AMCHP 2006 ANNUAL CONFERENCE
EARLY CHILDHOOD: BUILDING THE FOUNDATION FOR LIFELONG HEALTH
March 4-8, 2006
MELINDA REDDISH DOUGLAS: Good morning. I’m Melinda Reddish Douglas with Oklahoma State Department of Health. And today, I’m going to be presenting a study of deaths that occurred among sleeping infants. I want to acknowledge the coauthors of this study, Pam Archer in the Injured Prevention Service and Dr. Jeff Golfdom, the chief medical examiner in Oklahoma. And kind of what Jim was saying, it takes everybody’s opinion and experiences and expertise. So I’d like to acknowledge in the health department, working with child guidance, maternal and child health, injury prevention and all the very valuable resources within the Health Department.
Each year in Oklahoma, approximately 400 infants died before their first birthday. Congenital and chromosomal abnormalities, short gestation, and SIDS are those three leading causes of death. In Oklahoma the death rate though has declined 7% from 8.4 per thousand in 2000 to 7.8 per thousand. This study came about when several organizations and agencies were really exploring similar concerns. The chief medical examiner became concerned about the number of deaths among infants that were sleeping with adults that had come to their investigation attention and was working with injury prevention. At the same time, the Child Death Review Board and the Department of Human Services noticed an increase in infant co-sleeping deaths. And thirdly, then, the Safe Kids Coalition in Maternal and Child Health began working on materials to look at safe sleeping and Back to Sleep campaign. Now nationally, Safe Kids has added SIDS as a component because it is believed up to one-third of SIDS cases are actually misdiagnosed cases of asphyxia, basically because of the complications and difficulty and distinguishing certain cause of death, especially among certain infants. Well, research literature has shown us that there are unsafe sleeping environments, specific factors, such as sleeping on the stomach, infants being exposed to secondhand smoke or prenatal smoking, infants sleeping on furniture not designed for sleeping, such as sofas, chairs, loveseats, infants sleeping on soft bedding or with pillows or quilts, and infants sleeping where there are spaces between the mattresses and the furniture and walls which pose entrapment hazards. What is unclear though is should co-sleeping be added to these list of factors?
Private organizations and public agencies have developed policies and awareness campaigns that really span the gamut from discouraging co-sleeping to encouraging co-sleeping. For example, the February 2005 American Academy of Pediatrics Policy on Breastfeeding says that an infant should be in close proximity of the mother at all-time, and there was a news release that really encouraged co-sleeping. But there is another American Academy of Pediatrics Policy that discourages co-sleeping, saying it could be hazardous in certain circumstances. When we look at how is co-sleeping getting more prevalent, we looked at the National Infant Sleep Position Study by the National Institute of Child Health and Development and they compared 1993, where 6% of infants usually spent the night in bed with an adult to13% in 2000. They also found in 2000 at 20% of infants spend at least some time in the bed sleeping with adults, and at the time of the study, 45% reported that the infant has spent sometime during the past two weeks sharing beds with adults. Among those who shared bands, infants were twice as likely to be covered by a quilt or a blanket than non-bed-sharers, and bed-sharing was more prominent among young mothers, those with low income, non-whites, and when the infant was less than eight weeks old.
In Oklahoma we did a local focus group to kind of interview and understand what were some of the issues with co-sleeping. And reasons for co-sleeping were to promote bonding and breastfeeding, to follow cultural practices, to alleviate the fear of SIDS, to be able to hear the baby, and to promote better sleep for the baby and parents. Although it wasn’t the case for the focus group, they did mention, well, some people may not have a crib.
So we designed a study to review infant deaths that occurred while sleeping and included the characteristics of the sleeping environment, especially co-sleeping, meaning infants sleeping on the same surface as adults or other children. And the purpose of our study was to investigate one, the number, the circumstances, and then really to guide a campaign in order to address this issue. For this study we used medical examiner data. Medical examiners investigate deaths that are violent or suspicious and nature, unattended by a physician, unexpected during therapeutic procedures or other very specific circumstances. Now the medical examiner is charged with determining the manner of death, meaning was the death accidental or intentional, such as homicide or suicide, or the intent is unknown. They also determine the cause of death, meaning was it asphyxia, SIDS, drowning, or an unknown cause, meaning they really were unable to determine that cause of death.
