16)10/23 (43)45/114* (39)
 When bathed
  Immediately22/43 (51)7/41 (17)3/21 (14)32/105* (30)
  First day26/43 (60)10/41 (24)10/21 (48)46/105 (44)
  Second day8/43 (19)18/41 (44)4/21 (19)30/105 (29)
  Within first 7 days34/121 (28)37/51 (73)20/34 (59)91/206* (44)
  More than once in first 7 days3/33 (10)12/36 (33)8/19 (42)23/88 (26)
Hygienic practices
 Washed hands before handling baby
  Never71/129 (55)18/56 (32)8/35 (23)97/220* (44)
  Sometimes32/129 (25)19/56 (34)11/35 (35)62/220 (28)
  Always6/129 (5)4/56 (7)9/35 (26)19/220* (9)
  Before feeding9/129 (7)8/56 (14)3/35 (9)20/220 (9)
  After changing5/129 (4)7/56 (13)3/35 (9)15/220 (7)
 Baby's clothes changed after bathing44/55 (80)41/44 (93)22/22 (100)107/121 (88)
 Baby's clothes changed daily66/91 (73)23/30 (77)28/30 (93)117/151 (77)
 Baby wore diapers130/130 (100)53/53 (100)34/34 (100)217/217 (100)
 Diapers changed when soiled96/114 (84)45/48 (94)30/31 (97)171/193 (89)
Umbilical cord care
 Product applied to umbilical cord85/113 (75)34/48 (71)27/34 (79)146/195 (75)
  Alcohol69/85 (81)10/34 (29)7/27 (26)86/146* (59)
  Antiseptic other than alcohol2/85 (2)7/34 (21)6/27 (22)15/146* (10)
  Kohl4/85 (5)14/34 (41)2/27 (7)20/146* (14)
 Diaper covered umbilical stump83/112 (74)15/52 (29)19/32 (59)117/196* (60)
Skin care and management
 Skin care product applied90/130 (69)27/53 (51)32/34 (94)149/217* (69)
 Product used 1–3 times daily66/85 (78)22/28 (79)29/32 (91)117/145 (81)
 Day skin product applied
  First day31/90 (34)3/27 (11)11/32 (34)45/149 (30)
  Second day42/90 (47)12/27 (44)11/32 (34)65/149 (44)
  After second day17/90 (19)12/27 (44)10/32 (32)39/149 (26)
 Reason skin product applied
  Improve skin condition17/85 (20)12/32 (38)0/43 (0)29/160* (18)
  Prevent cold0/85 (0)2/32 (6)2/45 (4)4/160 (3)
  Prevent skin infection48/85 (56)13/32 (41)21/45 (47)82/160 (51)
  Smooth the skin11/85 (13)2/32 (6)11/45 (24)24/160 (15)
  Make the baby warm0/85 (0)1/32 (3)1/45 (2)2/160 (1)
  Custom/habit6/85 (7)1/32 (3)7/45 (16)14/160 (9)
 Skin complications noted
  Skin inflammation or wounds8/119 (7)8/51 (16)5/32 (16)21/202 (10)
  Diaper rash7/118 (6)12/49 (24)8/32 (25)27/199* (14)
 Management of skin complications
  Applied product to affected area2/7 (29)3/12 (25)5/9 (56)10/28 (36)
  Changed diaper more frequently5/7 (71)9/12 (75)4/9 (44)18/28 (64)
 Willing to use prescription skin care product96/117 (82)46/51 (90)30/32 (94)172/200 (86)
Eye care
 Eye drops given14/102 (14)7/43 (16)2/29 (7)23/174 (13)
  For redness2/13 (15)0/3 (0)1/2 (50)6/18 (33)
  For pain5/13 (38)1/3 (33)0/2 (0)3/18 (17)
 Frequency of eye drop administration
  1 × /day10/14 (71)4/4 (100)1/2 (50)15/20 (75)
  2 × /day1/14 (7)0/4 (0)1/2 (50)2/20 (10)
  > 2 × /day3/14 (21)0/4 (0)0/2 (0)3/20 (15)
Recognition of neonatal problems or danger signs
 Newborn died3/130 (2)0/50 (0)0/34 (0)3/214 (1)
 Newborn was ill5/130 (4)5/50 (10)0/34 (0)10/214 (5)
 Any abnormality7/109 (6)5/47 (11)7/33 (21)19/189 (10)
 Observed problems/danger signs
  Congenital anomaly0/98 (0)2/50 (4)0/33 (0)2/181 (1)
  Abnormal urination16/107 (15)20/49 (49)6/34 (18)42/190 (22)
  Vomiting17/126 (13)9/46 (20)8/33 (24)34/205 (17)
  Dehydration20/124 (16)3/49 (6)2/32 (6)25/205 (12)
  Decreased crying13/128 (10)7/50 (14)4/32 (13)24/210 (11)
  Difficulty breathing2/128 (2)20/52 (38)2/34 (6)24/214* (11)
  Umbilical oozing7/128 (5)8/47 (17)6/34 (18)21/209 (10)
  Low temperature12/129 (9)4/52 (8)4/34 (12)20/215 (9)
  Diarrhoea5/124 (4)9/50 (18)5/34 (15)19/208 (9)
  Eye discharge10/128 (8)2/51 (4)2/34 (6)14/213 (7)
  Eye redness8/128 (6)2/51 (4)1/34 (3)11/213 (5)
  Fever4/129 (3)6/54 (11)2/32 (6)12/215 (6)
  Skin color changes8/125 (6)4/51 (8)1/34 (3)13/210 (6)
  Cough5/126 (4)3/49 (6)2/33 (6)10/208 (5)
  Abdominal distension7/126 (6)2/49 (4)1/34 (3)10/209 (5)
  Umbilical stump tenderness5/121 (4)4/13 (31)1/8 (13)10/142 (7)
  Umbilical stump redness7/126 (6)2/11 (18)0/8 (0)9/145 (6)
  Umbilical stump bleeding3/128 (2)2/11 (18)1/8 (13)6/147 (4)
  Impetigo3/128 (2)2/50 (4)1/33 (3)6/211 (3)
  Disseminated rash4/129 (3)2/49 (4)0/34 (0)6/206 (3)

