AMCHP 2005 ANNUAL CONFERENCE
DELIVERING RESULTS, IMPROVING PREGNANCY & BIRTH
February 19-23, 2005
JAMES KEMP: My name is Jim Kemp, and I'm a pediatric pulmonary physician from St. Louis. I'll be working with Pat Tackitt, and we'll talk about the importance of death scene investigations. And we are going to focus on sudden infant death. Most of the deaths we are going to talk about happen when the babies are asleep.
Dr. Shepard is going to talk a bit about diagnostic shifting, and I'll lay the historical background for diagnostic shifting. It's probably happening in your city. Once called SIDS, now called suffocation and called SIDS again. It's important to remember that when SIDS was defined, the scene investigation was by and large not available. So the definition, the original definition of SIDS did not include much scene investigation.
I'm going to talk about the epidemiology and physiology that's pertinent to the scenes Pat is going to show you. So it will be scenes, data, scenes, data, that kind of stuff.
When we think about babies dying soley and unexpedetly during sleep, you can't think of one true cause. Death certificates require something at the top of the list. There has to be one cause. But all of the deaths of babies that occurred suddenly and unexpectedly during sleep will incorporate this triple risk model, I think.
A baby at risk at a vulnerable period of his or her time of development is exposed to exogenous stressor and it leads to a lethal outcome.
It's hard on death certificates to incorporate all this thinking because you have to put just one thing at the top.
Once SIDS became available as a diagnoses, diagnostic shifting took off. SIDS was first defined in 1969 and there became -- there was an international classification of diseases code available in '73. With the introduction of a new rubric for SIDS in '73, most of the deaths previously assigned strangulation or suffocation were diagnosed by California coroners, diagnosis by California coroners were called SIDS. This is from a paper by Jess Krause looking at California data from the ’70s and '80s. So deaths that had been called suffocation because it was assumed the babies suffocated became SIDS death in California. That happened across the country.
The historical framework for this is the following. There was a really important paper from New York City in the 40's by a man named Abramson that showed that the majority of babies dying during sleep in New York City died prone and face down. But the paper that sort of began the SIDS era is a landmark paper by Maria Valdez Depina in pediatrics in 1967 in which she said look at we -- based on medical history and post mortem data, we don't know why most of these babies die. But Dr. Depina a was apparently not aware of Dr. Abramson's work and her paper doesn't really talk about death scene investigation. So SIDS was defined based on negative medical history or inadequate medical history and nothing in the post mortem that said this is why the babies died.
So what happened after the SIDS era began in the late '60s and early '70s that as Jess Krause mentioned that SIDS was common and suffocation was really uncommon.
Beginning about 1990, based on information that suggested sleep position and certain items of bedding were very important, the possible contribution of suffocation was sort of reintroduced based on the scene investigation data that became available. I think the unfortunate thing is that most states still scene investigations are not done in the majority of deaths, and in some places they are not done hardly at all or in some states, they sort of got into it and sort of stepped back. But the progress has been made in babies dying less often has been based on scene investigation data.
Now this slide, I know the X and Y axes are not visible, but this is data from Henry Krause Roger Byard from South Australia and the X axes is (inaudible) and it goes from '84 to '98, and the y axis is a number of deaths. The top curve is all post neonatal deaths, post perinatal deaths. And you can see that from the start in '84 to '98, it went down.
The middle curve is SIDS, and beginning about 1990 which is just left to the word year there, SIDS started going down in south Australia and it continued to go down through 1998.
The bottom curve is accidental deaths, and as you see that has sort of crept up and that continues to increase in South Australia. Krause and Byard concluded that although diagnostic outcomes have altered, that is, we are changing what we call these deaths, is more likely the result of more careful interpretation of the extensive investigation that are no undertaken rather than arbitrary reclassification.
So interventions caused babies to sleep on their back and fewer of them died, so the top line is going down and there were fewer deaths felt to be mysterious deaths, so the middle death is going down. And with the institution of more scene investigations, the accidental component of these deaths is more appreciated. So the accidental is going up although the background rate, the total post perinatal deaths is going down in South Australia.