Now the medical examiner is statewide, and all of the reports are combined into one electronic database. We used the medical examiner database and selected deaths that occurred during a four-year period, from January 2000 through December, 2003. Further selection criteria included only Oklahoma residents who were less than 12 months of age at the time of death. We selected the manner of death, coded as accidental, and the cause of death, coded as asphyxia, or the manner of death, coded as unknown, and the cause of death coded asphyxia, other, or unknown. The selection criteria excludes SIDS because it is a natural cause of death and excludes stillbirths. Hard copies of the Medical Examiner reports of investigation were obtained for cases meeting this first set of steady criteria.
And investigation reports included demographic data, seeing observations, a narrative, and autopsy report. And by reading the narrative reports, we were able to determine what the infant’s last known activity was. And only those who were sleeping were included in the study. For coding events of sleep, and of co-sleeping, we used the following criteria for determining status. Unsafe sleeping was coded if the infant was not sleeping alone in a crib or bassinette, was not put to sleep or found on back, or pillows, stuffed quilts, blankets, or comforters were in the crib or bassinette. Co-sleeping was coded in the infant was sharing the sleep surface with another person, and that surface could include beds, couches, chairs, or other surfaces.
So in the database, we identified 124 possible cases, based upon that coded manner and cause of death. And of which, 113 were sleeping prior to death. Only two infants were found to be in safe sleeping environments. For five cases, there was not sufficient details in the records to determine if the sleeping environments were safe or unsafe. So 94% were determined to be in unsafe sleeping environments. The manner of death was coded as unknown in 81% of the cases, and the cause of death was coded as unknown or other in 80% of the cases.
So the following data in slides described the characteristics of those 106 infants found to be in unsafe sleeping environments. This graph shows the age in months and the sex of the infants. Overall, 62% of the deaths occurred among infants who were less than three months of age, and peaked at one month old. Female deaths are indicated by the red bars, and male deaths indicated by the yellow bars.
Males accounted for 61% of the cases. Twenty-eight percent of those deaths among males occurred at two months of age. Among females, nearly half of the deaths occurred before two months of age. Sixty-three percent of the deaths occurred among whites. Whites accounted for 52% of the unsafe sleeping deaths in 2000. By 2003, whites accounted for 74% of the unsafe sleeping deaths. However, for our study period, African-Americans had the highest rate, at 1.2 per thousand births, which was three times higher than the rate among whites and among American Indians. The overall rate was 0.5 per thousand. Most of the deaths occurred at night, rather than during a nap period. Ninety-seven percent occurred in a private home, mostly the infant’s own home. Two percent occurred in licensed childcare, and one death occurred in the hospital, when teen parents fell asleep with the baby in the hospital bed.
Now this pie chart shows the sleeping surfaces that the infants were on prior to death. And less than one-fifth were in cribs or bassinettes. Beds were the most common surface at 50%. Now deaths occurred when infants were placed asleep on waterbeds and playpens, mattresses on the floor. One death occurred in a stroller, and one in a baby’s swing. Now among the deaths that occurred in cribs or bassinettes, two-thirds of the cribs and bassinettes had items added to them that made them unsafe. All of the bassinettes and 57% of the cribs had items added such as a standard pillow added to a bassinette, or a twin comforter, receiving blankets, and eight stuffed animals added to a crib.
Now this pie chart shows the possible mechanism of injury as mentioned within the reports, the medical examiner reports. It’s not a definite cause of death, but a possible mechanism of injury. Thirty-five percent of the cases involved possible or definite overlay. Overlay is when another person has covered the baby, either with a part of their body or arm. Possible is when they’re co-sleeping and the medical examiner said that this is possible. Definite is when a third person discovered the adult or child overlaying the baby. So that would be definite. If the medical examiner documented possible or definite overlay in the report, then we classified that mechanism of injury as overlay, even if there were other hazards in the bed, such as quilts or blankets.
