Should We Try to Save 22 Weekers?

Dr. John Lantos

35 years ago, the President's Commission on Bioethics issued a report on life-sustaining treatment decisions for children. And they had this two-by-three table in. They talked about how doctors need to decide whether treatment's beneficial, futile, or somewhere in between, the gray zone, and then about whether parents should have the last say on whether treatment should be provided. And the real topic for today is whether treatment for 22-weekers belongs in this gray zone where outcomes are ambiguous or uncertain and where the President's Commission and many bioethicists say in those situations parents' preferences should determine whether treatment's offered or whether it belongs in the futile category in which case we should not offer it. Can get some evidence. This is one of the few areas where we actually have some data on where neonatologists collectively think about this treatment. This was data from an NICHD Neonatal Research Network study that was published in the New England Journal a couple of years ago, looking at which babies received active treatment by gestational age. And it's sort of a complicated graph, but the y-axis is the percent of the babies who got active treatment, and then it's by gestational age, 23, 24, 25, 26, each dot represents one center and the bars are the standard error bars. So 24 to 26 weeks in these 24 hospitals, almost all babies at almost all centers received active treatment. At 23 weeks, some centers treated every single baby, but about two-thirds of centers seemed to personalize the decision or to have variation in the decision. and the range of babies who got active treatment went from 100 down to about 25%. The biggest variation was in the 22-weekers, where five hospitals didn't treat any babies born at 22 weeks. And remember, these are the elite hospitals in the Neonatal Research Network. Seven hospitals treated every baby born at 22 weeks, and the rest of the hospitals were somewhere in the middle. So collectively, it would seem, doctors are uncertain about whether treatment for these babies is beneficial, futile, or somewhere in between. And it's pretty clear from that data that in the United States today, there's no standard approach to babies born at 22 weeks. And by that first graph from the President's Commission, one could conclude that when there is this sort of professional disagreement, Parents' values should prevail. We should offer treatment, saying some people think it's a good idea, but we should be willing to forego treatment because many people think it's a bad idea. What is the data that the people who think treatment ought to be provided use to say that such treatment is not futile? Well, in this same study from the Neonatal Research Network, Overall survival for babies born at 22 weeks was just 5%. And some people say that's pretty good evidence of medical futility. However, if you looked at the babies who got active treatment, the overall survival rate was 23%. That looks a little bit better and gets you sort of away from the idea that this treatment is absolutely futile, and one center we just heard about current outcomes, which I'll show you a minute from Iowa, but at the time that this study came out, Iowa was reporting 48% survival, vastly different from the rest of either the neonatal research network or the centers that report to Vermont, Oxford, which are shown in those other two bars

Dr. John Lantos

down here. And in a paper just published last month from the Iowa group, they're now reporting 70% survival for babies at 22 weeks, 82% at 23. And most of the babies, as you can see in the column over there on the right, do not have severe neurodevelopmental impairment. But I'll get back to that in a minute. How do they do it in Iowa? Again, I'm not a neonatologist, but it seems like there are some features of the Iowa program that have not been studied, but that the people there seem to believe are important. The first is collaboration with their maternal fetal medicine specialists so that all women with labor and impending delivery start getting antenatal steroids starting around 20 or 21 weeks.

Dr. John Lantos

There's discussion about C-section if indicated, even at these low gestational ages. There's parental informed consent for NICU treatment. They have what they call a golden hour protocol, which I'll show you in a minute, and then there's a tiny baby unit within the NICU where all the health professionals spend their entire professional lives caring for these very tiny babies. complicated slide the details for my purposes are less important than the fact that there is a protocol there is proactive and anticipatory planning for the treatment of these babies there's attentiveness to psychosocial features they take pictures of the baby the mom gets to see and touch the baby they start lines they give antibiotics they do exquisite temperature control they have guidelines for co2 management and again I'm not focusing on the details of the medical management, because my point here is that we don't know which aspects of this treatment work. But what seems to work is having a plan, giving some thought to what to do, and not be in the situation that many doctors in many hospitals are in when a baby's born at 22 weeks, where people go, oh dear, what are we going to do? We don't have a plan. Should we resuscitate? Should we not? Should we talk to the parents? How do we deal with it? Having a plan seems to help.