Your system is low on virtual memory. I should say great. Super. Okay. Now I need to go back a few.
All right. So the concern many of you have in your state if you're going --if you went from suffocation back to SIDS, because you're not doing death scene investigations, then the preventability access of the analysis of these deaths will be lost.
In the city of St. Louis, and the medical examiners are very good here, there is three deaths that basically we look at when we do our studies. There's SIDS, accidental suffocation and cause of death undetermined. And frankly, in St. Louis in particular it's almost serendipitous what these deaths are called, because we have deaths that seem identical to another. One case will be called suffocation, another case will be called SIDS. So we don't try to make the medical examiners be uniform in their approach because it doesn't work. It doesn't work in St. Louis, I bet it doesn't work in Michigan, it just doesn't work across the country.
But if we look at all these deaths together, this will catch all the deaths less than one year of age related to sleep. And I want to emphasize that case control studies that focus on SIDS such as the recent Chicago Infant Mortality Study, CIMS, there was 260 SIDS deaths, but there was 75 more deaths that were called accidental suffocation and undetermined, and why some were in the SIDS group and some in the accidental suffocation group is serendipitous according to my reading of this literature. So what we want to focus on in St. Louis is post neonatal deaths during sleep are related to beds or sleeping and not just the old ICD 9 code 79A which is SIDS, and I think this approach has allowed us to make progress within a system where there's a lot of serendipity involved.
PATRICIA TACKITT: I'd like to take you to a few death scenes with me and I'll frame it or give you a little background.
This is the child death review team process, and the photos you see in this presentation will not be in your handouts that's because they are recreations in people's homes, and some of these identify themselves. We don't want to make their tragedy worse. So you won't have the photos, but you will have the information presented here.
We are looking at why do we do scene reenactments, how do we do them and what are the considerations. And there was a handout I had planned and unfortunately it didn't get in the packet. So what I'm going to tell you, if you contact Keeping Kids Alive.ORG or I'll give you my e-mail afterwards or make sure I send it to you. It's about a five page handout and it talks about when we look at child death review the kind of considerations we have based on Soodees (phonetic) and SIDS.
So these are findings at death scenes. This is a common finding. When a police officer goes out and visits a scene, they often will take photos of just what you see here. And my question to you is, can you tell me what happened? You know there was a bed, you don't know how many people were in it, you don't know where the baby was. My point is that with this kind of information, it's not enough and frequently the parents are still at the hospital and whoever is left at the house are talking to the police department, so the information, the level of the data is not enough for the medical examiner to be able to tell what caused the death.
This was a mom where the case was signed out basically at SIDS at the medical examiner's office because I had not been able to locate her. It took me three weeks to find her, she had moved after the death. I asked them to hold the death certificate. I went back and finally was able to locate here. We took this bassinet out of storage and she showed me exactly what happened. See if you can tell better what happened here.
This is a 17 year old mom. This is her very first baby. She was shown in the hospital how to double swaddle the baby, so she's demonstrating that for me, to do it very tightly. Her apartment is 85 degrees because she thinks babies need to be warm. She put the baby in the bassinet on its side and in the morning --she also covered it with another blanket. She found the baby like this.
Now this case had already been signed out as SIDS. They amended it because the baby's face is straight down, nose and mouth obstructed and what do you find under the sheet? A pillow. Her aunty always did that with all of her babies and they all lived, so she thought it was safe.
So my question to you is can you tell from this scene reenactment what happened. The parent, the expert, is telling us what happened. We hand her the doll and she tell us two things. How she placed the baby down and how she found the baby. Takes the bias out, helps explain to the medical examiner how things work.
this is another case and it kind of brings to the point that you have to sometimes look carefully at the scene reenactments. This is the only case I've been out on with the police. We came to this home, they've been to the emergency room, I met them at the home when they came home. Mom told us the five month old had been in the bassinet, and I said how did you find the baby? And she said this is how she found it. We looked at the bottom of the bassinet, there was no way the baby could have fallen through the bottom. So she was telling us the baby had gone overboard from the bassinet and gotten it's head caught on the way down. The police and I had a little problem with that, so they took her downtown and and after about an hour she divulged that she had been bed sharing with the baby in a twin bed, but they didn't get anymore information for the medical examiner. So we knew she was in bed, but what happened? Was the nose and mouth obstructed? Was her arm over the chest? Had the baby fallen off the bed? What had happened? So they aren't used to getting that level of data for the poice so we could tell the medical examiners what happened.
