Research Friday: Does intervention prolong life?

Kosho T & Carey J (2016)

For our third Research Friday post, we thought we'd look at a recent paper that addresses a question that many physicians and parents often ask, and over which even more frequently disagree: does medical intervention actually improve survival in children with trisomy 18 (or trisomy 13)?

Parents who request interventions for their children often encounter opposition, commonly because physicians consider these interventions to be of no, or minimal, benefit for these children. In our second meeting with our own neonatologist (who had firmly stated her opinion that any intervention, beyond possibly a feeding tube and oxygen, was too aggressive), we asked:

"Would you be willing to change your mind if treatment improved survival by 20%?"

She replied: "It doesn't."

So, according to this recent paper from Kosho and Carey (authors who, admittedly, may have some pro-intervention bias) - who is right?

About the paper

The full paper is available, for a fee, here:

Does medical intervention affect outcome in infants with trisomy 18 or trisomy 13? (Kosho T & Carey J, 2016)

An abbreviated version is available free on pages 6-7 of the recent newsletter (PDF) of the International Trisomy 13/18 Alliance - an enormously helpful organisation whose website is full of good research that is always worth checking out.

The paper focuses primarily on two recent studies. I hope to look at one of these, Meyer et al (2016), next week: a US population study of survival of children with trisomy 13 and trisomy 18. The other, Subramaniam et al (2016), looks at a cohort of infants with trisomy 18 in Alabama.

There is also a brief overview of other literature regarding interventions and survival for children with trisomies 13 and 18.

Criticism of Subramaniam et al (2016)

Subramaniam was an interesting paper (and I will try to get hold of a copy) as it is one of the very few studies to conclude that aggressive intervention does not prolong life. It did conclude that when you compared infants who received no intervention to those who received some interventions, the intervention group did better (had longer lives). However, it subdivided the intervention group into those who received 'aggressive' and those who received 'non-aggressive' interventions, and found that each subgroup had the same median survival length; they therefore concluded that aggressive interventions do not extend life in children with trisomy 18.

Kosho and Carey had some criticisms of this paper: notably, that the author's extensive definition of aggressiveness included any corrective surgery, feeding tube placement, cardiopulmonary resuscitation, TPN and mechanical ventilation aimed at prolonging life. The cases were all described. Kosho and Carey decided to look at them themselves, dividing the cases into 'palliative care at birth and no mechanical ventilation' and 'no palliative care at birth, with mechanical ventilation'. The survival then was 8 days in the former group and 399 days in the latter group. A significant increase in survival for 'aggressive intervention'!

Does this indicate that 'full' treatment improves survival but 'semi-active' achieves little?

This is interesting. I have seen myself how 'semi-active' treatment seems to achieve very little. Semi-active treatment often occurs because parents want full treatment and hospitals don't want to give it (hence the half-way house) or because parents themselves elect for this approach, believing that full treatment is too invasive, but not wanting full-on palliative care.

For some children this approach meets their needs adequately, but for a lot, it seems to leave them in a limbo of inadequate care, with the child never getting well enough to go home. They are the cases that doctors characterise as 'active treatment' at conferences when they say that treatment becomes a crusade and forces children into suffering. Often, this middle ground is where the suffering is.

Children who have high-level treatments suffer, sure, but often less than children who receive insufficient treatment for their needs. For example, if a child is struggling to breathe and really needs intubation and mechanical ventilation but is given CPAP, this extends the child's suffering rather than solving the problem.

These two different interpretations of this study's results perhaps suggests that 'half-way house' care does not extend survival, but full care does make a difference. This would be in line with the Goc et al (2006) study, which also suggested that intervention did not extend survival markedly (although their results are highly questionable) but withheld 'aggressive' treatments from children with a confirmed diagnosis.

Kosho and Carey on Meyer et al (2016)

The Meyer study is a population-based study from the US, where routine treatment for children with trisomy 13 and trisomy 18 is becoming increasingly common, though not universal.

