Informed choices

Web sites and newspapers have been full of speculation this week about the rising rate at which babies with Down syndrome are being born in England and Wales and the reasons why this might be. The National Health Service quickly contradicted initial reports, claiming the rise was not as high as was being claimed. It also claimed that the suggestion that the rise was due to changing attitudes was wrong. So what is going on? Is the birth rate rising, by how much and why? Why has the claim that attitudes are changing drawn such criticism?

Selected media coverage

• BBC Online: Society more 'positive on Down's'
• Times Online: Down's births increase in a caring Britain
• The Independent: Down's syndrome: parents think again
• Sky News: More Parents Keep Down's Babies
• Mail Online: More mothers are choosing to keep their babies when diagnosed with Down's Syndrome
• The Sun: Down's births are on the rise
• The Independent: Shame on the doctors prejudiced against Down Syndrome
• Mail Online: My daughter shows why we should celebrate Down's syndrome children - NOT abort them

Stories about rising births and changing attitudes emerged last Monday in conjunction with a BBC Radio documentary. The Down's Syndrome Association (for England, Wales and Northern Ireland) issued a press release that, along with comments in TV and radio interviews, was widely quoted in the media and drew attention to a survey that the Down's Syndrome Association claimed supported the interpretation that the rising birth rate was evidence of changing attitudes.

The data

Data about babies diagnosed with Down syndrome has been collected by the National Down Syndrome Cytogenetic Register (NDSCR) since 1989. Their most recent annual report includes data up to 2006 and was published earlier this year. Data for the most recent year is still "provisional" because of a higher number of unknown outcomes of pregnancies recorded as being prenatally diagnosed with Down syndrome. However, we can make reasonably accurate assumptions about these 'unknowns' and thus estimate the likely total live births for 2006. As the NDSCR report states:

"Assuming that their proportion terminated remains as before 2006, the likely number of Down syndrome live births in England and Wales in 2006 would have been 767 (46+703+6% of 293), a prevalence of 1.2 per 1000 livebirths occurring in England and Wales in 2006."

A total of 767 births is 7% higher than the 717 births recorded in 1989 when national statistics were first collected and serum screening tests were first introduced.

Analysing the data

The data collected by NDSCR includes an increasing proportion of unknown outcomes in recent years. (I understand that this is mostly due to difficulties obtaining records from a single private clinic in London). To look at trends over time, we therefore need to estimate what the actual outcomes were for this diagnosed pregnancies to avoid distorting the overall picture. This is not too difficult to do and is exactly what I did for the careful analysis of the NDSCR data in an editorial by Sue and myself that was published in Down Syndrome Research and Practice in October.

Of course, the actual number of live births of babies who have Down syndrome tells us little about trends over time - they could go up or down just because all births in a fluctuating population go up and down over time. The proportion of all babies who are born who have Down syndrome (the live birth rate) is what we need to look at.

In our editorial, we chose to look at data from 1992, when maternal serum screening became the most common screening test in use (replacing maternal age only screening). Since then, prenatal screening has been universally introduced (to include mothers of all ages - not only those over 35). During this period, however, the live birth rate has risen from 9.2 in 10,000 to 11.5 in 10,000 in 2006 - a rise of 25% (graph).

Explaining the data

So, a careful analysis of the data makes it clear that the rate at which babies who have Down syndrome are being born (as a proportion of all babies born) is rising and rising fairly substantially. But, what does this tell us about why this might be and what the consequences of genetic screening policies are?

Changing attitudes?

In the press release preceding the media interest this week, the Down's Syndrome Association claimed that it had "consulted 1,000 of its members on why they had decided to continue with their pregnancies after being given a high-risk pre-natal screen for Down's syndrome". It is unlikely many of those responding to the survey actually had a definite diagnosis following a "high-risk" screen. By my calculations, between 1992 and 2006 there were 1,023 pregnancies in England and Wales where parents continued with the pregnancy after a diagnosis confirmed by CVS or amniocentesis.

Most respondents, therefore, must have had a "high-risk" screen result alone to guide their decisions. During this period a screen result considered "high risk" would been one with odds higher than 1 in 250 or 1 in 300. Many of those responding were therefore likely to have been given odds that they might quite reasonably have thought to be low enough to be of little concern. It is perhaps therefore not surprising that the "almost half of those questioned said they did not think they would have a child with Down's syndrome and that's why they continued." Therefore, almost half of respondents' decisions had little to do with attitudes, but more to do with the imprecision of screening tests. This survey also only offers a selective snapshot of current attitudes. It does not appear to directly say anything about changes in attitudes over time.