When I was able to find her after the police were done, it took me about three weeks because she moved again as well. She went upstairs and showed me that basically she went to sleep with the baby in a twin bed on the floor in this position. And as she went to sleep, her arm must have gotten somewhat heavy because you relax, and when she woke up, her arm was like his. She was very willing to tell me. She needed help to kind of get some services, but she was very willing to tell me. And again we handed her the doll.
I asked her with she didn't tell us the first day and she said because she had been told not to sleep with the baby and she didn't want it to be her fault, and you can understand that. And the more we educate people about how to sleep safely with the baby, the more we are getting people giving us back after they baby has died exactly what we told them was safe instead of what happened. So you have to look very carefully at your scene reenactments and make sure they match the lividity matter.
So I'm saying in infants found unresponsive, the context in which the death occurred is often lost. When they do CPR and move the baby, we don't know often what happened because when law enforcements comes back, they think this is a medical issue so they are looking at medical things with the emergency room and the medical examiner's office, but they don't bring this data to the medical examiner so they know what happened. Careful reenactments may help us gain the information of the context, but sometimes people misrepresent it. So it's important to check. CDR team needs a process that will bring accurate scene information consistently to the ME and the team. And we again need to find effective ways to do better death investigation so there's something behind our stats that's accurate.
This is our first set of our scenes. This is nose and mouth totally obstructed. This is probably a few weeks ago now. Baby was at Christmas time lying on ts back at the other end of the sofa with mom. Baby is four months old, and when mom found the baby, the baby turned over. That's all it took. The baby was face down in the pillow. Mom found her in enough time that she was still warm. She was in the hospital for 24 hours, but she was brain dead.
This is a mom with a 17 day old infant. She had been playing with the baby on her chest on a Sunday morning. She had her other child an the other end of the bed. This is a pull out sofa bed. Baby has fallen asleep on top of mom while mom fell asleep as well. And when she woke up, the baby was here, and the nose and mouth are totally obstructed.
And this is again fairly recently. This is a Pack and Play of some type, generic Pack and Play. This is a baby on top of a very thick quilt. They did not have enough sheets, they had thrown them in the washer. So they put the baby on top of the quilt and the baby was on its side, and as it rolled during the night on to its tummy, its face went down into a very thick quilt. Dad got up about 4:00 in the morning for a job he found the baby this way and turned the baby over, and didn't turn a light on because he didn't want to wake her up. Didn't realize she was deceased, and when he left his mother woke up and came down the stairs about 15 minutes later and found the baby deceased. If we only took the information from her and didn't track back each step of the way, we would have thought the baby was supine and that's its nose and mouth were unobstructed. But when we went through the entire thing from the time he went to bed, where the baby was each step of the way, then we could tell this was a baby whose nose and mouth were blocked. And we had the police measure the thickness of the quilt, it was about an inch and a half, very soft.
JAMES KEMP: So what Pat had shown you was a series of scene reconstruction were babies were found both prone and face down. It turns out if you look back at SIDS series, deaths that were called SIDS, not by me but by the person who wrote the paper and by the medical examiners. They are fairly common in SIDS series. Going back to the '50s, 29 percent of babies in Cleveland were found prone and face down. Classic studies from England in the '60s, 20 percent were face down. Another set of classic studies from Seattle, early '70s, 27 percent were face down in. In the New Zealand cot death study, over half of the babies were found prone and face down. In Australia, Tasmania in particular, 39 percent were prone and face down. And a study that MJ Shears did with Consumer Product Safety Commission, almost 30 percent of babies in the US mid '90s were prone and face down.