Umbilical cord care

Products were applied to the umbilical cord of three-quarters of newborns during the first week of life. Antiseptic use was common; alcohol was the antiseptic of choice (80% of antiseptic applications). Kohl (local eyeliner) was also used on the cord occasionally (9% of neonates). The diaper frequently covered the umbilical stump.

Skin care and management

Skin care product use was common (69%), usually on the first or second day of life and primarily to prevent skin infections (Table 3). Products were usually applied 1–3 times daily. Most mothers (86%) expressed willingness to apply a prescription skin oil or emollient to prevent infections. Localized skin injury/inflammation and diaper rash were relatively common, particularly in the poorer governorates of Lower Egypt, and were managed by frequent diaper changes or with medication.

Eye care

Eye drops were given to 11% of newborns, generally once daily for redness or maternal perception of pain.

Recognition of neonatal problems or danger signs

Early neonatal mortality was low (1%); all deaths occurred in Fayoum. Mothers noted illness in 5% of all newborns. Congenital anomalies were rare, but reported exclusively in Luxor, an irregularity warranting further investigation. Prevalence of particular symptoms of illness suggests that gastrointestinal illness was relatively common: vomiting, dehydration and ‘abnormal urination’ were reported by > 10% of mothers, and diarrhoea by 9%. Signs of neonatal septicaemia and/or pneumonia (often indistinguishable), including vomiting, fever/hypothermia, decreased crying, distended abdomen, difficulty breathing, cough, tachypnea, umbilical tenderness/redness or low vitality, were described in 4–10% of newborns. These more objective signs may more accurately indicate disease. Signs of meningitis due to sepsis (e.g. convulsions) or tetanus (e.g. stiffness/rigidity) were rare in this population.