Perhaps unsurprisingly if you believe that treatment improves outcome, this study had the highest reported survival rates for children with trisomy 13 and trisomy 18 yet seen in a population study: 13.4% for T18 at 12 months and, even more surprisingly, 12.4% at 5 years. These figures certainly deserve a closer look (next week!) but they are very good survival figures, particularly for a population study for an area in which there is not a standard approach to treatment, and for which some infants will, by parental choice or otherwise, have received full-on palliative care.

Do these figures reflect the increase in interventions, and can we conclude from this improvement in survival that the increase in interventions is directly leading to an increase in survival, or is this simply a coincidental finding? Kosho and Carey suggest in their paper that the increasing survival of children with trisomy 13 and trisomy 18 in population studies (they point out the high survival figures of Wu et al (2013)) suggest that interventions are prolonging life in this group.

Can we confidently state that the high survival in Meyer (2016) suggests that intervention improves survival?

I could buy this in the Meyer study. My experience suggests that increasing numbers of children in the US are having interventions, therefore we could link these high five-year survival rates to the interventions.

However, I just don't buy that increasing numbers of children in the UK (Wu et al is a UK population study) in the period studied were receiving full interventions - or any interventions at all, for that matter. I think the Wu et al high survival statistics deserve closer analysis and, in fact, may be the key to establishing whether the Meyer results are really down to the interventions or whether we are in danger of confusing correlation with causation in this case.

My view is - and I say this having not yet spent long looking at the Meyer study, and it's a while since I spent time on Wu - that Kosho and Carey are in great danger of doing exactly this. I do not think we can safely say that children are surviving longer because of interventions based on increasing survival in population-based studies.

Factors that may contribute to longer survival in population-based studies may, for example, include the improvements in prenatal screening, which mean that many fetuses are simply removed from the studies by virtue of termination of pregnancy after diagnosis. Fetuses who present with more, or more severe, medical issues are perhaps over-represented in this group. Thus, those who remain undetected are more likely than ever before to be the healthiest children, and this translates into better survival statistics.

There are likely to be other factors also contributing to increasing survival in population-based studies. I would argue that Wu et al in many ways discredits Meyer. It shows that the increase in survival in Meyer may simply be due to the fact that the general trend is for longer survival, regardless of intervention.

Review of other evidence and conclusions from Kosho and Carey

The paper then goes onto look at the other evidence for treatment, including the increasing evidence base for cardiac surgery in improving survival for children with trisomy 13 and trisomy 18. They quote a number of both familiar and less familiar papers, although not in detail, and there is no discussion about whether cardiac surgeries were performed later or earlier, or selection criteria, etc.

They quote an interesting paper, Nishi et al (2014), which looked at oesophageal atresia and showed that children with trisomy 18 who had full repairs had better one-year survival figures than those who had palliative surgeries. They also refer to Nelson et al (2012), which does show increasing intervention for children who have trisomy 13 or trisomy 18, but does not relate the interventions to survival. They of course refer to Kosho's own work of 2006, which remains one of the most useful research papers to cite regarding the impact of intervention overall.

Kosho and Carey conclude in this paper that "the body of evidence indicates that intervention does make a difference in survival in infants with trisomy 13 or trisomy 18", although cautiously adding that this makes it a reasonable option to discuss with parents - hardly a ringing endorsement.

Closing thoughts

My thoughts are that, in 2016, we haven't really moved forward very far from 2006. 10 years on from the original Kosho paper, we have a lot more literature to get to grips with, but really we can still only say from the limited evidence that intervention does seem to prolong life in children with trisomy 13 and trisomy 18. But the evidence remains limited in quality and far from convincing. We don't seem to have made great strides in establishing which interventions help which children; in understanding the risks of cardiac surgery for individual children with trisomy 18; in developing a good understanding of the unique needs of these children.

As to this paper, it is a useful run through the literature. A good starting point for families exploring intervention, and a useful rebuttal of the Subramaniam paper (which will no doubt be used against families), but it does not really answer the question conclusively as to whether intervention prolongs life in children with trisomy 18. I think that is because this question is unanswerable at the present time.