Amongst a group of babies who have gotten a particular interest lately, since most babies actually are used to being put down on their back, if babies are unused to sleeping in the prone position and they turned to the prone position or they go to daycare at grandma's house and are put in the prone position, that's the inexperience being prone, they are at very high risk of being found prone and face down, anywhere from 43 to 71 percent. And in San Diego in a recent paper by Henry Krause, 37 percent of babies were found prone and face down.
So this is not a small, you know, tiny group in SIDS populations, and obviously this would be a group of deaths where, you know, if there is a whole other group of babies that are called accidental suffocation, it might be a group of deaths that are virtually indistinguishable from these, but these were called SIDS.
We were involved a number of years ago in looking at sheepskins, this is a long wool sheepskin. We modeled the death of over half the babies in New Zealand were prone and face down and the majority, over 70 percent, were on sheepskins. And this is a head from an infant resuscitation mannequin, and if you've done baby CPR classes, you know there's an airway associated to this. We attached this airway to the rabbit, and we had the rabbit breathe through its tracheotomy through a doll's head because the contour of the face makes a difference in this model. And when you do the this study you get the following: This is a C02 tracing, a carbon dioxide tracing. And the upper tracing is a normal tracing. You go up and it goes like this and ends at five percent and the rabbit breathes in, and this is with the rabbit breathing through the doll's head with the face up, and then it exhales down and starts to breathe in. It gets very close to 0, and that's fresh air with essentially no carbon dioxide. If you take the doll's head and put it face down on the sheepskin, you get the lower curve, and you can see the end is about 8 percent which corresponds to a blood gas of CO2 of 56 and the inspired CO2 which should be 0 is up to five percent. So the partial pressure of Co2 when you inspire is 35 millimeters of mercury. So this is what rebreathing looks like when you monitor capnometry and this was a very dangerous predicament for these rabbits to be in, and I suspect it was very dangerous for these babies to breathe into these environments also.
PATRICIA TACKITT: These photos you'll see the baby's nose and mouth are not fully obstructed, but partially obstructed. So mom who had been out for the evening, she forgot she left the sheets in the wash and so she didn't have them dry. She put the baby on top of a double thickness of a quilt. You can see if you look at the doll that both of her eyes are visible, you can see fully her nose for both nares. But when mom found that baby, and I would offer the crib was right next the bed but wasn't being used, that basically she found the baby like this. Can you notice the difference in the positioning? So 50 percent of her ability to get rid of carbon dioxide and breathe in fresh air appeared to be compromised.
This is a dad who is sleeping with a child who is six weeks old. Mom had gone out for the evening. Again, a bassinet next to the bed. And this baby at some point in time late in the evening, dad had fallen asleep, rolled off the pillow and again had its head in a position where the nose and mouth were partially compromised. Six weeks old, breast fed baby, but unfortunately again the outcome was a tragedy.
This is a mom with an ill child. She is 16 years old. She believes even though there is a crib right next to her that she should sleep with her baby. So she fell asleep cuddling her baby like this. And I find that quite commonly, that people's arms are over the baby. She went to sleep about 4:00 in the morning, she got up at 7:00 and changed the baby, fell back asleep, grandma came home at 10:00 and found the baby in this position with its face partially into the pillow and mom's arm was still across the middle. I think for babies that are ill they, again, are at more risk because they may not have the ability, they may have longer apnea periods and their oxygen level doesn't stay in the same range.
JAMES KEMP: And so if the baby's nose and mouth is straight down, the physiology is much clearer than you might imagine if the baby had a cold. When you get a cold, you know one side of the nose is plugged up more than the other. So if the baby's good side is down, that could obviously be a problem. People studied air flow through noses. It's something that's very interesting actually, and it changes diurnally around the clock. So, the physiology is much more straightforward if the nose and mouth is straight down.
There are other explanations for death in the prone position including heat stress, and it's felt babies radiate heat less well when they're prone and most of their face is covered.
Where the nose and mouth is we think is the most important distinction you can get from the prone death in the death scene investigation. Even if the nose and mouth isn't faced straight down, there's still a dramatically increased risk of the babies dying in the prone position. For example, babies awaken from sleep less robustly if they are in the prone position than when they are on their back. And certain kinds of apnea, particularly something called a laringealchemo reflex apnea, which is apnea response to having secretions in the throat will be longer if the baby is in the prone position compared to being on its back.
