Content uploaded by Sarah J M Shaefer
Author content
All content in this area was uploaded by Sarah J M Shaefer
Content may be subject to copyright.
Sudden Infant Death Syndrome in Child Care Settings
Rachel Y. Moon, MD*‡§; Kantilal M. Patel, PhD‡§; and Sarah J. McDermott Shaefer, RN, PhD储¶
ABSTRACT. Background. The incidence of sudden
infant death syndrome (SIDS) in the United States has
decreased with decreased prone sleeping. Extrapolating
from Census Bureau data, ⬃7% of SIDS should occur in
organized child care settings (ie, child care centers or
family child care homes). However, 2 states have re-
ported higher rates of SIDS in child care.
Objectives. To determine the percentage of SIDS
deaths occurring in child care settings, and to ascertain
associated factors.
Design. A retrospective study of SIDS deaths from
January 1995 through June 1997 was conducted. Data
were abstracted from SIDS databases in 11 states. Char-
acteristics of SIDS cases occurring in child care settings,
including sleep position, were compared with those oc-
curring in the care of parents. Univariate and multiple
logistic regression analyses were performed.
Results. A total of 1916 SIDS cases were analyzed for
this study. Of these deaths, 20.4% occurred in child care
settings. Compared with deaths in the care of parents,
those occurring in child care settings were more likely to
occur on weekdays between 8:00
AM and 4:00 PM; infants
were older; not black; and their mothers were more ed-
ucated. Infants in child care were more likely to be found
prone in univariate analysis, but the association was not
significant in multiple logistic regression analysis. How-
ever, in multiple regression analysis, infants in child care
were more likely to be last placed prone or found prone,
when the usual sleep position was side or supine.
Conclusion. A large proportion (20.4%) of SIDS cases
occur in child care settings. Factors associated with SIDS
in child care settings include older age, race, and highly
educated parents. Previous studies have reported that
unaccustomed prone sleeping puts infants at high risk
for SIDS; this characteristic was found to be associated
with SIDS in child care and may partly explain the high
proportion of SIDS cases in child care settings. Parents
must discuss sleep position with any caretakers of their
infants. In addition, further efforts to educate child care
providers about the importance of supine sleep for in-
fants must be ongoing. Pediatrics 2000;106:295–300; sud-
den infant death syndrome, child care, prone position,
sleep position.
ABBREVIATION. SIDS, sudden infant death syndrome.
I
n 1992, the American Academy of Pediatrics pub-
lished a recommendation to place all healthy in-
fants in the supine or side position for sleep,
because the prone infant sleep position is associated
with an increased risk of sudden infant death syn-
drome (SIDS).
1
Because subsequent evidence sug-
gested that side position put infants at higher risk
than supine, this recommendation was modified in
1996 to state that supine was preferred over the side
position for sleep.
2
In 1994, a national public educa-
tion campaign, Back to Sleep, was launched through
a coalition of the US Public Health Service, the Amer-
ican Academy of Pediatrics, the Association of SIDS
and Infant Mortality Programs, and the SIDS Alli-
ance. Since these recommendations were given and
the Back to Sleep campaign was begun, the percent-
age of infants sleeping prone has decreased dramat-
ically. Recent surveys indicate that the percentage of
infants placed prone for sleep by parents decreased
from ⬃70% in 1992 to 24% in 1996.
3–6
Concurrently,
the incidence of SIDS decreased ⬃40% from 1992 to
a rate of .69 per 1000 live births in 1997.
7
In the United States, the number of women with
children ⬍6 years old in the labor force has increased
from 2.9 million in 1960 to 10.3 million in 1996.
8
In
1994, there were 1.7 million infants ⬍1 year of age
with employed mothers.
9
As the number of mothers
in the labor force has increased, so has the number of
children in child care. From 1977 to 1992, the number
of child care centers more than doubled from 25 000
to 51 000.
10
According to the US Census Bureau, in
1994, 17% of US infants ⬍1 year of age were attend-
ing some type of organized child care, approximately
one half in child care centers, and one half in family
child care homes (nonrelatives).
8,9
If one estimates
that infants in child care spend ⬃40% of their time
(10 hours/day or 38 –40 hours/week)
11
in that set-
ting, one can extrapolate that ⬃7% (17% of infants in
child care ⫻ 40% of time) of deaths attributed to SIDS
can be expected to occur in organized child care
settings. However, data from 2 states have suggested
a higher rate of deaths in child care settings than
would be expected.
12
This is concerning, especially in
light of data that suggest that many child care pro-
viders may be placing infants in the prone sleep
position at least some of the time.