Regional variation

In Fayoum, babies were more statistically likely to be cared for by the mother and breastfed immediately and exclusively (Table 2). Along with these positive practices, reported handwashing practices were statistically significantly worse in Fayoum, and better in Aswan. There was significant regional diversity in the products applied to the umbilical cord; kohl was favoured in Luxor versus alcohol in Fayoum (where the stump was also usually covered by the diaper), and antiseptic use was common in both Luxor and Aswan. Skin care product use was especially frequent in Aswan.

Discussion

Caregiver

Maternal death is a strong predictor of infant death, underlining the importance of the mother as caregiver, although this can be mitigated if mothers have support (WHO 1994). Throughout the programme area, the mother was the principal home caregiver for neonates, but this was less common in the poorer governorates of Lower Egypt. Our results indicated participation of other family members in various neonatal care activities (e.g. bathing, changing clothes, eye care), as also found in Bangladesh (Winch et al. 2005). Interestingly, the daya was involved in one-quarter of eye care cases, perhaps seen as having specialized expertise. More data is needed on specific activities of alternate caregivers, and treatments caregivers hesitate to give at home, as this would inform potential behaviour change communication and case management strategies. While the mother is the single most important caregiver, behaviour change communications are best targeted to a wider range of caregivers.

Feeding

The WHO recommends immediate breastfeeding, but fewer than half of neonates here were immediately breastfed (WHO 2002). Nevertheless, this rate exceeded that reported by Langsten (1998) (33%), and was similar to rates reported in the 2005 Egyptian DHS survey (43%) and a JSI Rapid Household Survey in Luxor (32–71%) (unpublished data 1999). Breastfeeding delay of > 24 h occurred in < 20% of neonates, fairly low compared with South Asia, where early breastfeeding rates are often < 10% (Black & Kelley 1999; Saving Newborn Lives 2001). While exclusive breastfeeding rates were higher than the reported in the 2005 DHS (66%), their data encompassed a different timespan (infants < 2 months old vs. < 7 days old).

Non-exclusive breastfeeding heightens the risk of late neonatal death (Table 1) (Victora et al. 1987; Wilson et al. 1998; Cesar et al. 1999; Coutsoudis et al. 1999; Huffman et al. 1999, 2001; Leach et al. 1999; Perera et al. 1999; Raisler et al. 1999). Definitions of exclusive breastfeeding appear to vary, as some Egyptian mothers considered breastfeeding plus regular or irregular complementary feeding as being exclusive (Hakim & El-Ashmawey 1992). The number of women reporting the administration of specific supplementary feedings (n = 111) plus the number of women who reported exclusive breastfeeding (n = 166) exceeded the total number of mothers interviewed (n = 217), which suggests some women believed that they were exclusively breastfeeding despite administering supplements. This is a common perception in developing countries, and is not particular to this study, as pre-lacteal and supplemental feeds may be viewed as having medicinal or ritual value and are not considered food items (Davies-Adetugbo 1997; Semega-Janneh et al. 2001; Fikree et al. 2005). Pre-lacteal feeds have been documented by other Egyptian studies; some report feeding of colostrum afterward, while others suggest mothers believe colostrum is inappropriate (Harrison et al. 1993; Langsten 1998).

Early and exclusive breastfeeding has been increased through maternal education in hospitals as well as home visits (Lutter et al. 1997; Morrow et al. 1999; Haider et al. 2000). Trained peer counsellors may be particularly beneficial to support mothers receiving advice from family and community members or even health staff that is contrary to their own wishes to exclusively breastfeed (Haider et al. 2000). These data suggest that educational programmes to promote breastfeeding may be suitable for Egyptian hospitals and communities, using birth attendants, community health workers, and/or peer counsellors.