Even if one nostril is uncovered, rebreathing is likely to happen for reasons that I mentioned, if one side is plugged. And remember, normal inspired carbon dioxide percent is 0 percent. In certain scenarios, such as if you use a long wooled sheepskin and even if the head is turned 45 degrees from one side to the other, so that the nares are in a midst of sort of a forest of wool, you can get very significant rebreathing. 8.5 percent in a study done by David Bolton in babies when the head is to the side in a four centimeter long sheepskin and 10 percent of inspired Co2, and remember 0 is normal. For a six centimeter long sheepskin, that's about that long, these sheepskins commonly have wool that long. So even when the nose is turned to the side, the air doesn't waft up into the room and sort of mix with the general air, it tends to be kept near the baby's face.
If you look at ordinary bedding from the United States, the most common scenario we see is that babies have predominantly flexor upper extremity tone, and if there's a little receiving blanket underneath them, they sort of bunch it around their face, okay, and they sort of create this little bunch of blanket around their face and it can produce potential lethal rebreathing. We studied that with Pat Carol from Minneapolis.
So rebreathing, heat stress, and less robust reflexes that cause the baby to arouse from sleep or to protect its airway. These are all potential mechanisms that are less likely if the baby is on the back. So not having the nose and mouth completely covered is not a requisite for having unexplained death if you will in the prone position
PATRICIA TACKITT: Our next section is going to be babies that are on their back or on their side and they get the covers over heir head or things that are in their bed near their face. This is an infant that was trying to take a nap and its sibling kept waking it up so mom put it in the roommates room and put the babies hands covered by the quilts because she felt it was too cool in there. When she came in to find the baby about an hour later, she found the baby like this and the baby was unresponsive.
This is a scene I went out on about six months ago and mom told me the baby was on the back in the bassinet and there was -- this is how she also found it she said. I would like to point out if I had just taken the information and gone back to the medical examiner's, what would this have been called? SIDS, that's correct.
It's not that I'm against SIDS, it's that I'm definitely looking to get what the real information of what happened is. And so I rubbed her back and I said you know, I have been to a lot of homes and this is not what we are finding. Could you please understand this is really important for babies, so we can prevent deaths. And after about three to four minutes, almost five minutes she finally said this is what happened. And she said I just wanted my baby to be warm and it was winter and there was cold coming in, so she had used this to line the inside of the bassinet. This baby was only about six weeks old -- actually it was three weeks old and this had gotten over the baby's face and when she woke up on the couch next to the baby, this is how she found it. But if I stopped before that, we would have had a SIDS. It's not that I'm against that, but we have to have the truth.
This is a mom who was sleeping across the bottom of a queen size bed, her baby is in the middle going top to bottom, mom fell asleep studying. Here is her baby. It had pillows all around it so if it fell out of the bed -- it kept it from falling out of the bed. And it had a lot of toys in there. And grandma said she thought it was cute because the baby was chewing on the ear of the stuffed animal, but when she walked in a little later, this is how she found the baby and it was unresponsive. I don't often see this, but there can be enough thing in stuffed animals and the blanket that helped again block the airway and rebreathing .
JAMES KEMP: So there's actually a surprising amount of disagreement in St. Louis at least about whether these deaths should be called suffocation or SIDS, that is when the baby is on its back or on its sides with blankets or something over its head. And the inclusive argument is that if the majority of babies could free their heads and get access to fresh air, then this baby who probably did suffocate would be called SIDS because the scenario would not be dangerous to the majority of infants. That is, this is not something that is going to be lethal to the majority of babies, so there's probably something primarily wrong with this baby and therefore we'll call it SIDS rather than suffocation.
Now, again, that fits into, as far as I'm concerned, serendipitous discussions, but that's in fact what happens in St. Louis and I'm sure it happens in a lot of places.