12
Because many
From the *Department of General Pediatrics and Adolescent Medicine,
Children’s National Medical Center, Washington, DC; ‡Center for Health
Services and Clinical Research, Children’s Research Institute, Children’s
National Medical Center, Washington, DC; §Department of Pediatrics,
George Washington University School of Medicine and Health Sciences,
Washington, DC; 储Department of Pediatrics, University of Maryland School
of Medicine, Baltimore, MD; and ¶Association of SIDS and Infant Mortality
Programs, Baltimore, MD.
The results from this manuscript were presented in part at the Association
for SIDS and Infant Mortality Programs meeting; March 19, 1999; Bethesda,
MD; the National SIDS Alliance Meeting; April 9, 1999; Atlanta, GA; and the
Ambulatory Pediatric Association meeting; May 1, 1999; San Francisco, CA.
Received for publication Jul 20, 1999; accepted Nov 29, 1999.
Reprint requests to (R.Y.M.) Department of General Pediatrics and Adoles-
cent Medicine, Children’s National Medical Center, 111 Michigan Ave, NW
Washington, DC 20010. E-mail: rmoon@cnmc.org
PEDIATRICS (ISSN 0031 4005). Copyright © 2000 by the American Acad-
emy of Pediatrics.
PEDIATRICS Vol. 106 No. 2 August 2000 295
infants spend a large percentage of their time in
various forms of child care, it is worthwhile to ex-
amine both the prevalence of SIDS in these settings
and the factors relating to these deaths. We hypoth-
esized that a substantial proportion of SIDS cases
occur in child care settings and that infant sleep
position may play a role in SIDS occurring in child
care.
METHODS
We performed a retrospective surveillance study of all deaths
attributed to SIDS from January 1995 through June 1997 in 11
geographically diverse states (Arizona, California, Colorado, Flor-
ida, Maryland, Massachusetts, Minnesota, Michigan, Missouri,
New Hampshire, and New Jersey). The directors or coordinators
of the SIDS programs of these states were members of the Asso-
ciation of SIDS and Infant Mortality Programs. Each state SIDS
center received information from the medical examiner regarding
all SIDS cases reported in that state. For each death, data were
obtained from birth and death certificates and supplemented by
parent interview. Trained SIDS grief counselors or investigators
from the medical examiner/coroner’s office conducted parent in-
terviews within 1 month of the infant’s death. Interviews were
conducted in the infant’s home or by telephone. Data for this
study were abstracted from information previously collected by
the individual states. No new data were collected. Data collected
included birth history, demographic information, smoke expo-
sure, sleep position (usual and last placed), position found, loca-
tion of death, time of death, caretaker at time of death, prenatal
exposures (alcohol, drugs, or tobacco), breastfeeding, medical
problems, and recent changes in the child’s routine. Location of
death was categorized as in a child care setting or not in a child
care setting. Not in child care was defined as under the care of a
parent or guardian. These deaths usually occurred at home, but
some occurred in hotels, the homes of friends, or cars. Child care
settings were divided into: at home with a nanny or babysitter
(relative or nonrelative), in the care of a relative in the relative’s
home, in a family child care home, or in a child care center (Table
1).
13,14
Frequencies of the demographic variables (means ⫾ standard
error of the means) of demographics were tabulated. The outcome
measure was location of death (ie, dying of SIDS in a child care
setting vs dying of SIDS in the care of the parent/guardian). We
performed univariate and multiple logistic regression analyses to
identify demographic variables and factors associated with the
outcome measure.
This study was approved by the institutional review board of
Children’s National Medical Center.
RESULTS
Data were provided by the 11 participating states
on a total of 2315 SIDS cases, representing all of the
SIDS deaths in those states between January 1995
and June 1997. Seventeen percent of the cases (399)
were excluded because the location of death was not
documented, resulting in a sample size of 1916 cases.
The excluded cases were similar to the cases in the
sample with regard to age at death, birth weight,
gestational age, gender, maternal age, and the day of
the week of death. However, the mothers in the
excluded cases were less likely to have a high school
diploma (P ⬍ .001). The total number of SIDS deaths
declined each year of the survey, from 91.25 deaths/
month (1095 total) in 1995 to 71.67/month (860 total)
in 1996, and 59.5/month (357 total) for the first half
of 1997.
Of the 1916 cases, 60% were male and 40% female.
