Unsafe Sleep Practices and an Analysis of Bedsharing Among Infants Dying Suddenly and Unexpectedly: Results of a Four-Year, Population-Based, Death-Scene Investigation Study of Sudden Infant Death Syndrome and Related Deaths

DISCUSSION

We report sleep practices evident at the death scene in 119 infant deaths from the city of St Louis and St Louis County, Missouri. These deaths occurred after the beginning of the Back-to-Sleep campaign. Twenty-five SIDS victims were found prone with their nose and mouth into underlying bedding.³⁶ The face-down position is associated with a high odds ratio for death (11.2).¹⁰Another 16 of the deaths were attributed to accidental suffocation. We assume that supine sleep on firm sleep surfaces designed for infants could have prevented most of the 41 deaths in these 2 groups that make up our first 2 categories in Table 4^14,15,40 (Figs 1–4).

Five of the deaths attributed to SIDS occurred with the infant's head completely covered by bedding. Like the face-down position, this scenario causes rebreathing of exhaled air^20,41 and significantly increases the adjusted odds ratio for dying to 2.18.⁴ It seems likely that some of these types of unexpected deaths can also be prevented.

Thirty-five infants were found prone with face to side. It is difficult to predict how many of these deaths would have been prevented because a separate population-attributable risk for the prone face-to-side position has not been reported. Regardless of how many infants in this category would have avoided sudden death, there is international agreement that they would have been at reduced risk if they had slept supine.⁴⁰ And there is at least a statistical likelihood that many would not have died, with possible explanations provided by the fact that arousal mechanisms and airway protective reflexes are more robust among supine than among prone infants.^42,43

Our results suggest the extent of the problem of sudden death among infants using unsafe sleep practices in St Louis. Eighty-four percent of the victims in this series (Table 4) had a diagnosis of accidental suffocation, or were prone and face down or to the side, or were found with head covered, or while sharing a sleep surface. For each of these circumstances, a case has been made, or could be made, that the sleep practice in question is causally associated with sudden death. This suggests that the deaths may not have occurred if certain high-risk sleep practices had been avoided, and that the majority of deaths were preventable. Although it is certainly true that infants continue to die suddenly and unexpectedly in the Back-to-Sleep era, it is apparent from this series that only a small minority (8.4%) were found alone in bed, in a nonprone position, with external airway unencumbered. This finding, in particular, highlights the need to continue to emphasize safe sleep practices.⁴⁴

The impact of bedsharing on risk for sudden infant death remains controversial. Three case–control studies suggest that bedsharing increases risk for sudden death,^24,26,45 but the risk is lessened when the high rate of maternal smoking in these studies is considered. In England, in particular, the rate of smoking among mothers whose infants died while bedsharing is so high that the risk for nonsmoking mothers cannot be calculated from the data.²⁴ In the United States, a case–control study²⁷ from Washington, DC showed increased risk especially when black infants bedshare. Finally, preliminary results from the Chicago Infant Mortality Study, a large, recent case–control study, strongly indicate an effect of bedsharing that is independent of smoking.⁴⁶ There are no recent published results addressing risk for infants sleeping alone outside of cribs, but data from the US Consumer Products Safety Commission suggest that the risk may be high, particularly for accidental suffocation, and data substantiating this risk have been presented in a preliminary report.⁴⁷

Strategies in addition to Back-to-Sleep are needed to avoid deaths in sleep environments known to be dangerous. For example, all would agree that every infant deserves access to a safe bed. Nevertheless, designing and manufacturing safe cribs are only part of a strategy, because only one quarter of our victims (29 of 119) were placed to sleep before death on surfaces designed for infant sleep. Only 2 of our cases of accidental suffocation involved infant cribs. More than one half of all deaths in all categories (Table 2) involved adult beds, cushioned chairs, or sofas,²⁴ including 13 of 16 accidental suffocation cases. Strategies, such as moving the infant so his or her feet are at the foot of the bed,¹⁸ will not be effective for the majority of infants in the United States who die outside of cribs. Although separate risk calculations have not been made for use of sleep surfaces other than those meeting standards for infant safety, the case against this practice seems compelling.^{13–17,47,48}

