OBJECTIVE AND DESIGN: To determine the incidence of and possible risk factors for unexpected sudden infant deaths (SID) and severe apparent life-threatening events (S-ALTE) that occurred within 24 hours of birth. This was a monthly epidemiologic survey.
PATIENTS AND METHODS: Throughout 2009, every pediatric department in Germany was asked to report such cases of unexplained SID or S-ALTE in term infants after a good postnatal adaptation (10-minute Apgar score ≥ 8) to the Surveillance Unit for Rare Pediatric Conditions in Germany. The latter has a capture rate of > 95%. S-ALTE was defined as acute cyanosis/pallor and unconsciousness, requiring bagging, intubation and/or cardiac compressions. Hospitals that reported a case were asked to return an anonymized questionnaire and discharge letter as well as the autopsy protocol in SID cases.
RESULTS: Of 43 cases reported, 17 fulfilled entry criteria, yielding an incidence of 2.6 in 100 000 live births. There were 7 deaths (ie, 1.1/100 000); 6 of the 10 S-ALTE infants were neurologically abnormal at discharge. Twelve infants were found lying on their mother's chest or abdomen, or very close to and facing her. Nine events occurred in the first 2 hours after birth; 7, were only noticed by a health professional despite the mother being present and awake.
CONCLUSIONS: SID or S-ALTE may occur in the first 24 hours after birth, particularly within the first 2 hours. Events seem often related to a potentially asphyxiating position. Parents may be too fatigued or otherwise not able to assess their infant's condition correctly. Closer observation during these earliest hours seems warranted.
- sudden infant death
- apparent life-threatening event
- skin-to-skin contact
- sleep position
WHAT'S KNOWN ON THIS SUBJECT:
Sudden infant death and a severe apparent life-threatening event may occur in term newborns within 24 hours of birth.
WHAT THIS STUDY ADDS:
Using a national reporting system, the incidence of sudden infant death and a severe apparent life-threatening event within 24 hours of birth was found to be 2.6 in 100 000 live births in Germany. More than half of these events occurred in the first 2 hours of life.
Sudden infant death (SID) has its peak incidence at 2 to 4 months of age. After risk-reduction campaigns, its incidence has decreased to ∼32 in 100 000 live births, ie, 220 cases per year in Germany in 2008.1 In contrast with sleeping recommendations for older infants, newborns are often treated differently, at least shortly after birth. In many maternity wards, priority is given to maternal-infant bonding to intensify the relationship between the mother and her infant and to facilitate breastfeeding.2 In doing so, the newborn is placed on his mother's chest or abdomen in a prone position with skin-to-skin contact.
We observed 2 severe apparent life-threatening events (S-ALTEs) in our delivery suite in initially well-adapted term newborns while lying prone in skin-to-skin contact with their mother. In an informal inquiry among neighboring hospitals, several similar cases were reported. We thus wondered about the incidence of SID and S-ALTE in newborns within 24 hours of birth, but found only case reports,3,–,6 1 retrospective inquiry,7 2 single center studies8,9 and 1 regional study.10 We thus performed a prospective epidemiologic study to determine the incidence of such events in Germany. We also aimed to generate hypotheses on possible risk factors for such events with the aim better to prevent them in the future.
As part of the Surveillance Unit for Rare Pediatric Conditions in Germany (ESPED; for details, see Ref 11), all pediatric departments in Germany received monthly reporting cards between January 1, 2009, and December 31, 2009, to notify the study center of any case of unexplained SID or S-ALTE that occurred within 24 hours of birth in a term infant (≥37 weeks' gestational age) after a good postnatal adaptation (10-minute Apgar score ≥ 8). A S-ALTE was defined as an acute state of cyanosis or pallor and unconsciousness, which was felt to require bagging, intubation, and/or cardiac compressions. Regular analyses of its capture rates showed that reporting completeness of the ESPED surveillance system consistently exceeds 95%.11
Hospitals that reported a case received an anonymized questionnaire that asked about the birth, the postnatal situation, and the circumstances of the event. In addition, we asked for the infant's hospital discharge letter and the autopsy protocol in case of SID; these documents had to be anonymized before sending them to us. Having collected these data, we first excluded cases not meeting inclusion criteria (eg, those who had only been stimulated).
To determine the incidence of these events, we used data from the National Bureau of Statistics on the number of births in Germany.1 On the basis of the completed questionnaires and other documents on affected newborns, we tried to identify potential risk factors for these events. The study protocol, including an informed consent waiver, was approved by the ethics committee of Tuebingen University Hospital.