The mean age at death was 83.5 days (range: 2–365
days). A total of 51.8% were classified as white;
29.4%, black; 11.2%, Hispanic; 2.5%, Asian; 2.1%, Na-
tive American; and 1.8%, multiracial or other. Mean
birth weight was 3093 g (range: 600-7569 g), and
mean gestational age was 39 weeks (range: 23– 44
weeks). More than three quarters of the sample
(76.6%) was full-term at birth. Mean maternal age
was 23.4 years (range: 14– 48), and mean paternal age
was 27.4 years (range: 14 –67). Of the mothers, 37.1%
had not completed high school, 35.7% had a high
school diploma, and 27.3% had completed some
postsecondary school education. Nearly one half of
fathers (43%) had a high school diploma, whereas
24.4% had not completed high school and 32.6% had
some postsecondary school education. There was
documentation of prenatal exposure to tobacco in
39.5% of cases, prenatal alcohol exposure in 11.1%,
and prenatal illicit drug exposure in 16.3%. Of the
infants dying of SIDS, 64.7% had been exposed to
household cigarette smoke, and only 18.8% had been
breastfed at any time.
In our sample, 40.1% died during the daytime
hours of 8:00 am to 4:00 pm, 10.5% from 4:01 pm to
midnight, and 28.1% from 12:01 am to 8:00 am. The
time of death was unknown for 21.2%.
In our sample, 20.4% of deaths occurred in various
child care settings, with 1.3% occurring with a nan-
ny/babysitter, 4.3% in a relative’s home, and 14.7%
in organized child care settings (12.1% in family child
care homes and 2.6% in child care centers). The pro-
portion of child care deaths did not increase during
the survey, with 203 of 906 deaths (22.41%) in 1995,
130 of 692 (18.79%) in 1996, and 56 of 315 (17.78%) in
1997. The proportion of child care deaths in the in-
dividual states ranged from 9.4% (Florida) to 40.2%
(Minnesota; Table 2). Approximately 60% of SIDS in
child care occurred in family child care homes (Table
2). The prevalence of prone as the usual sleep posi-
tion was not increased in the child care cohort.
Among the infants in child care, 32.0% usually slept
prone, 37.7% supine, and 29.5% on the side, com-
pared with 40.7% of those not in child care who
usually slept prone, 23.0% supine, and 32.6% on the
side. A total of .8% of those in child care and 3.7% of
those not in child care had no usual sleep position.
When we analyzed demographic characteristics
and other factors in a multiple logistic regression
model for association with location of death (Table
3), we found that, compared with SIDS that did not
occur in child care, SIDS occurring in child care was
more likely to occur on weekdays during the hours
TABLE 1. Definitions for Child Care Arrangements
13,14
Not in child care Under the care of a parent or guardian
(usually at home)
Nanny Under the care of a nanny or babysitter
(relative or nonrelative) in the infant’s
home
Relative home Care of children by a relative in the
relative’s home
Family child care
home
Care of 6 or fewer children in the
caregiver’s home (nonrelative).
Caregiver is often referred to as
family child care provider
Child care center A licensed nonresidential facility
that provides care for ⬎10 children
Organized child care Child care that occurs in a family child
care home or in a child care center
296 SUDDEN INFANT DEATH SYNDROME IN CHILD CARE SETTINGS
of 8:00 am to 4:00 pm, and infants were more likely to
be born at term, older at the time of death, and less
likely to have been exposed to tobacco in utero. Black
infants were underrepresented and white infants
were overrepresented in the group of infants in child
care. Parents of infants dying in child care were more
likely to be better educated than parents of SIDS
victims who died at home.
On univariate analysis, infants in child care were
more likely to be found or last placed prone. How-
ever, when multiple logistic regression was per-
formed, controlling for race, age at death, gestational
age, maternal age and education, paternal education,
prenatal tobacco exposure, and time of death, this
was no longer a significant factor. We did find that
being found (P ⫽ .011) or last placed (P ⫽ .0037)
prone when the usual sleep position was side or
supine was associated with death in child care. Of
the infants who were last placed prone, 59.5% usu-
ally slept supine and 40.5% usually slept on the side.
Of the infants who were found prone, 38.0% usually
slept supine and 62.0% usually slept on the side,
suggesting that many of these infants rolled to the
prone position while asleep. When the usual sleep
position was side or supine, infants in child care
were twice as likely to be found prone and 5 times as
likely to be last placed prone. The mean ages of
infants last placed prone and found prone when the
usual position was side or supine were 96.5 days and
99.8 days, respectively; in the multiple logistic re-
gression model, age at death was not a significant
factor.