Maternal age?

The difficulties about drawing the conclusions from this survey that were being drawn in much of the media quickly led to a statement from the NDSCR insisting that it is "incorrect to report that more women are declining antenatal screening for Down's syndrome" because of rising births. A similar statement was put out by the National Health Service.

Criticisms of the 'changing attitudes interpretation' of the data were echoed in articles in The Telegraph, The Independent and by Ben Goldacre in his Bad Science column in the Guardian, among others.

Most of these criticisms point to maternal age as the key factor influencing rising birth rates.

So, are rising birth rates all down to changes in maternal age?

It is certainly true that the primary factor known to influence the chance of giving birth to a baby who has Down syndrome is a mother's age. The chance is around 1 in 1,480 for mothers aged 20 years. By the age of 25, this has increased slightly to around 1 in 1,340. By the age of 30, the chance has increased to 1 in 940 (11/10,000 live births). Between the age of 30 and 35, the chance increases nearly threefold, reaching around 1 in 350 (28/10,000 live births). Between the age of 35 and 40, the chance more than quadruples, reaching around 1 in 85 (118/10,000 live births). Live birth incidence is therefore clearly dependent on the maternal age distribution at birth and particularly sensitive to the distribution among births to mothers aged over 30 years (graph).

Between 1989 and 2003, the percentage of all births in England and Wales to mothers aged 35 and over increased from 9% to 19%. It can be calculated that between 1992 and 2006, the natural (if no selected abortions) live birth rate rose 50% - from 14 in 10,000 to 22 in 10,000 (graph). Meanwhile, the proportion of all diagnoses that were identified prenatally increased from 44% to 60%. Most (91%) pregnancies that are conclusively diagnosed are aborted.

Negative attitudes?

Critics of the 'attitude change' theory suggest that the rising prenatal diagnosis rate and the high abortion (following invasive diagnosis) rate are evidence that attitudes are as unfavourable to unborn babies who have Down syndrome as ever.

Perhaps. However, an increasing prenatal diagnosis rate could be (at least partially) due to an increasing shift to first trimester screening tests (from second trimester). Babies who have Down syndrome are more likely to be naturally lost during pregnancy than babies who do not have Down syndrome. Therefore many of the babies who have Down syndrome and alive towards the end of the first trimester (~40%) do not survive to term. From the end of the second trimester around 20% survive to term. Therefore, earlier screening detects many more babies who have Down syndrome, but many more of them would not naturally survive to term, so it does not necessarily increase the rate at which the births of babies with Down syndrome are 'prevented'. The medical screening establishment has determined that more parents having to make life or death decisions about their babies in the first trimester is somehow preferable to fewer parents having to make these decisions in the second trimester.

Another problem with relying on the rising prenatal diagnosis rate to argue that attitudes are not changing (or even that attitudes are becoming less favourable towards babies who have Down syndrome), is that public policies have shifted during the period in question. Only in recent years has it been government policy that all hospitals offer prenatal screening to all women, regardless of age. Throughout the period in question, screening has continued to become more widely available and one might suppose this has something to do with an increasing prenatal diagnosis rate.

The 91% 'abortion rate' is on face value startling. However, it is virtually irrelevant as an indicator of parental choices. Indeed, it might be expected to be nearer 100%. The quoted rate is the proportion of parents choosing an abortion given a firm diagnosis. By this stage in the screening process, the parents have (1) chosen to opt into prenatal screening; (2) received a positive screening test result; and (3) chosen to accept an invasive diagnostic procedure that carries a 1%-2% chance of causing a miscarriage in order to obtain a firm diagnosis. It might be presumed that if the consent for the invasive diagnosis is 'informed' then it is based on the intention to abort given a firm diagnosis.

Unfortunately, we do not collect national figures on screening uptake, but figures for hospitals in London seem to range between 60% and 80% uptake. Of those who screen positive, not all women choose to accept an invasive diagnostic test. Again, we do not collect national figures on screening uptake, but figures for hospitals in London seem to range between 50% and 70% uptake. Put another way, therefore, maybe 80% x 70% x 90% = 50% of women make all of the choices necessary to selectively terminate a pregnancy diagnosed with Down syndrome.

So, where does this leave us? Rising prenatal diagnosis rates and a high abortion rate do not necessarily contradict a positive shift in attitudes towards unborn babies who have Down syndrome. However, they are inconsistent with a dramatic shift towards more positive attitudes. It is certainly clear that a rising expected natural live birth rate (due to rising maternal ages) significantly limits the extent to which shifting attitudes might explain the rising live birth rate. Unfortunately, the survey put forward in support of this week's media stories provokes more questions than answers about trends in attitudes towards unborn babies who have Down syndrome.