Thermal control

Delay of 1 h or more between delivery and initial contact between mother and newborn, as reported in approximately half of cases, was consistent with the common lack of initiation of immediate breastfeeding and the presence of hypothermic and hypoglycemic symptoms (Table 1) (Mondlane et al. 1989; Charpak et al. 1996; McCall et al. 2005). Although reasons for delay were not identified, studies from other developing countries suggest that awareness of hypothermia and its basic preventive measures is low among providers, despite the negative impact of hypothermia on neonatal outcome (Dragovich et al. 1997). To reduce infections (Table 1) (Sloan et al. 1994; Conde-Agudelo et al. 2003), an evaluation of thermal control practices, including feasibility and acceptability of post-delivery skin-to-skin care in the home, is warranted, although there are few examples to date documenting experience with introduction of skin to skin care in the community (Quasem et al. 2003; Darmstadt et al. 2006).

Hygiene

The pervasive lack of handwashing, particularly after diaper changes and before feedings, is a marker of poor general hand hygiene (Table 1). Handwashing with an effective cleanser, whether or not it is antibacterial (Curtis & Cairncross 2003; Luby et al. 2005), is a simple and cost-effective way to prevent skin colonization, provided it is not done so frequently as to cause skin irritation (Larson 1999). Chlorhexidine is ideal for hand decontamination, as it kills bacteria for hours after use (Doebbeling et al. 1992). Emollients may reduce bacterial shedding and promote epidermal barrier function, but those containing anionic surfactants can compromise chlorhexidine's effectiveness (Hall et al. 1986; Frantz et al. 1997; Darmstadt & Dinulos 2000). Because of cost and limited availability of chlorhexidine in Egypt, programmes could promote use of a mild, non-antimicrobial soap for soiled caregivers’ hands, followed by emollient use on hands, where available (Larson 1999). Future research should define acceptability and feasibility of caregivers’ use of soap and water and hand emollients.

Umbilical cord care

Hygienic umbilical cord care neonatal mortality (Table 1), but best cord care practices are still debated (Garner et al. 1993; Mullany et al. 2003, 2006a). Recent evidence suggests that chlorhexidine is an effective cord-cleansing antiseptic (WHO 1998; Darmstadt & Dinulos 2000; Mullany et al. 2003, 2006a,b); however, its use was not reported in our study which predated new evidence (Mullany et al. 2006a,b), and only half of newborns had their umbilical cords cleansed, usually with alcohol, as recommended by the Egyptian Ministry of Health and Population. Alcohol treated stumps, however, can be malodorous and heal more slowly (Belfrage et al. 1985; Lacour 1998; Lund et al. 1999). While triple dye (gentian violet, brilliant green and proflavine hemisulfate) is superior to alcohol in controlling cord colonization, and may be preferable where available (Paes & Jones 1987), it is occasionally toxic. Available evidence suggests most antiseptics are no more effective than soap and water or dry cord care (Zupan & Garner 2004); however, recent community-based data from Nepal demonstrates that chlorhexidine is superior to either regimen in reducing cord infection and sepsis (Mullany et al. 2006a), is long-lasting and is safe (Mullany et al. 2006b). Reasons for the use of kohl on the umbilical cord should be investigated.

Bathing

Generally, bathing practices were appropriate, since immediate bathing after birth is unnecessary and places newborns at risk (Table 1) (Henningsson et al. 1981; WHO 1997; Darmstadt & Dinulos 2000). The common use of soap and water in this study for bathing may injure the skin (Morelli & Weston 1987). Soaps, including milder baby soaps and shampoos, can remove epidermal barrier lipids, irritating and drying skin. Bathing with soap is generally unnecessary given the neonate's low sebaceous gland output, and it may increase skin bacterial counts (Davies et al. 1977; Lilly et al. 1979).

Bathing also may hasten the removal of lipid-rich vernix. Although its role is poorly understood, vernix appears to mechanically insulate and protect the skin, optimize hydration, enhance epidermal barrier function and development and minimize transepidermal water loss, and protect the skin from trauma, hypothermia and infection (Joglekar 1980; Okah et al. 1994; Hoath 1997; Bautista et al. 2000; Darmstadt & Dinulos 2000; Visscher et al. 2005). Thus, measures should be taken to preserve the vernix on the skin.