Because child care generally occurs during the
hours of 8:00 am and 6:00 pm, we conducted a sep-
arate analysis of the 1017 deaths occurring during
these hours. As expected, a larger percentage in this
subsample (306 or 30.1%) died in the care of a child
care provider (1.3% with a nanny, 4.7% in a relative’s
home, 19.5% in a family child care home, and 4.6% in
a child care center). However, compared with the
entire sample, we found no significant differences in
mean age at death, gender, race, parental age, paren-
tal educational level, or exposure to prenatal and
postnatal tobacco smoke. When multiple logistic re-
gression was performed on this subset, SIDS in child
care was significantly more likely to occur in older
infants (P ⫽ .02), with older (P ⬍ .01), more educated
(P ⬍ .001) mothers. Infants were less likely to be
exposed to secondhand smoke (P ⫽ .01) and less
likely to be black (P ⬍ .01). In this cohort, infants in
child care were much more likely to be found prone
when the usual sleep position was side or supine
(P ⬍ .005). However, they were not more likely to be
last placed prone in this cohort.
For 99 of the infants, information was available
regarding the length of time they had been in their
individual child care situation. In this small sample,
16 (16.2%) died on the first day in child care, and an
additional 18 (18.2%) on days 2 through 7, for a total
of 34 deaths (34.4%) occurring in the first week of
child care. Of the 99 infants, approximately two
thirds of these deaths (63.6%) occurred in family
child care homes. When deaths in the first week of
child care were compared with deaths after the first
TABLE 2. SIDS Deaths, Individual States
State Number
of
SIDS
Deaths
SIDS Deaths
Not in Child
Care
(% ⫾ Standard
Error of the Mean)
SIDS Deaths in Child Care Settings
(% ⫾ Standard Error of the Mean)
Total Nanny Relative
Home
Family Child
Care
Child Care
Center
Arizona 87 68 (78.2% ⫾ 4.4%) 19 (21.8% ⫾ 4.4%) 4 (4.6% ⫾ 2.3%) 3 (3.4% ⫾ 1.9%) 11 (12.6% ⫾ 3.6%) 1 (1.2% ⫾ 1.2%)
California 578 468 (80.9% ⫾ 1.6%) 110 (19.1% ⫾ 1.6%) 7 (1.2% ⫾ .5%) 30 (5.2% ⫾ .9%) 49 (8.5% ⫾ 1.2%) 24 (4.2% ⫾ .8%)
Colorado 156 112 (71.8% ⫾ 3.6%) 44 (28.2% ⫾ 3.6%) 0 (0%) 9 (5.8% ⫾ 1.9%) 29 (18.6% ⫾ 3.1%) 30 (5.2% ⫾ .9%)
Florida 169 153 (90.5% ⫾ 2.3%) 16 (9.4% ⫾ 2.2%) 0 (0%) 10 (5.9% ⫾ 1.8%) 5 (3.0% ⫾ 1.3%) 1 (.6% ⫾ .6%)
Maryland 197 169 (85.8% ⫾ 2.5%) 28 (14.2% ⫾ 2.5%) 3 (1.5% ⫾ .9%) 4 (2.0% ⫾ 1.0%) 20 (10.2% ⫾ 2.2%) 1 (.5% ⫾ .5%)
Massachusetts 100 75 (75% ⫾ 4.3%) 25 (25% ⫾ 4.3%) 1 (1% ⫾ .99%) 5 (5% ⫾ 2.2%) 7 (7% ⫾ 2.6%) 12 (12% ⫾ 3.3%)
Michigan 76 65 (85.5% ⫾ 4.0%) 11 (14.5% ⫾ 4.0%) 2 (2.6% ⫾ 1.8%) 2 (2.6% ⫾ 1.8%) 4 (5.3% ⫾ 2.6%) 3 (4.0% ⫾ 2.3%)
Minnesota 164 98 (59.8% ⫾ 3.8%) 66 (40.2% ⫾ 3.8%) 2 (1.2% ⫾ .9%) 11 (6.7% ⫾ 1.9%) 53 (32.3% ⫾ 3.7%) 0 (0%)
Missouri 197 149 (75.6% ⫾ 3.1%) 48 (24.4% ⫾ 3.1%) 3 (1.5% ⫾ .9%) 6 (3.0% ⫾ 1.2%) 39 (19.9% ⫾ 2.8%) 0 (0%)
New Hampshire 26 22 (84.6% ⫾ 7.1%) 4 (15.4% ⫾ 7.1%) 1 (3.9% ⫾ 3.8%) 0 (0%) 2 (7.7% ⫾ 5.2%) 1 (3.9% ⫾ 3.8%)
New Jersey 166 146 (88.0% ⫾ 2.5%) 20 (12.1% ⫾ 2.5%) 2 (1.2% ⫾ .9%) 3 (1.8% ⫾ 1.0%) 15 (9.0% ⫾ 2.2%) 0 (0%)
Total 1916 1525 (79.6% ⫾ .9%) 391 (20.4% ⫾ .9%) 25 (1.3% ⫾ .3%) 83 (4.3% ⫾ .5%) 234 (12.2% ⫾ .8) 49 (2.6% ⫾ .4%)
ARTICLES 297
week, there was no statistical difference in the loca-
tion of death. The average age for infants who died
on the first day or first week in child care was 106.9
days, compared with the average age of 126.9 days
for infants who died after the first week in child care
(P ⫽ .07). For the 34 who died in the first week in
child care, 28 (82.3%) were found prone; of these, 19
usually slept in the supine or side position.