Dr. John Lantos

The whole country of Sweden reports similar results. Here is a complicated graph, but for my purposes, just look at these four bars, which are babies born at 22 weeks. The first two are in the period 2004 to 2007. Actually, the first and the third, the second and the fourth are in a later period. In 2004 to 2007, overall survival was 10%. But among the babies who were admitted to the NICU, just like in the NICHD study, many more survived. Take-home lesson, if you have a tiny baby and don't admit them to the NICU, they're going to die.

Dr. John Lantos

If you do, their outcomes are much better, and as of 2016, survival rates in the whole country of Sweden for babies who were born at 22 weeks and admitted to neonatal intensive care units is now 60%. The common element, as I say, is not any particular type of treatment. Other centers, Cologne in Germany, the whole country of Japan, report increasing survival rates at this low gestational age. Everybody does things differently. In Iowa, they use jet ventilation from the first breath. In Cologne, they give less invasive surfactant and put babies on CPAP. We don't know what works, but it does seem that having a protocol is better than not having one. So here's what makes this really interesting for me as a bioethicist rather than as a neonatologist. I would view these as very promising preliminary results. Here's a disease being born at 22 weeks, which previously had been thought of as almost universally fatal. And now some centers are starting to report that they're getting very promising early preliminary results. And in most situations in medicine, when that happens, people say, wow, let's find out what they're doing. Let's go study it. Let's try it. Saving people's lives is better than letting them die. Here's some promising results. Let's do it. And what seems to be happening with babies at 22 weeks instead is most other centers don't want to try it. The NICHD doesn't seem to want to study it. And professional societies even misrepresent the data in developing guidelines for such treatment. So here's a statement that came out from the American College of Obstetrics and Gynecology, Society of Maternal Fetal Medicine, after the Rosavi study, the New England Journal study that I just showed you, citing that study in making the claim that delivery before 23 weeks

Dr. John Lantos

results in neonatal death irrespective of newborn resuscitation, even though 5% survived when there was no resuscitation, 25% when there was, and significant morbidity among survivors is universal. The study showed that 60% of survivors didn't have severe impairment. So how you can write a statement citing a study that has data that contradicts exactly what you're saying is mind-boggling to me. So it's a mystery. Is there any other situation in medicine where patients have a disease that's almost uniformly fatal? Some centers report 40, 50, or now 70% survival. Other centers don't offer treatment. Some say it's unethical to offer treatment. And many bioethicists support them. It's hard to think of one. It's really weird.

Dr. John Lantos

But I know what you're thinking. Like all the survivors must be severely disabled, right? And survival rates are only one aspect of this, and that's just wrong. Here were the outcomes for babies in the 24 centers of the neonatal research network. Among the babies who were treated, as I said, at 22 weeks, 23% survived. About a third of the survivors had severe impairment. That means two-thirds of the survivors didn't have severe impairment. The data are a little complicated though because people report results using different denominators. Again, the question is do you report outcomes for all live births or do you report outcomes for live births where people actually admitted them and treated them in neonatal intensive care units? And then if you report the results, what outcome do you use? Often in neonatal studies and, for example, as I'll show you in a minute in the Neonatal Research Network Outcomes Calculator, people use a combined variable of death or severe neurodevelopmental impairment as sort of a total lump sum of bad.

Dr. John Lantos

But you can also report overall survival without severe neurodevelopmental impairment or rates of neurodevelopmental impairment among survivors. And those yield very different results. Here's an example. If you go to the website for the Neonatal Research Network outcome calculator and you put in 500 gram, 23-week singleton,