Integrity

Issues of prenatal life and death tend provoke intense debates. However, even in the heat of these arguments, I'm sure most would agree that it is important that every effort is made to be as accurate as possible about the facts (and the evidence supporting them). It is particularly incumbent upon public health authorities (and other organisations from whom people may seek trustworthy information) to be rigorous about the accuracy of statements they issue.

It is therefore particularly disconcerting that the UK's National Health Service and the NDSCR should (knowingly) issue statements quoting misleading figures. As noted above, the 4% rise in live births since 1989 quoted by the NHS and NDSCR is wrong, unless you ignore the distorting effects of a large number of unknown outcomes in the provisional data for 2006. By NDSCR's own figures, the correct figure is 7%. The increase in live birth rate over this period was 10%. The explanation I was given was that the NDSCR thought that journalists would not understand why the raw data had to be adjusted to present an accurate reflection of what (most likely) actually happened. So NDSCR and the NHS would rather distort the truth than have to explain things properly to journalists? Not good enough.

Informing choices

If the NHS cannot be trusted to get its own numbers straight, then can it be trusted to ensure 'informed consent'? As acknowledged in the guidelines for antenatal care, recently updated by the UK's National Institute for Health and Clinical Excellence, prenatal genetic screening "should start with the provision of unbiased, evidence-based information about the condition, enabling women to make autonomous, informed decisions".

There are certainly healthcare professionals who think only consent to prenatal screening can possibly be informed (that only the ill-informed refuse). The NHS web site talks about "why antenatal screening is necessary" (rather than "why antenatal screening may be chosen"). The NHS' so-called 'best practice' information leaflet devotes less than 3 pages to informing parents about Down syndrome. Most of this limited information is about maternal age and chromosomes rather than people. People who have Down syndrome are not featured talking about how they view their lives. Nor are their families. There are no photographs of people who have Down syndrome. This 'best practice' guidance then devotes 12 pages to describing the screening and diagnostic process in detail. The message? On balance, parents should be far more interested in learning about the screening process than the condition actually being screened for.

Of course, simply having a national screening programme for Down syndrome tilts the balance. Surely, if the NHS goes to all this effort and expense to offer screening it is something people should want to do?

I think Dominic Lawson was right to point out in his commentary this week that there is a common (and ill-informed) prejudice against people who have Down syndrome among the medical community. Some of the poor attitudes uncovered by a survey of American mothers of children who have Down syndrome are not dissimilar to those reported elsewhere, including the UK.

In the USA, the recently adopted Prenatally and Postnatally Diagnosed Conditions Awareness Act aims to improve the availability of balanced information and access to support providers. The Act provides for funding with an "emphasis on funding partnerships between health care professional groups and disability advocacy organizations" to ensure that "up-to-date, evidence-based, written information concerning the range of outcomes for individuals living with the diagnosed condition, including physical, developmental, educational, and psychosocial outcomes" and "contact information regarding support services, including information hotlines specific to Down syndrome or other prenatally or postnatally diagnosed conditions, resource centers or clearinghouses, national and local peer support groups, and other education and support programs".

There is a case to be made for similar legislation to be introduced to strengthen requirements for balanced information in the UK and many other countries.

Public policies

To some, individual choice seems to trump all other considerations. Personal freedom is sacrosanct. To the medical screening establishment and the prenatal screening industry, 'informed choice' is the ultimate disclaimer - the outcomes of the policies cannot be the responsibility of the policymakers because all they are doing is facilitating informed individual choices.

This is obvious nonsense. We do not facilitate prenatal selection of individual babies for any old thing. Just Down syndrome (and a few other conditions) are targeted by (current) public health policies. We do not (officially) leave sex selection to 'informed choice'.

Government, the medical screening establishment and the prenatal screening industry therefore carry responsibility for the outcomes of their policies. These national public health policies impact every one of ~700,000 pregnancies annually. Around 8% of all women screened are panicked by a 'positive' screen result. Barely 1 in 20 of these women are actually carrying a child with Down syndrome. As a direct result, many babies who are not affected by Down syndrome are subjected to invasive diagnostic tests. These test carry risks and therefore some of these babies are lost as a result of the testing. Our recent estimates suggest that around 400 babies who do not have Down syndrome are being lost due to invasive testing each year in England and Wales as a consequence of public health policies that 'achieve' a reduction of 660 live births of babies who had Down syndrome.

What are 'disabled lives' worth?