When we stratified the data by child care type, we
found that the demographic characteristics of infants
who died in a relative’s home and those who died in
the care of their parents/guardians were similar. The
infants who died in the care of a parent, guardian, or
relative were more likely to be black, exposed to
secondhand smoke, and born to younger and less
educated mothers. These characteristics contrasted
with those of infants who died in the care of a nanny
or while in organized child care, who were more
likely to be white, born to older, more educated
mothers, and lacking a history of exposure to house-
hold smoke.
DISCUSSION
In our study, we found that the proportion of
infants dying of SIDS in organized child care settings
was disproportionately high, considering the num-
ber of infants reportedly in child care. This is espe-
cially striking, because the demographic characteris-
tics of this group (white, born to older, more
educated parents, and without a history of smoke
exposure) would typically place these infants in the
lowest risk category for SIDS. These characteristics
can primarily be predicted by the demographics of
working parents who use child care, in that working
parents tend to be older and more educated. In ad-
dition, because most families require child care on
weekdays during typical work hours, it is to be ex-
pected that deaths in child care were associated with
these periods. Indeed, when we looked at the subset
of deaths taking place between the hours of 8:00 am
and 6:00 pm, we found that a larger percentage of
these deaths occurred in child care settings.
We found that the demographic characteristics of
infants cared for by parents were comparable to
those in a relative’s home. US Census Bureau and
other national surveys have reported that low-in-
come and minority families are more likely to rely on
relatives to care for their children than are more
affluent or white families.
9,14,15
It is concerning that 60% of the child care deaths
occurred in family child care settings; family child
care homes accounted for 12.2% of all SIDS deaths, as
opposed to 2.6% occurring in child care centers. Ac-
cording to US Census Bureau data, approximately
one half of infants in organized child care are in
family child care homes, with the other half in child
care centers,
8,9
so one would expect similar numbers
of SIDS deaths. There are differences between family
child care homes and child care centers regarding
licensure, regulation, and care provider characteris-
tics that may be important. As opposed to child care
centers, family child care providers may or may not
TABLE 3. Factors Associated With SIDS in Child Care
No Child Care
(n ⫽ 1525)
Child Care
(n ⫽ 391)
P Value* Odds Ratio
Age at death (d) 91.95 ⫾ 1.56 (2–362) 120.28 ⫾ 3.06 (2–365) .0045 1.005 (1.001–1.009)
Birth weight (g) 2958 ⫾ 21.4 3251 ⫾ 33.3
(425–7569) (1260–4876)
Gestational age (wk) 37.9 ⫾ .1 (23–43) 38.8 ⫾ .1 (24–42) .001 1.15 (1.06–1.25)
Maternal age (y) 24.46 ⫾ .17 (14–45) 27.88 ⫾ .32 (14–48) .090 (NS) 1.03 (.995–1.07)
Maternal education .0035 1.45 (1.13–1.86)
⬍High school 40% ⫾ 1.3% 14.3% ⫾ 1.8%
High school diploma 38.3% ⫾ 1.2% 30.9% ⫾ 2.3%
⬎High school 21.7% ⫾ 1.1% 54.8% ⫾ 2.5%
Paternal age (y) 28.06 ⫾ .24 (14–67) 30.35 ⫾ .39 (15–61)
Paternal education .0265 1.31 (1.03–1.65)
⬍High school 29.8% ⫾ 1.2% 9.0% ⫾ 1.5%
High school diploma 46.2% ⫾ 1.3% 32.9% ⫾ 2.3%
⬎High school 24.0% ⫾ 1.1% 58.1% ⫾ 2.5%
Prenatal tobacco exposure 43.6% ⫾ 1.3% yes 19.1% ⫾ 1.98% yes .0013 .46 (.29–.74)
Prenatal illicit drug exposure 19.0% ⫾ 1.0% yes 7.2% ⫾ 1.3% yes
Time of death .0001 3.11 (2.26–4.27)
Midnight to 8:00 am 34.2% ⫾ 1.2% 7.9% ⫾ 1.4
8 am to 4:00 pm 37.4% ⫾ 1.2% 57.7% ⫾ 2.5%
4:00 to midnight 9.9% ⫾ .8% 17.2% ⫾ 1.9%
Unknown 18.5% ⫾ .99% 17.2% ⫾ 1.9%
Death on weekday 71.0% ⫾ 1.2% weekday 89.2% ⫾ 1.6% weekday
Race .0187
Black 32.1% ⫾ 1.2% 12.2% ⫾ 1.7%
White 46.8% ⫾ 1.3% 69% ⫾ 2.3%
Hispanic 13.7% ⫾ .9% 9.6% ⫾ 1.5%
Other 7.5% ⫾ .5% 9.3% ⫾ 1.5%
Cosleeping (yes) 32.1% ⫾ 1.2% 4.5% ⫾ 1.1%
Found prone (yes) 59.8% ⫾ 1.3% 74.4% ⫾ 2.2% .37 (NS)
Last placed prone (not usual) 1.1% ⫾ .3% 5.1% ⫾ 1.1% .0037 4.59 (4.52–13.70)
Found prone (not usual) 6.0% ⫾ .6% 12.5% ⫾ 1.7% .0114 2.09 (1.18–14.29)
NS indicates not significant.