Dr. John Lantos

They don't go down to 22 weeks, and you put in boys and girls, and steroids are not steroids. You get this result for 23-week singletons who survive unimpaired, 5% for boys whose moms didn't get steroids going up to 18% for girls whose moms did get steroids. Couple things to notice about this, girls whose moms get steroids do four times better than boys whose moms didn't, so to say, what's the survival rate at 23 weeks is Not an adequate question, that was really the point of the study that published these data. But look what happens if instead of asking how many 500 gram 23 week singletons survive unimpaired, you ask a slightly different question. How many 500 gram 23 week singletons who survive are unimpaired? This is using the same data set, although this data is not available on the Neonatal Research Network neonatal calculator as a number. You have to calculate it yourself by subtracting out all the babies who died and then looking at the rates of neurodevelopmental impairment among survivors. If you think of this as a problem for informed consent and you want to give parents accurate statistics on outcomes, which statistic should you give them? Here's a way to think about this just as a simple thought experiment. Imagine 100 babies, and that should say, instead of 90 day, 90 of them die, so 90% mortality. And among survivors, three have severe neurodevelopmental impairment. You can say two totally true, totally factual things about this. One is that 93% of babies born like this either die or have severe neurodevelopmental impairment. Or you can say 70% of survivors have no neurodevelopmental impairment. Both are true. They might lead to a very different decision by both doctors and parents. Another quirk, studies usually don't account for non-treatment substandard treatment or decisions to withdraw life support. Here's an example from the Epicure study, the big study in the UK on outcomes. Among 22-weekers, they had 152 live births. 6% of the moms got steroids. Fewer than half were delivered in a tertiary care center. Only about a quarter got active treatment. And of those, only about half made it to the NICU. And overall survival, they reported, was three out of 152 babies, or 1.5%. That sucks. And anybody who reads that would say there's no point in offering this treatment. even if one-third survived without major morbidity. But this would be a little like reporting outcomes for leukemia, where you said these babies were treated by general practitioners in their offices with vitamin C, and they didn't do very well. If you don't give babies state-of-the-art treatment when they have a severe life-threatening disease, that's not a good measure of whether the treatment is effective. It's a measure of whether you're providing the treatment.

Dr. John Lantos

Bottom line is, in all of these studies, most of the babies who survive do pretty well, but you may not know it from the way the data are usually reported. What do parents want? That is, if we think this belongs in the zone of shared decision making where parental preferences should determine the outcome, what do you think parents would say if you gave them accurate outcome statistics? Well, we know from some studies, and actually there are a number of studies comparing the attitudes of health professionals with the attitudes of parents. And most look something like this. This was a study where parents of extremely low birth weight babies, control parents, doctors and nurses who worked in the NICU were asked to agree or disagree with the statement, I believe an attempt should be made to save all infants regardless of birth weight. Most doctors and nurses strongly disagreed with that statement. Most parents agreed or strongly agreed with that statement. Although parents are much more variable than doctors and nurses, and about 25% did not agree with that statement. So the only way to know what a particular parent would think would be to sit down, talk about the outcomes, and ask them. Most parents say we should try to save babies at all costs and are less troubled by disabilities than health professionals. This was a study that asked both health professionals and parents to rank these three states. Which do you think is worse, for a baby to be dead, for a baby to survive with severe global impairment, which in the study they defined, as you can see, or to survive with moderate global impairment. So think in your mind how you would rank those now. Which is worse, death or severe global impairment?

Dr. John Lantos

Most doctors and nurses say severe global impairment is worse than death. Many fewer parents say that. We can say maybe they just don't understand. That may be true, although these studies have now been done prior to birth, after birth, a year after birth, 10 years after birth, and Saroj Sehgal now has a book out 30 years after birth. And it seems that at the very least, these preferences are durable. Whether they're durable and just simply wrong is the sort of question that only a philosopher could answer. but it seems that if we believe parental preferences should guide treatment, we should recognize our own biases about the value of life with severe disability or our own beliefs may not be the same as parents. One last concern people have about this care is that it may just cost too much. Again, this is simply a worry that the data don't confirm, at least not by standard measures of medical cost effectiveness. Here's the sort of data that's been gathered looking, and this hasn't specifically been done for 22-weekers, but the epidemiology and the reason why costs are as they are, I think, would apply even more to 22-weekers than to bigger or older babies. This was a study of cost effectiveness from Lex Doyle's group in Australia. The key finding, the details on this one, they broke it down by different birth weight increments, gestational age increments, different measures of cost-effectiveness. However you did it, treatment of tiny babies comes out somewhere between five and ten thousand dollars per quality adjusted life year, dollars per qualia. A study from economists at Harvard who weren't even involved in neonatal care got about the same number, six thousand dollars per quality adjusted life year. Compare that to what the Harvard economists calculated was the value of things like pap smears at $17,000 per QALY or treatment of hypertension at $17,000 or bypass surgery, which is $40,000 per QALY. By whatever measure you use, the treatment of tiny babies looks to be remarkably cost effective. much more cost effective than adult intensive care. So old people in ICUs don't do so well. The cost effectiveness of adult intensive care is really bad. And the reasons for that are shown in this slide. This is a graph that shows ICU patients and NICU patients. And this is the chance of survival with each passing day that you're in the unit. So if a baby is in the unit on day one, and this was babies under 750 grams