The question is if it's called SIDS, will the aspect of preventability be lost? So it's very important for these people making these determinations to have some sense about what babies can or cannot do to get access to fresh air. It's important because this again is not a small fraction of babies dying suddenly and unexpectedly. If you look at SIDS cases, not those that are called suffocation, 14 percent of SIDS in Australia, a quarter in Scotland, over a quarter in the Netherlands, that is a high percentage that is called SIDS from the -- a high percentage of deaths called SIDS the baby is supine or on its side with its head covered. The assumption is that babies were unable to free their head and they did not arouse to free their head. But the basis for these forensic decisions oftentimes is not clear. In the city of St. Lewis, we have ten deaths of babies on their back with their head covered. Five are called SIDS and five are called suffocation. And that's what we had to sort of deal with.
What do babies do when they get blankets over their head? It's important to consider developmental and behavioral physiology. In other words, what they can do changes from the time they are two and a half months of age to the time they are five months of age. The investigators in Norway took babies, put them on their back, put douvets or comforters over their head, and at two and a half months, a third of the babies or less than a quarter of the babies are able to get the blankets off their head. By five months, over half the babies could get the blankets off their head, the comforters off their head. So these airway protective behaviors develop. It's not clear to me how this data is being used by forensic experts who seem to know what babies can do to get their airway clear. It nevertheless is important. And these insights are relevant to 14 to 28 percent of deaths that are called sudden infant death syndrome.
PATRICIA TACKITT: Our next section of slides will be on bed sharing and overlay by either adults or children. This is a family, this is their first night home from the hospital, they stayed one night. This is a breast fed baby. Dad knew the baby was in bed. Mom said she put the baby between them and made sure he was aware. He was watching television and drifted off to sleep. When she woke up, he was on top of the baby's face and she felt the foot kick her and the baby was taken to the hospital for three days on a vent and pronounced brain dead and deceased. But they both said they were just exhausted after the labor and coming home that day and he happened to fall asleep on top of her and roll over.
This was a couple summers ago. Grandma was sleeping with her three daughters. She has a fourth daughter who is about 18. This is her second baby that grandma is baby-sitting. Grandpa is under the window, what you see in the far part of the slide, and his head was the same direction as grandma. So there's six people in this bed. The three daughters of grandma and the baby that is right behind her. She never intended to fall asleep, but at 5:00 a.m., and again she's cuddling the baby with the arm over the middle section of the baby at 5:00 a.m. She drifts off and then the eight-year old came -- this looks like a two-year old, but she's trying to take my doll. The eight-year old said, where is the baby? And she said, I raised all four children in bed with me and I would never believe that bed sharing could be unsafe.
This is a mom who recently, again, in the Detroit area, was sleeping with her baby and she cuddled it in her arms and drifted off to sleep. And her mom is taking the position of the male that was with her because he wasn't able to be with us for the scene reenactment. You can tell this is a pillow top mattress. What we found was when mom fell asleep, she was like this. When she found the baby, the baby was like this. But when I always go back after the first set of photos and say, was this exactly how it was? She said, oh, no there was a pillow under my arm. That brings that up into the baby's face which makes it much more difficult for the baby to lift its head and get out of this situation. So it is important to ask that question after you do your initial set of photos.
This is a recent death as well. This is a baby ten months old and 1.3 pounds when she was born at 23 and a half weeks. She was doing beautifully, there were no red flags. Mom went off to do an errand and when she came back two hours later, dad had gotten in bed so he could feed the baby when she woke up. He had drifted off to sleep. When mom came in, this is how she found the baby. He cuddled up to the baby in his sleep, never waking enough to realize what he had done and his arm was over the baby's chest. So even though the nose and mouth were unobstructed, her chest could not rise to breathe.
JAMES KEMP: So this is a pretty controversial area. I think it's important to be informed and it's important to agree with me after that I think. No, it's important to be informed about this information. I think this is the next area, the area of bed sharing is where progress is going to be made in deaths during sleep.
I want to bring up today in a couple of epidemiological studies, the first the Chicago Infant Mortality Study which is a big case control study, 260 cases of SIDS one control. In this study, 132 of the babies died while bed sharing, but the controls, fewer than a third of them died while bed sharing. So half the babies died while bed sharing, but fewer than a third of the controls were bed sharing.