* All differences were significant (P ⬍ .05) on univariate analysis; P values for multiple regression analysis are shown.
298 SUDDEN INFANT DEATH SYNDROME IN CHILD CARE SETTINGS
be licensed and are less regulated.
9
As recently as
1992 (the most current information available), ⬃30%
of children in family child care homes were cared for
by informal or unlicensed providers
16
; these provid
-
ers would not be regulated or registered by any
agencies. Consequently, there is no systematic ap-
proach to dissemination of Back to Sleep information
to these providers, compared with those regulated
by a state agency. In addition, surveys as recent as
1988 indicated that child care center employees tend
to be younger (80% are 40 years old or younger) and
well-educated, with more than one half of child care
teachers and assistant teachers having attended col-
lege.
17,18
Although there is no more recent formal
data on age and educational level of child care pro-
viders, these data from 1988 seem to still be fairly
accurate (J. Rickter, personal communication, 1999).
Because many of the child care center employees are
young mothers,
17
they may be more likely to be
aware of supine sleeping guidelines from experience
with their own children. Family child care providers
are often older women (50% are over 50 years old)
and tend to be less well-educated.
16,19
Eighty percent
of family child care providers have children who are
school-aged or older,
19
so they are unlikely to have
had personal experience with supine infant sleeping.
It is unclear why there is such a wide range of
child care deaths among the various states. The pro-
portion of child care deaths ranged from 9.4% in
Florida to 40.2% in Minnesota. Although there is no
state information regarding infants ⬍6 months or 1
year of age in child care, Minnesota has a high pro-
portion of children ⬍6 years with working mothers
(69%, compared with a national norm of 60%).
20
However, other states with proportions of children
⬍6 years old with working mothers nearly as high,
such as New Hampshire (67%), Maryland (66%), and
Florida (63%)
20
do not have comparable rates of child
care deaths. In addition, there is no correlation be-
tween the proportion of SIDS in child care and the
number of child care centers and licensed family
child care providers in each state.
20
Although there are guidelines regarding the diag-
nosis of SIDS,
21
we acknowledge that there may be
variability among state medical examiners in declar-
ing SIDS as the primary cause of death. Although
experts have agreed that complete autopsies should
be prerequisite for the diagnosis of SIDS,
21
they are
by no means universally performed by medical ex-
aminers in every state.
22
We do not have data regard
-
ing the number of autopsies performed in our sam-
ple; however, nationally, ⬃90% of sudden
unexplained infant deaths are autopsied.
22
In addition, because this was a retrospective sur-
vey, we relied on information available in the partic-
ipating states. Data collected by the state SIDS cen-
ters are considered very accurate, compared with
National Center of Health Statistics data.
23
In addi
-
tion, for each SIDS death, the state SIDS center ob-
tains data from birth and death certificates and sup-
plements the information with parent interview.
Although we believe that the data collected by the
SIDS centers are accurate, we acknowledge that there
may be slight reporting bias at several levels and/or
inaccuracies in the data obtained from the official
documentation. Although it is possible that there
may have been some differential reporting of deaths
in child care settings (ie, location of death more likely
to be reported if it occurred in child care), it is un-
likely to account for the large number of child care
deaths. In addition, location of death was deter-
mined both by death certificate information and pa-
rental interview, making differential reporting less
likely. However, we acknowledge that some of the
data were incomplete. Because of the emotional
trauma that would be involved with calling parents
several years after the infant’s death to ask additional
questions, we did not attempt to obtain further in-
formation. Therefore, we could not analyze informa-
tion regarding several important risk factors, includ-
ing postnatal tobacco exposure; preexisting
respiratory conditions; and the presence of quilts,
blankets, pillows, and other fluffy items in the crib.