Just over one-quarter of the reports lacked sufficient details to identify a possible mechanism of injury. Now 10% of the infants were entrapped, some wedged between the mattress and wall or furniture, some between couch cushions in the back of the sofa. Six percent were put to sleep on a bed or on a mattress but were found on the floor. Among those found on the floor, half were found in piles of clothes, the other half on plastic materials. Now there were differences between the age of the infant and the mechanism of the injury that was identified. So significant differences were among infants who were less than three months of old, age, definite or possible overlay was recorded more than other mechanisms. Among infants who were three months of age or older, entrapment was recorded more than other mechanisms. Several health factors or infant factors were documented in the medical examiner reports. Evidence in the records show that 6% were being breast fed, 6% were low birth weight babies, 15% were premature births, and 25% had some indication of a respiratory illness. And as indicated by the percent of not specified, these factors were not consistently recorded in the medical examiner records.
Now the records gave indications of other factors on the caregivers in the family. Fourteen percent of the records show that the caregivers were using drugs or alcohol just prior to the death. Eleven percent had indications of previous drug or alcohol problems in the home, 21% of the families had come in contact with Child Protective Services. Either the family had had other children removed from the home, current open cases, or referrals were made to the hospital at the time of the infant’s birth. Three percent of the records documented previous infant deaths due to SIDS, and 3% of the records documented exposure to second-hand smoke.
Again, these factors were not consistently documented. Overall, co-sleeping occurred among 64% of these unsafe sleep deaths. From 2000 to 2003, the number of co-sleeping deaths has increased as indicated by the red bar. In 2000, slightly less than half of the deaths involved co-sleeping, and by 2003, the number of co-sleeping deaths were about three times higher than the number of deaths that occurred in unsafe sleeping environments that did not include co-sleeping. The largest increase in 2003 was among whites. And the co-sleeping deaths in 2003 were significantly higher than through 2000 and 2002.
Now the following slides are going to look at just co-sleeping deaths. So our analysis on just the co-sleeping deaths show that just over half of the infants were less than two months of age, which means this is a younger age group than the overall unsafe sleep deaths. Males were still represented more than females at 57%, and the rates among African-Americans and American Indians were higher than those among whites.
Seventy-one percent of the co-sleeping deaths occurred in beds, 9% on other furniture, 6% on a mattress on the floor, and 4% on a waterbed. And one where the infant was sleeping in a crib with its twin. Nine percent were on other surfaces, or the surface was not identified.
For 55% of the deaths, co-sleeping deaths, the medical examiner documented possible or definite overlay. Four percent mentioned entrapment, 3% mentioned pillows, 6% stated that the infant was found on the floor, and 1% was other. Thirty-one percent of these cases lacked sufficient details to identify a possible mechanism of injury. We looked at those 31 that lacked injury and mechanism of injury. So the n here is actually 21. Some of those factors present in cases that lacked possible mechanism of injury details included one third had current or past child protective services involvement.
Nearly 20% occurred when the family and the infant were sleeping at someone else’s house. Either they were there for the holidays, they were there for a party, or it just said they were staying over at someone’s house. Ten percent of the infants had previously documented episodes of apnea, and 10% were found facedown, it was documented found facedown on mattress. And 5% occurred on a waterbed. So half of the co-sleeping deaths occurred when the infant was sleeping with one other person. Forty percent of the deaths occurred when there was a total of three people in the bed, and 10% occurred when there were four people.
One example is that an infant mother, grandmother, and grandfather shared a bed. The infant was sleeping with one or more adults in three-quarters of the cases. And in 18% of the cases, the infant was sleeping with another child. And in 9% of the cases, the infant was sleeping with at least one or more adults and one or more other children.