This effective bed sharing is independent of cigarette smoking. But if you add in cigarette smoking, it increases the risk. If there's multiple bed partners or bed partners besides mother, the odds ratio is quite high. And Pat showed you one dose of scene recreations. The odds raise from anywhere from twofold increase to almost an over elevenfold increase. And remember from the Chicago Infant Mortality study, there was 75 additional suffocations that were not analyzed as part of that study.
In St. Louis, the reasons that high risk babies which means poor inner city African American babies, bed share, the mothers put them to to bed with them is that they don't want the baby to die a crib death so they take the baby out of the crib and put it in mom's bed. They believe they sleep better when they sleep with the child which is probably actually not true based on some focus group stuff we have done. In St. Louis the big thing is the mothers couldn't afford a bed, that's why they had the baby in bed with them. And if they lacked confidence in their ability to provide a safe sleep environment, they didn't know how to put the baby down safely or what they thought was safe, they were more likely to bring the baby close to them in bed.
Another important epidemiological study that often gets quoted is a study from the United Kingdon called the Sezne study which is a confidential inquiry into sudden deaths in infants. And they also have a very big case control study where there's 325 deaths and a couple of controls for each death.
Using the parental bed increased the odds of dying to almost ten. The odds ratio was 9.78 with the upper limits of competence was almost 24. This very high risk of dying while using the parental bed became statistically insignificant if neither parents smoked. But, in England amongst the babies bed sharing, 91 percent of them had a mother that smoked. So there was only four deaths out of 325 of a baby dying while bed sharing who had a mother who didn't smoke. So rather than saying it's safe to bed share if you don't smoke, the proper interpretation of this data is that the high prevalence of smokers among the parents of infants who died makes it impossible to derive precise estimates of risk associated with bed sharing for families who do not smoke. Not that it's safe to bed share if you don't smoke, but we can't do the statistics because there were so few babies in our studies that had mothers who didn't smoke who died while bed sharing. This data has been misinterpreted to say what can be done safely and the authors themselves warn us against doing that.
In a big study done with MJ Shears with the Consumer Product Safety Commission, we looked at their suffocations deaths, not SIDS deaths but the suffocations deaths, and we also looked at something called the National Infant Sleep Practices Data Set which is a home -- phones to home -- telephone calls to about 800 babies and it's done a couple of times a year to the homes of the babies. And this allowed us to get a risk calculation for sleep practices in suffocation. The National Sleep Practices data set allows us to have a denominator. In the study, we looked at 340 babies that were reported to the Consumer Product Safety Commission and compared how they slept to over 4,200 to living infants and the mechanism we found amongst babies dying of suffocation was way more likely to be entrapment within the bed or getting the head covered instead of overlying. Overlying happened, but entrapment within the bed was a greater risk factor.
We calculated that the suffocation risk was 40 times higher if babies used an adult bed rather than using a crib. So in the other group, the group that is not called SIDS, the risk of the suffocation is gargantuously higher if babies use an adult bed instead of a baby crib. Babies do die in cribs, but they die of suffocation. It's much more likely to happen outside a crib
PATRICIA TACKITT: Our last section then is entrapment or wedging in beds, sofas or chairs. In this situation dad had gone downstairs for a brief amount of time, maybe 15 minutes, and the baby was left with the bottle propped but still awake. When he had gotten back he found she had fallen off the bed surface and gotten wedged between the wall and mattress. I have about ten of those that happened in the last four years. So it isn't atypical to be able to find that.
This is a mom recently who was sleeping in the bed with the baby. And I don't find too many that happen this way for bed sharing. But this is her first infant, she's a late teen and basically during the night she moved from the side position on to her tummy for the mom. The baby got scooped up against the wall and the mattress at the top of the bed and was found at about seven weeks old unresponsive.