Skin care

Products of unknown toxicity were applied to the skin of most neonates (690, Table 1). Even in the United States, where a study documented that an average of eight commonly available skin care products containing 48 environmental chemicals was used during the first month of life, their toxic potential is often unknown (Cetta et al. 1991). On the contrary, there is evidence, including from Egypt and Bangladesh, to suggest that topical application of skin barrier enhancing oils can prevent neonatal nosocomial infections in pre-term infants (Darmstadt et al. 2004, 2005b). Neonatal massage with skin ointments and oils is common in developing countries (Darmstadt & Saha 2002; Mullany et al. 2005), thus qualitative and toxicity studies of neonatal skin products could facilitate development of behaviour change programmes using non-toxic, locally-available emollients, such as sunflower seed oil. Minor skin problems (injury, inflammation, diaper rash) affected a quarter of newborns during the first week of life, indicating a need for caregiver education on skin care and case management.

Eye care

Eye drops were given to approximately 10% of neonates, usually for ‘red eye’, but the study could not determine the precise cause of the erythema or the outcome of the inflammation. Given the above finding that few mothers reported administering any eye care (i.e. antibiotic prophylaxis for ophthalmia neonatorum) after delivery, future studies might investigate this topic. Some proportion of these infections could be preventable; treatment immediately after birth with antibiotic drops or ointment (Table 1) could prevent unfortunate sequelae (Oriel 1984; Isenberg et al. 1995).

Recognition of neonatal danger signs

The neonatal mortality rate (NMR) in this study was lower than expected (14/1000), based on the results of the 2005 Egyptian DHS showing a NMR of 20/1000, and data from the Minya (32/1000) and Qaliubia (25/1000) Governorates (Stanton & Langsten 1998), but this may be attributable to our relatively small sample size. That all deaths occurred in Fayoum warrants further investigation.

Rates of neonatal morbidity found in this study are similar to rates of serious illness documented by studies in India (Bang et al. 1999) and in Bangladesh (Baqui et al. 2006). Illness rates in our study were lower than in another study in Egypt, where rates of neonatal illnesses seemed to be overestimates (Stanton & Langsten 1998). Mothers correctly associated signs of serious illness with poor health status in their newborns, perhaps due to education about newborn danger signs from birth attendants (Sibley & Sipe 2002), but Egyptian women may overestimate illness in their children and in themselves (Stanton & Langsten 1998); mothers may have difficulty ascertaining and interpreting neonatal symptoms. However, other studies have indicated that caregivers can recognize illness symptoms with simple training (Bartlett et al. 1991; Bang et al. 1993, 1999). In Luxor, a JSI survey revealed wide variation (4–95%) in ability to recognize 4 of 5 danger signs in neonates (unpublished data 1999). Studies assessing accuracy of Egyptian caregivers in identifying serious neonatal illness will aid development of simple algorithms for caregivers and primary health care workers to recognize and manage neonatal illness.

Regional variation and generalizability

Our survey captured significant regional differences highlighting widespread variation in maternal and neonatal care practices. Some differences between governorates may reflect real differences, interviewer bias or an insufficient sample size. In general, a number of practices in Fayoum were statistically significantly different from the poorer governorates of Aswan and Luxor, perhaps attributable to their respective distances from Cairo; whether differences in outcomes (e.g. higher rates of morbidities) are attributable to differences in practices or to disparities in socioeconomic status is difficult to differentiate from these results. Formative research might further investigate qualitative reasons for differences in reported practices, leading to development of effective behaviour change strategies (Parlato et al. 2004; Piwoz 2004). Many of the local practices observed (e.g. administration of pre-lacteal feeds, lack of caregiver handwashing with soap and water, application of kohl or other non-antiseptic substances to the umbilical stump, application of potentially harmful skin care products) are common to many developing countries and strategies for behaviour change could be transferable between countries. However, significant in-country variation cautions against designing one-size-fits-all programmes and policies at the national or international level. Appropriately tailoring behaviour change programmes to suit local understandings, traditions and resource availability is crucial.