Children's hospitals notified us of 43 cases that had occurred in 2009; 17 of them met inclusion criteria. Of these, 7 were deaths; 3 after unsuccessful resuscitation, the other 4 had initially been resuscitated but had intensive care discontinued on day 6 to 52 of life because of severe hypoxic brain damage. Postmortem results, available for all infants who had died, invariably confirmed a diagnosis of SID. Of the 10 survivors, 6 were described as neurologically abnormal on clinical examination at the time of discharge. Head ultrasounds, performed in all infants, did not reveal intracranial hemorrhage or malformation as a potential explanation for the event. Detailed information about pregnancy, birth, and the circumstances of the event for these 17 infants is provided in Tables 1 and 2.
All infants were singletons, and all (except 1 small-for-gestational age newborn) had appropriate weight for gestational age; 7 were male. Fifteen infants had been born via vaginal delivery (2 with vacuum extraction, 1 with forceps). For 13 mothers, the index case had been their first birth. None of the mothers was reported to have consumed alcohol or illicit drugs during pregnancy; 15 were reported as nonsmokers (for the other 2, no information on maternal smoking was provided). Two mothers had received sedatives (tramadol/oxacepam and valerian, respectively) in the last 24 hours before birth, whereas 13 had not; information was missing for the remaining 2. For the 2 infants who had been born via cesarean delivery, epidural analgesia was used in 1 case, whereas information was missing in the other 1. In 9 of 17 infants, the event had occurred within 2 hours of birth; 9 events had occurred in the delivery suite, the remaining in the perinatal ward.
Twelve newborns were found lifeless while lying on their mother's breast/abdomen or very close to and facing her. One infant was found while lying supine next to his mother, whereas 2 were lying supine in their own bed. Two infants became lifeless while being held in their father's arm.
In 7 infants, the event was discovered by a health professional (with the mother not noticing her infant's condition despite her being present and awake); 6 infants were found by their mother, 4 of them after her waking up from a nap. Three events were discovered by the father, and 1 by a maternal roommate.
The median interval between last seeing the child apparently well until finding him lifeless was 15 minutes (range: 1–180 minutes) for the total group and 6 minutes (range: 1–60 minutes) for events that occurred in the first 2 hours of life. In the 9 cases where the infant had last been seen by a health professional (physician, midwife, nurse), the median interval was 15 minutes (range: 3–180). No infant had been on a monitor.
Among the excluded 26 cases, there were 4 preterm infants who met inclusion criteria except for their premature birth, and 3 term infants who had recovered after vigorous stimulation only. In 3 additional cases, the infant had not been found lifeless, but only in very poor condition. Here, an underlying diagnosis (pneumonia, respiratory distress syndrome, and pulmonary hypertension of the newborn, respectively) was identified. In the vast majority of the remaining cases, either the 10-minute Apgar score was <8 or the event occurred beyond 24 hours of age.
Given a reporting rate of >95% and 665 126 live births in Germany in 2009,1,11 the 17 newborns reported here who suffered unexpected death or a S-ALTE on their first day of life correspond to an incidence of at least 2.6 events per 100 000 live births (including 1.1 deaths per 100 000 live births). This incidence likely reflects only the most critical such events, and we assume that the number of cases may have been higher had we also included less critical events, eg, children who recovered with stimulation only.
Our incidence data are comparable to the incidence of SID/S-ALTE in the first hours after birth reported by Branger et al,7 but are based here on the first nationwide epidemiologic study yet published on this issue. Given that 220 SID were reported in Germany in 2008,1 the number of events occurring on the first day of life seems relatively high.
What could be reasons for this? There are some known risk factors for SID, such as prone sleeping and covering of the face, which can coincide at the moment of skin-to-skin contact between mother and newborn, often implemented shortly after birth to improve maternal-infant bonding.12 There are several publications in which similar circumstances are described and supposed as being causally related to S-ALTE/SID occurring on the first day of life, particularly if the infant's nose is pressed against the mother's breast or abdomen, which may lead to acute upper airway obstruction.4,6,8 In our inquiry, 12 of 17 events had occurred in such a potentially asphyxiating position. One contributing factor in this regard could be overweight of the mother. Unfortunately, we did not obtain information on maternal BMI.
Several studies, including ours, report a high proportion of primipara among mothers whose child suffered such an event.6,8,10 This may be related to their lack of experience in positioning their infant and in realizing possibly hazardous situations, or to the fact that current birth rates in western societies are low.
Another contributing factor could be postnatal fatigue of mother and child, potentially resulting in an infant being unable to free himself from an asphyxiating situation, or a mother being incapable of adequately observing her newborn infant. In our study, 9 of 17 events occurred in the first 2 hours after birth, and 5 while the mother was asleep (with death as outcome in 4 of these 5 events). Only 2 mothers had received sedatives in the last 24 hours before delivery.