Dr. John Lantos

15 years ago, chances of survival were low. But with each passing day that a baby's in the NICU, the chances of them surviving goes up. Most babies who die, die quickly. If you're old, and you're in the ICU and you're on a ventilator, with each passing day, your chance of getting out of the ICU alive goes down. So the net result is most dollars spent on old people in ICUs are spent on people who are going to die. Most dollars spent on babies in NICUs are spent on people who are going to survive. So if you're concerned about cost effectiveness or if you're talking to policymakers who are You should present them data like this. If you look at a 22-weeker with APGARs of three and six compared to an 85-year-old who comes to the ER in the middle of an MI whose survival rate at best is about 15%, and then tell them that in Iowa for the 22-weekers you're getting 70% survival,

Dr. John Lantos

For policymakers, the key lesson is Sutton's Law. Some of you may know the famous American bank robber, Willie Sutton. When he was asked, why do you rob banks, he said, because that's where the money is.

Dr. John Lantos

If you're going to do cost effectiveness, go where the money is. So a modest proposal would be to evaluate treatment for 22-weekers the way we evaluate every other treatment in medicine. That is, if people are starting to get promising results, try to figure out what's working and emulate it. Be prepared to treat these babies if you're going to develop a program and in deciding who to resuscitate, listen to the parents, examine the patient and make individualized decisions based on the best data and the best clinical judgment. Let me just finish by saying there are a few elephants in the room in talking about this, and here are three, institutional culture, abortion politics, and then the artificial placenta as a disruptive technology. Institutional political culture, if you're gonna ask parents and consider this to be a decision that's in the domain of optional treatment, then you got to have a program in place to provide the best available treatment. And if antenatal steroids are part of that, that requires collaboration with your maternal fetal medicine colleagues. Because to offer parents treatment but not give them the treatment that's going to give the best possible outcomes is, I think, unprofessional and sort of a cruel charade. If you're gonna do this, it requires an institutional commitment with collaboration between NICUs and OB, steroids routine after 20 weeks of gestation, and tiny baby units in the NICU that have expertise in providing such care. Second elephant in the room here is abortion politics. This may be more relevant in the United States than in other countries, but the Roe v. Wade framework says babies are not viable until the end of the second trimester, which is 24 weeks. If babies are surviving at 22 weeks, that throws that whole framework into question. Although it sort of asks the question the wrong way. It doesn't say are babies viable at earlier gestational ages. Some people say we shouldn't save babies at earlier gestational ages because otherwise It will restrict abortion access. As an example of this in popular culture, when the Rosavi paper was published in the New England Journal of Medicine, the New York Times wrote an article about it. Their headline was, preterm babies can be viable at earlier birth, but their sub-headline was not, as I think it should have been. This is an amazing breakthrough in neonatal intensive care. Instead, their sub-headline was, study could affect the debate on abortion. Final elephant in the room, babies born in bio bags. Most of you have probably seen these experiments that are being done at the Children's Hospital of Philadelphia so far just on sheep, but they take the sheep at the equivalent of 22 or 23 weeks gestational age, anesthetize the mom, do a C-section, immediately put in UV, and UA lines and essentially put the fetal lamb in a bag of amniotic fluid, use the fetal heart as the pump, so it's like an ECMO circuit without an ECMO pump. They can put nutrients in through the circuit. They grow these lambs up from the equivalent of 23 weeks to the equivalent of 32 weeks. And then for delivery, you just open the Ziploc bag and the lambs go

Dr. John Lantos

scampering out into the field looking happy. It's generated some great headlines like this one.

Dr. John Lantos

Oh, yeah. Conclusion, survival rates improving, non-treatment is a self-fulfilling prophecy. Most parents favor treatment. Why not study it, find out what works, and with parental permission, try to save more babies. Thank you very much.