This infant was breast fed and found in the crib about 3:00 a.m. Mom was the loan caregiver. Dad was in the Service and was out of state. Mom had breast fed the baby and then was burping the baby on top of her chest and fell asleep. And the baby was found like this when she woke up to the phone call in the morning, and I asked her to go back and lie down how she was and it was obvious her body was against the baby's nose and mouth. She had a very small bed. Nobody had warned her about the fact that if you fall asleep when the baby is nursing or after the baby is done nursing, but the baby is on top of the chest and falls off. She had done this several other mornings and the baby when it was on it's side was still alive, so she thought that meant it was safe to do. Many of the babies that died were in the exact same positions days or weeks or months before and had lived through them, and so people thought that meant it was safe to continue to do.
This is a mom who is nursing twins. She was told that the babies should be in a crib, but she decided if she slept in this recliner chair, she would be able to nurse the babies, and she nursed them and one of them got turned around in her arm and she went to sleep this way, and during the night one of the babies ended up down here. So many people say to me, but we know it's safe because our baby lived through it. And I like this slide, even though I hate the tragedy that occurred, because one baby lived and one baby didn't. Does it make it a safe thing to do because one baby lived through it? Those babies could have died any of the two months prior that she slept the same way, they just didn't happen to those other nights.
JAMES KEMP: Pat do you have anymore scenes to show?
PATRICIA TACKITT: No.
JAMES KEMP: All right. So entrapment within adult beds was the most common mechanism in the Consumer Products Safety Commission study of suffocation.
I just want to emphasize that a fifth of all deaths in Chicago that happened during sleep and babies were sort of left out of the analysis of the data that was published.
We've done studies comparing the softness of adult beds and clearly there's more soft areas in adult beds causing more subtle entrapment in adult beds than in baby beds. And obviously in adult beds, there's more blankets to cover the head of the baby causing a subtle entrapment if you will. In St. Louis, a third of the babies dying in adult beds were on top of pillows and apropos, one of Pat's scene recreations.
In St. Louis, unsafe adult beds are a very big issue. In a four year study of SIDS, they are -- and accidental suffocation undetermined, we had 119 deaths. All these had a very detailed death scene investigation, and only 8.4 percent of all the deaths in St. Louis happened with the baby on its back sleeping alone with its head uncovered. So there's a big role amongst the 92 percent remaining deaths for safe sleep practices. Of the 119 deaths, only 35 happened on sleep surfaces meant for babies to sleep on. So the vast majority happened on adult beds or on couches. So the cribs seem to be safe, other sleep surfaces are not.
There's a big push in many states and nationally to reduce racial disparity, and this is one area where racial disparity has gotten some attention, and I don't want to just say for the sake of discussion, but I think it's true, that Back to Sleep is not the answer in reducing racial disparity when babies die suddenly and unexpectedly during sleep. The focus of the NIH's table out here, for example, is emphasizing back sleeping amongst African Americans. But African American babies died a much higher rate while sharing adult beds which are softer, but also which can entrap the baby, and there's big people or people in that bed.
Minorities in general such as Mollys in New Zealand and African Americans in particular are not protected by supine sleep when bed sharing. No one has ever shown, no one has ever shown if you put the baby on its back in a shared bed, that it has any effect on the baby's risk of dying.
In St. Louis, African American babies are just as likely to die while bed sharing on their back as in the prone position. So that ain't the answer for the deaths that are happening in St. Louis at least.
In St. Louis, 92 percent of fatalities during sleep involve sleep practices that increase risks significantly. There's a very high rate of bed sharing deaths, and back sleeping is not protective during the 60 percent of deaths in St. Louis that happened while bed sharing.
The question is how many more deaths are preventable? And that's what we like to think about when we talk about these sleep practices. And I leave the P word, preventable, out here because preventable was something when you talked about SIDS that you weren't allowed to use until the last couple of years. I guess I'm concerned as many of the members of the panel are about how recent trends in designating manner of death among SIDS undetermined or accidental suffocation with or without overlying, they mask those deaths which are preventable. If SIDS is seen as being unpreventable, we have to be very careful not to put deaths that have a preventable component into the SIDS category. This calls attention to how the manner designation may affect efforts to reduce child deaths.
This is an ongoing issue in St. Louis. It's on going issue in Michigan, and I'm certain that it's a point of controversy in whatever state you happen to be from, is that if it's called SIDS it seems to be unpreventable, but if the other things are going on, it's scene as being preventable. I think we are done.