An interview with the chief medical examiner and a 23-year veteran medical examiner was conducted. Their experiences were that infant deaths were increasing and co-sleeping was increasing, especially among the white population. They thought that the individual medical examiners’ beliefs and experiences and opinions reflected in how and what they reported. During the interview, it was mentioned that the medical examiner’s office had a state appropriations cut in 2003 and could not afford full drug and metabolic testing. Therefore, budget cuts may have limited the ability of the medical examiner to determine the manner and cause of death. For example, in possible SIDS cases, where since the SIDS diagnosis is exclusive of all other causes of death, it has to have laboratory testing.
So we verified this anecdotal information from the medical examiner using other data sources. According to the vital statistics data indicated by the VS all infant row and blue text, the overall infant death rate and numbers were down from 2000 to 2003. As seen on the ME all infant (inaudible) black text, the ME investigated a larger percentage of all infant deaths. So something was changing in the characteristics of the deaths that they now became under the medical examiner. And the 2003 number of unsafe sleep deaths that the medical examiner investigated, as indicated by the ME unsafe sleep death, or unsafe sleep row in red text increased and accounted for a greater percent of the investigated deaths.
Because of that lack of funding and limited funding and testing for metabolic testing in 2003, and that potential change in coding and manner of cause of death, we plotted the medical examiner, the unsafe sleep death, as seen in the yellow line, and the ME SIDS deaths, as seen in the red line. And the unsafe sleep deaths increased dramatically from 2002 2003, while the SIDS deaths decreased 56% during that time.
So our conclusions of the study were that unsafe sleep deaths occurred most often among infants less than three months of age, half of the unsafe sleep deaths occurred in beds, and among the deaths that occurred in cribs and bassinettes, items such as quilts and pillows that made the environment unsafe were added. We found differences in the mechanism of injury among infants who were less than three months of age and those who were older. And we concluded that co-sleeping deaths are increasing, mainly among the white population, and we’re not clear of this increase as an actual increase or an artifact of the medical examiner coding of the cause and manner of death.
So we have limitations in this study, especially with the medical examiners, that there is not a standardized documentation of what factors are collected among investigations. And these factors include the positions, place to sleep, the availability of cribs, and other risk factors. And the medical examiners, they wanted everyone to know that they’re limited by what people at the scene choose to tell them, that that is their limitation. Limitations also include the ability to consistently distinguish between SIDS and overlap and unintentional asphyxia in child abuse. And of course the study was limited by the budget cuts to the medical examiner’s office.
Now other states have noticed similar trends in co-sleeping deaths, and have responded in a variety of ways. And some have found that recommending that infants do not to sleep with adults have caused great controversy, especially among advocates of attachment parenting, or breastfeeding, resulting in a prevention-ness message being lost.
There is also the potential shift in diagnosis from one of SIDS to co-sleeping, and one study found that one-third to one-half of SIDS deaths occurred while infants were co-sleeping. Another change may be the use of the unknown cause of death. It is found that the unknown cause of death can be more easily changed on medical examiner records or official records than other causes of death. And of course, further research is needed to identify possible markers that can assist in the determination between SIDS and asphyxia. So for our recommendations, we recommended that an educational campaign for businesses, especially those marketing to parents of newborns, in a store in Oklahoma City, of the 40 cribs present on display, not one was a safe sleeping environment. As illustrated by this picture, where there are quilts and bumper pads and pillows placed within the cribs.
We also recommend increasing the public awareness of co-sleeping deaths and creating messages that describe safe sleeping environments that can reduce the risk factors for overlay for SIDS and asphyxia. And these messages would be promoted through collaborative efforts, and we’d recommend creating messages that the target audiences can hear, and do, and the messages that breastfeeding, attachment parenting, and co-sleeping advocates will not protest.
I think it has questions here, but we’re going to save the questions to the end, and again, I wanted to acknowledge Jim and the work that’s being done at the health department, trying to find a blend between those messages of what is the message. Because with car seats, there can be dummies with sensors and mock crashes to say this is a safe car seat, this is not a safe car seat. With infant sleeping, there is no test to find out what is safe or not safe. So, I think we’re going to hand it over to Michigan, who also found the same types of things, but have moved past this process. Thank you.