Our conclusion that a disproportionate number of
infants die from SIDS in child care is limited by not
having living age-matched controls or, alternatively,
by not matching the proportion of infants in child
care to all deaths by census tract. In addition, our
calculation that only 7% of SIDS cases should occur
in child care is based on the assumptions that all
children in child care are in that setting 40% of the
time and that both SIDS and time spent in child care
is distributed equally throughout the day. However,
neither of these is true, and this may slightly affect
the proportion of SIDS that should occur in child care
settings.
It is difficult to determine exactly what factors in
child care may place infants at risk for SIDS. There
may be risk factors intrinsic to infants of working
parents that are unrelated to child care. However, it
seems that prone sleeping is an important prevent-
able factor. In our sample, we found that infants who
died in child care were much more likely than other
infants to be unaccustomed to prone sleeping. Stud-
ies by Mitchell et al
24
and L’Hoir et al
25
have re
-
ported that infants inexperienced with the prone po-
sition may be at extremely high risk of SIDS when
placed in the prone position. Many child care pro-
viders still may be unaware of the importance of
supine sleeping and may place infants prone, when
they usually sleep in the supine or side position at
home, for reasons of infant comfort.
12
In addition,
among the 99 infants for whom there was informa-
tion about the length of time in child care, we found
that approximately one third died in their first week
in child care, one half of these occurring on the first
day in child care. An extremely large percentage
(82.3%) of this group was found prone at the time of
death; parents reported that most of these infants
usually slept in a nonprone position.
The ages of 2 to 4 months have been identified
epidemiologically as the highest-risk period for
SIDS; part of this may be explained by the phenom-
enon of unaccustomed prone sleepers. Many infants
are now coming home from the newborn nursery as
nonprone sleepers; however, ⬃20% of these initially
nonprone infants change to the prone position at 2 to
4 months of age.
26
These infants may be at increased
ARTICLES 299
risk because they have not yet developed the upper
body muscle strength routinely seen in initially
prone sleepers, which may be protective in lifting the
head and clearing the airway when sleeping prone.
27
Although the national SIDS rate continues to de-
cline, efforts must continue to increase public aware-
ness about the dangers of prone sleeping for infants.
Pediatricians must reinforce with parents the impor-
tance of continued nonprone sleeping through the
first year of life. Parents must discuss nonprone sleep
position with any caretakers for their infants,
whether these be relatives, child care providers, or
occasional babysitters. It must be emphasized that
nonprone sleepers may be at greater risk if ever
placed prone. In addition, further efforts to educate
child care providers must be ongoing.
ACKNOWLEDGMENTS
This study was funded by a grant from the Gerber Foundation.
We thank the staff of the following state SIDS centers for their
assistance with data abstraction: Arizona Department of Health
Services (Robert Schackner, SIDS Director); California SIDS Pro-
gram (Ben Carranco, Program Consultant); Colorado SIDS Pro-
gram (Sheila Marquez, RN, Executive Director); SIDS Program,
Florida Department of Health (Annette Phelps, ARNP, MSN);
Center for Infant and Child Loss, Maryland (Jean Edwards, Pro-
gram Coordinator); Massachusetts Center for Sudden Infant
Death Syndrome (Mary McClain, RN, MS, Project Coordinator);
Michigan Department of Community Health (Cheryl Lauber, SIDS
Program Coordinator); Minnesota SIDS Center (Patrick Carolan,
MD, Medical Advisor; Kathleen Fernbach, PHN, Director); SIDS
Resources, Inc, Missouri (Lori Ahrens, Program Director); New
Hampshire SIDS Program (Audrey Knight, MSN, CPNP, Program
Coordinator); and SIDS Center of New Jersey (Linda Esposito, RN,
Education, Research, Communications Coordinator).
We also thank Bruce Sprague for data management, and Peter
Scheidt, MD, and Tina Cheng, MD, for their thoughtful comments
on the manuscript.
REFERENCES
1. American Academy of Pediatrics, Task Force on Infant Positioning and
SIDS. Positioning and SIDS. Pediatrics. 1992;89:1120 –1126
2. American Academy of Pediatrics, Task Force on Infant Positioning and
SIDS. Positioning and sudden infant death syndrome (SIDS): update.