Most deaths on adult beds, chairs, and sofas occurred while the infant was sleeping with another person (68.1%). It is difficult to imagine a shared sleep surface used in the United States that would be as safe as standard cribs, in good repair, in preventing falls and entrapment. Therefore, although it is controversial whether bedsharing per se increases risk for sudden death, there seems little question that the sleep surfaces used by infants dying while bedsharing fail to meet widely recognized standards of safety.^16,17 It is also apparent that shared sleep surfaces themselves are softer and more likely to cause entrapment and, perhaps, rebreathing.⁴⁹Eighteen bedsharing infants were found dead on pillows or comforters, and 6 of 10 cases of entrapment by sleep surfaces were during bedsharing.²⁸ Twenty-two of 56 of deaths (39.3%) while bedsharing fit into 1 of the first 3 categories in Table 4, indicating that factors clearly to be avoided are present in many bedsharing deaths. If bedsharing is to be accepted as a safe sleep practice in the United States, guidelines are urgently needed regarding the firmness of the shared sleep surface, avoidance of pillows and soft comforters, and ways to avoid falls and entrapment of the infant.^16,50Finally, some of the deaths while bedsharing in our series probably involved exposure to tobacco smoke, a factor that greatly increases the risk associated with bedsharing.^26,51

In this series, nearly one quarter of deaths (23.2%) while bedsharing were diagnosed as suffocation or undetermined. If our findings can be generalized, then the potential risk attributable to bedsharing in sudden unexpected deaths should not be assessed from SIDS data alone, but also from cases with closely related diagnoses, eg, accidental suffocation. This is particularly important during times when new mechanisms for death are being discussed, and diagnostic drift may occur. The potential for confusion and diagnostic drift is further evident in other reports. For example, in a large recent study from England, among infants diagnosed as SIDS, 1 victim was found under a parent and there were 4 cases where the infant was “wedged between the parent and the back of the sofa.”²⁴ Why these deaths were diagnosed as SIDS and not accidental suffocation seems arguable and would require more specific information about the death scene.

Many reasons point to the possibility that entrapment by a bedmate is more common than is documented in our study (7 of 56 deaths while sharing a sleep surface). In 5 of 7 cases (71%) another family member discovered the entrapment of the deceased infant. Without a third-party observer, the bedmate may have changed position without being aware that the infant was once beneath part of his or her body. The bedmate may also deny the possibility of entrapment, for obvious reasons, if he or she finds the infant has died. Therefore, it seems plausible that the rate at which entrapment is documented would increase if more infants dying while sharing a sleep surface were discovered by a third party.⁵² Finally, of 29 infants dying supine with nose and mouth uncovered, 19 (65.6%) died while bedsharing, suggesting that bedsharing may lessen the benefits of non prone sleep among high-risk groups.

Sixty-five of the deaths reviewed (categories 2, 3, and 4) occurred in positions and microenvironments in which rebreathing of exhaled air, perhaps with associated thermal stress, is believed to be an important contributory mechanism.^32,53 Another important potential mechanism, namely delayed or blunted airway protective reflexes, is suggested by the 10 deaths occurring with the infant's head completely covered, a finding that causes marked increases in odds ratio for dying.^4,20 At this time, there is only preliminary information describing what infants can do to obtain access to fresh air when their heads and external airways are covered, and the range of skills among young infants and the timing of their appearance during postnatal development are incompletely understood.^41,54Nevertheless, as was the case in studies of prone deaths, risk associated with being found with head covered has been established by epidemiologic studies,^4,20 and potential lethal mechanisms have been explored in studies involving animals, mechanical models,⁵⁵ and human infants.^20,41Epidemiologic and physiologic studies, thus, strongly suggest that avoiding loose bedding, particularly if bulky, could prevent head-covered deaths.

Of the 10 deaths with head covered, 5 were diagnosed as SIDS and 5 suffocation. This points out the strong potential for overlap between these 2 diagnoses in head-covered deaths. To resolve this diagnostic quandary, more must be learned about what infants can and cannot do, on average, when their heads and external airways become covered. It may be that some head-covered deaths are similar to wedging deaths, because the infant cannot use its developing repertoire of behaviors to escape from a suffocating environment.

Increasing supine sleep can prevent some deaths diagnosed as SIDS. By broadening our analysis to related sudden deaths, including those attributed to accidental suffocation and those for which a cause is undetermined, we have shown that most infants with these 3 causes for death either died prone or were using other sleep practices that should be discouraged. Our findings suggest that the Back-to-Sleep message should be intensified and that risks from sleep surfaces other than standard cribs, particularly those used during bedsharing, should be included in future public health messages. Combining Back-to-Sleep with refined messages about safe sleep zones should further reduce the rate of SIDS and related causes of infant death.