Almost 50 years ago, Desmond et al13 described the physiologic changes that take place in the first 6 hours after birth in healthy term infants. They observed that the first 2 hours after birth are dominated by an initial wave of sympathetic activity after the stimuli encountered during the delivery process that rapidly dissipates and is followed by a period of diminished responsiveness to external stimuli. We speculate that this diminished responsiveness may, in some cases, extend even to a potentially asphyxiating situation as seen here.
Only 2 events occurred while the newborn was lying supine and in its own bed (which was standing in the mother's room), thus complying with current recommendations for SID prevention. In 1 of these cases, caregivers suspected the event as having been related to gagging, but no vomiting had been observed in this or any other infant.
We do not wish to imply that our observation should be interpreted as suggesting that mother and infant should be separated after birth. Our data only reveal that placing the infant to sleep in his own bed, preferably next to his mother, may be a somewhat safer sleep environment than placing him in close body contact to a fatigued mother.
All of these proposed mechanisms, however, remain speculative without adequate controls. We are currently continuing the study to collect such information.
A large proportion of infants with S-ALTE were neurologically abnormal at discharge, raising the question as to whether their abnormalities were a cause or a result of their event. A potential explanation for both a S-ALTE and a neurologic abnormality would be an intracranial malformation or an intraventricular hemorrhage,14,15 but this was excluded in all infants. Thus, and because all infants had had a good postnatal adaptation and seemed clinically healthy before the event, we consider it more likely that their neurologic abnormalities were secondary to prolonged hypoxia experienced during the S-ALTE.
Should the postnatal surveillance be intensified, especially in the first 2 hours of life to realize such events earlier? In our survey, the median duration between last seeing the infant apparently well and finding him lifeless was relatively short. However, in 8 cases, it was only the mother or father who described the child as being well before the event had occurred, and they may have misinterpreted their infant's condition. We did not inquire whether monitoring of the infants by caregivers had been as usual or had been reduced (because of overwork or other circumstances). Electronic monitoring might have avoided some events, but we feel that such surveillance in thousands of healthy neonates is disproportionate.
Animal experiments indicate that experiences made shortly after birth may have lifelong consequences through experience-dependent chromatin plasticity.16 Thus, supporting a close proximity of mother and infant shortly after birth enables them to experience mutually beneficial physiology. At the same time, however, close surveillance of the infant's condition by experienced staff, although costly, may be necessary to prevent S-ALTE and SID during this important period of mother-infant bonding.
We received information on cases only some weeks or months after the event. Therefore our data may be subject to recall bias. Given the dramatic nature of these events, however, we feel that most caregivers are more likely to have reported the circumstances of these events correctly.
The intimate period of postnatal bonding is important to strengthen the relationship between mother and child and to facilitate breastfeeding,2 but at the same time, the safety of the mother and her infant is paramount.
In the light of our data, we suggest to check a newborn's condition frequently, especially in the first 2 hours of life and in primiparous mothers who are inexperienced regarding safe handling of a newborn. Midwives should be made aware of the possibility of life-threatening events to occur, even in initially well adapted term neonates. Particular attention should be paid to the nose not being occluded by the mother's breast or abdomen, and parents should be informed about this. Similar to advice given in the Reduce the Risk campaigns, the newborn should not lie on or be snugged to the mother while she is sleeping; there should be a possibility to put the newborn in a bed if the mother is too exhausted after birth.
This work was supported by the Reinhold Beitlich Foundation, Tuebingen, Germany.
We thank the physicians, midwives, and nurses who reported these cases.
- Accepted January 3, 2011.
- Address correspondence to Christian F. Poets, MD, Department of Neonatology, Tuebingen University Hospital, Calwerstr. 7, 72076 Tuebingen, Germany. E-mail:
Dr Anette Poets was involved in study design, supervised the data collection and analysis, wrote the first draft of the manuscript, and had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis; Ms Steinfeldt was involved in designing the study and data collection; and Dr Christian Poets initiated and supervised the study and revised the manuscript.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
- SID =
- sudden infant death •
- S-ALTE =
- severe apparent life-threatening event
- 1.↵Statistisches Bundesamt Deutschland [German Office of Statistics]. Available at: www.destatis.de/. Accessed December 22, 2010
- Moore ER,
- Anderson GC,
- Bergman N
- Toker-Maimon O,
- Joseph LJ,
- Bromiker R,
- Schimmel MS
- Peters C,
- Becher JC,
- Lyon AJ,
- Midgley PC
- Hayden CK Jr.,
- Shattuck KE,
- Richardson CJ,
- Ahrendt DK,
- House R,
- Swischuk LE
- Mercuri E,
- Dubowitz L,
- Brown SP,
- Cowan F
- Copyright © 2011 by the American Academy of Pediatrics