Pediatrics. 1996;98:1216 –1218
3. Kepler J, Gibson E, Dembofsky C, Greenspan J. Changes in SIDS rate
and infant sleep practices since the initiation of the “Back to Sleep”
campaign. Pediatr Res. 1997;41:200A. Abstract
4. Silvestri JM, Mulvey KP, Tinsley L, et al. Assessment of sleep position
over time among infants at risk of sudden infant death syndrome (SIDS)
and healthy term infants. Pediatr Res. 1997;41:79A. Abstract
5. Willinger M, Hoffman HJ, Wu K-T, et al. Factors associated with the
transition to nonprone sleep positions of infants in the United States: the
National Infant Sleep Position Study. JAMA. 1998;280:329–335
6. Centers for Disease Control and Prevention. Assessment of infant sleep-
ing position-selected states, 1996. MMWR CDC Surveill Summ. 1998;47:
873–877
7. Guyer B, MacDorman MF, Martin JA, Peters KD, Strobino DM. Annual
summary of vital statistics: 1997. Pediatrics. 1998;102:1333–1349
8. US Bureau of the Census. Statistical Abstract of the United States: 1997.
117th ed. Washington, DC; US Bureau of the Census; 1997
9. Casper LM. Who’s minding our preschoolers? In: Current Population
Reports. Washington, DC: US Census Bureau; 1997
10. Casper LM, O’Connell M. State Estimates of Organized Child Care Facili-
ties. Washington, DC: US Bureau of the Census; 1998
11. Hofferth SL, Shawman KA, Henke RR, West J. Characteristics of Chil-
dren’s Early Care and Education Programs: Data From the 1995 National
Household Education Survey. Washington, DC: US Department of Edu-
cation, National Center for Education Statistics; 1998. National Center
for Education Statistics 98-128
12. Gershon NB, Moon RY. Infant sleep position in licensed child care
centers. Pediatrics. 1997;100:75–78
13. American Academy of Pediatrics, Committee on early Childhood,
Adoption and Dependent Care. Health in Day Care: A Manual for Health
Professionals. Elk Grove Village, IL: American Academy of Pediatrics;
1987
14. Hofferth SL, Brayfield A, Deich S, Holcomb P. National Child Care
Survey, 1990: A National Association for the Education of Young Children
(NAEYC) Study. Washington, DC: Urban Institute; 1991
15. US Department of Education, National Center for Education Statistics.
Child Care and Early Education Program Participation of Infants, Toddlers,
and Preschoolers. Washington, DC: US Department of Education; 1996
16. Galinsky E, Howes C, Kontos S, Shinn M. The Study of Children in Family
Child Care and Relative Care: Highlights of Findings. New York, NY:
Families and Work Institute; 1994
17. Whitebook M, Howes C, Phillips D. Who Cares? Child Care Teachers and
the Quality of Care in America: Executive Summary, National Child Care
Staffing Study. Oakland, CA: Child Care Employee Project; 1989
18. Ruopp R, Travers J, Glantz F, Coelen C. Children at the Center: Summary
Findings and Policy Implications of the National Day Care Study. Cam-
bridge, MA: Abt Associates Inc; 1979
19. Kontos S. Family Day Care: Out of the Shadows and Into the Limelight.
Washington, DC: National Association for the Education of Young
Children; 1992
20. Children’s Defense Fund. Children in the States 1998 Data Book. Wash-
ington, DC: Children’s Defense Fund; 1998
21. Willinger M, James S, Catz C. Defining the sudden infant death syn-
drome (SIDS): deliberations of an expert panel convened by the Na-
tional Institute of Child Health and Human Development. Pediatr
Pathol. 1991;11:677– 684
22. Centers for Disease Control and Prevention. Guidelines for death scene
investigation of sudden unexplained infant deaths: recommendations of
the Interagency Panel on Sudden Infant Death Syndrome. MMWR Morb
Mortal Wkly Rep. 1996;45
23. Brooks JG, Shaefer SJM, McClain ME, Hoffman HJ. Nationwide Survey of
Sudden Infant Death Syndrome (SIDS) Services. Rockville, MD: Maternal
and Child Health Bureau; 1994
24. Mitchell EA, Thach BT, Thompson JMD, Williams S. Changing infants’
sleep position increases risk of sudden infant death syndrome. Arch
Pediatr Adolesc Med. 1999;153:1136 –1141
25. L’Hoir MP, Engleberts AC, van Well GTJ, et al. Risk and preventive
factors for cot death in The Netherlands, a low-incidence country. Eur
J Pediatr. 1998;157:681–688
26. Ottolini MC, Davis BE, Patel K, Sachs HC, Moon RY. Prone infant
sleeping despite the “Back to Sleep” campaign. Arch Pediatr Adolesc Med.
1999;153:512–517
27. Davis BE, Moon RY, Sachs HC, Ottolini MC. Effects of sleep position on
infant motor development. Pediatrics. 1998;102:1135–1140
300 SUDDEN INFANT DEATH SYNDROME IN CHILD CARE SETTINGS