Limitations

While interviewing mothers < 7 days after birth minimized recall bias, this also precluded elucidation of care practices during the remainder of the neonatal period. The minor discrepancies in some responses to questions about breastfeeding may have reflected limitations in recall or common misinterpretation of definitions regarding some feeding practices. Although we validated the questionnaire before use, observational studies could be useful to further evaluate accuracy of questionnaire responses. Some studies have observed that respondents tend to over-report ‘good’ behaviours on questionnaires when compared with structured observations (Curtis et al. 1993), while others have found very close correlation between observed and reported behaviours, but caution that observational data is most useful for measuring common/repeatable behaviours (Daniels et al. 1990). Home-based observational newborn care studies are rare, but could aid in assessment of neonatal care practices and ascertainment of changes in process indicators used to monitor intervention impact. While this study had limited geographic representation and sample size, care practices identified were generally consistent with limited previous data from Egypt and other developing countries.

Next steps

The findings of this survey reveal a number of key research gaps (Table 4). High priority research objectives for feeding practices include identifying reasons for giving supplemental feedings (i.e. feedings other than breast milk), and developing educational strategies for promoting immediate and exclusive breastfeeding in health facilities and the home. Thermal control research questions include identifying reasons for delays in post-delivery mother-newborn contact and developing strategies to minimize these delays; feasibility and acceptability studies of skin-to-skin care in the community; and evaluating caregivers’ ability to recognize and manage newborn hypothermia. Newborn hygiene research gaps include identifying barriers to handwashing and umbilical cord antisepsis; identifying reasons non-antiseptic substances such as kohl are applied to the umbilical stump; and devising sensitive strategies to phase out such practices. Improving responses to danger signs in the home could involve developing and evaluating an algorithm for caregivers and community health workers, such as the raedat, to identify seriously ill neonates in the home, and determining care-seeking behaviour when a neonate is ill, including the identification of barriers to seeking and receiving care. Other priority research areas include identification of barriers to the use of chlorhexidine, feasibility and acceptability studies of emollient use on newborn skin. This research agenda, while specific for Egypt, is relevant for much of the developing world.

Table 4.   Identified needs for new and modified neonatal care behaviours in homes in rural Egypt
Behaviour areaIdentified need for introduction of new behaviour or modification of existing behaviour
BreastfeedingInitiation of breastfeeding delayed until the second day of life in nearly one in 10 newborns
Pre-lacteal feeds (e.g. sugar water, anise, caraway) given to half of newborns
Supplemental feeds in addition to breast milk
Thermal controlLack of mother-baby contact for more than 1 h after delivery of half of newborns
Lack of monitoring of newborn body temperature in all cases
HandwashingLack of handwashing before neonatal care in half of cases
Lack of handwashing after diaper changes and before feeding by more than 90% of caregivers, leading to increased risk for transmission of infections
Umbilical cord careCovering the umbilical stump with the diaper in half of newborns
Application of kohl (eyeliner) to the umbilical cord stump of 10% of neonates
Bathing and skin careUse of soap during bathing
Topical application of potentially toxic substances
Eye careLack of treatment of eye infections
Recognition of danger signsNothing known about neonatal illness recognition and care-seeking behaviour in this population

Conclusions

To our knowledge, this report presents the most comprehensive examination of neonatal care practices in Egyptian homes to date, and adds to the scant literature on newborn home care practices in developing countries. This survey provides valuable information to guide more effective resource allocation and implementation of neonatal health care programmes in Egypt. Some care activities, notably prompt initial breastfeeding, colostrum feeding and continued breastfeeding; and most bathing practices were in accord with accepted standards of care (WHO 1994, 2003; Beck et al. 2004). However, several practices could be modified to substantially improve neonatal health (Table 3). These practices are priority areas for further investigation and development of behaviour change intervention strategies in rural Egypt. This process of identifying suboptimal practices and developing appropriately tailored behaviour change strategies can be employed to improve neonatal health not only in Egypt, but more broadly in other developing countries.

Acknowledgements

This study was supported by USAID through the HM/HRCP Grant to John Snow, Inc., Cairo, Egypt. We thank Amira Roess for her contributions to data analysis in the United States.

Ancillary