Some have suggested that this is an 'unhelpful' or irrelevant consideration in the context of debates about genetic screening. To me, this gets to the heart of the matter. What does the willingness to accept the loss of large numbers of babies who do not have Down syndrome in the quest to eliminate those who do say about society's view of the value of the lives of people born (by virtue of their genetic makeup) with mental abilities on average lower than most and with additional risks of (significant but mostly successfully treatable) health concerns?

And this begs the next question: where will it stop? Today Down syndrome is apparently not worth living with. Tomorrow will it be dyslexia? Or, maybe just a propensity towards antisocial behaviour?

Down syndrome is an easy target. Spotting a whole extra chromosome is a relatively low tech exercise these days. High tech diagnostics promise to widen the net in the near future.

Modern science is rapidly identifying many thousands of genetic variations associated with health conditions and mental characteristics. The costs of personal genome sequencing are plummeting (and the speed of sequencing accelerating). Vast effort (and considerable public expense) is being deployed to perfect the (non-invasive) extraction of a sample of a baby's DNA via a simple maternal blood test. When these rapidly evolving technologies converge, the game will fundamentally change.

The absence of political debate is worrying. As commented elsewhere about existing screening policies:

"The main drive towards introducing screening appears to come from medical agencies, not from lay people or through democratic debate; there has been no parliamentary discussion or law on screening in Britain. Policies are formulated by advisory committees, which issue guidance. Medical reports emphasize the prevention of suffering and promotion of parental choice by offering termination of affected pregnancies. Public health reports stress economic aims: to reduce the 'life time costs of care' for people with Down's syndrome; to avoid costly litigation for 'wrongful birth' of undiagnosed babies; and to develop comparatively cheaper, although also profitable, screening technologies."

Just because we will soon be able to identify thousands of risk factors for this, that and the other, does not mean it is necessarily desirable::

"Once fetal DNA can be non-invasively obtained, screening practices will be able to generate a massive amount of information of uncertain importance. These data might cause more harm than good..."

An excellent paper discussing many of the issues relating to disability and genetic advances appeared in a recent issue of Nature Reviews Genetics. In the introduction, Jackie Scully highlights the pressing need to grapple with these issues:

"The need to think through the ethically complex relationship between disability and genetic knowledge has recently become more pressing, as theoretical advances in genetics begin to be parlayed into clinical interventions. Put bluntly, genetics now enables 'something to be done' about disability, and this demands clarity about the kind of 'something' that is socially and ethically desirable. Through the genetic data that are becoming available... an increasing number of gene loci can now be prenatally and postnatally detected. Such knowledge could, in principle, be used in large-scale population screening of foetuses or newborns, as well as for individualized testing, and in the more distant future to develop various forms of gene-based therapy. As the possible interventions supported by genetic science become more diverse and sophisticated, the ethical debate is shifting from its initial focus on the legitimacy of prenatal selection against genetic disabilities to grapple with the much more fundamental questions of the way in which society conceptualizes phenotypic variation and impairment."

Scully goes on to describe some of the concerns expressed about genetic diagnosis (my emphasis):

"At one extreme of the disability critique of genetic diagnosis lies the argument that these techniques infringe the future person's right to life. For some proponents, this argument is based on the premise that termination of foetal life is impermissible in all cases (and sometimes this is stretched to cover the non-transfer of embryos in PGD as well), but more commonly, disability critics will claim that selection against genetic impairments is discriminatory. For example, the bioethicist Adrienne Asch has long argued that an impairment is often no more predictive of the happiness or worth of a future life than a characteristic such as gender or social class; if we do not consider it ethically acceptable to exercise prenatal selection in terms of those characteristics, we should not automatically consider it right to do so for genetic anomalies that happen to be detectable before birth."

And Scully rightly points out that the ethical issues do not stop with prenatal diagnosis:

"The discriminatory critique extends beyond prenatal diagnosis to include other interventions. It is true that currently the practical impact of genetic science on disability is largely prenatal, but this may change if and when various forms of gene therapy or other genetic intervention come on stream. The question will then be how far we are prepared to go in terms of innovative (and expensive) health care to ameliorate or eradicate genetically based impairment. Other anxieties concern the potential for postnatal discrimination in insurance, education or employment on the basis of genetic sequence data."

A brave new world is rapidly approaching...

Comments

[Ruth Palatnik]

Thank you for a well-researched, well-thought-out analysis of the statistics and the implications. It is lkie a breath of fresh air in all the talk going around.

[Julie Oakley]

Very interesting post. And yes I think the NHS is guilty of bulldozing mothers into test after test and encouraging them to abort all babies that have Downs syndrome.