THE BCOHTA PROJECT

Provincial policy-makers asked the BC Office of Health Technology Assessment (BCOHTA) some orthodox questions regarding the effectiveness and costs of TMS genetic screening programs and, as has been alluded to, some novel questions regarding their social, ethical, and legal dimensions. In particular, context-specific information was requested to understand why, despite universal technical capability (at least in developed countries), prenatal genetic testing is neither available in all jurisdictions nor uniformly utilized within any particular jurisdiction (4;7;15;25;31).

The ensuing HTA report summarized several policy options for shaping future funding and suggested a provincial process to guide future prenatal genetic screening programs. The HTA project concluded that, at minimum, policy-makers have a duty (as in all health services) to ensure equitable access to prenatal genetic-screening programs, and consistent quality in provision.

A conclusion of possibly greater significance, however, was that the province needs to establish a socially acceptable mechanism to set limits on which fetal conditions are proper to detect inside, or by purchase outside, the publicly-funded system. In addition, policy-makers need to balance support for women who choose prenatal testing and abortion with corresponding support for women and families who choose to raise children with disabilities.

The structural design of the project was necessarily complex. The HTA authors (whose fields of expertise included genetic counseling, feminist anthropology, medical anthropology, general practice, and health economics, in addition to HTA critical appraisal) helped policy-makers in their understanding of TMS in relation to clinical effectiveness and population health, social issues, individual rights, and professional standards, as well as cost. These findings were related to four TMS funding options considered by the province of BC (Table 1).

From its earliest stages, it was also considered necessary that the project should be conducted with particular attention both to the interests of groups who may be seen in some respects as vulnerable, i.e., women and members of the Down syndrome community, and also to the concerns of TMS providers.

The issues examined were those relevant to large-scale population screening of women considered at low pretest risk of carrying an affected fetus. Not addressed were issues of particular relevance to women identified as being at high pretest risk, owing to individual or family history of affected births. In the interest of clarity, the options are discussed as much as possible in relation to Down syndrome, the most common condition traced through TMS.

In developing the evidence, BCOHTA incorporated both primary research conducted using quantitative and qualitative methods, and systematically gathered secondary research from published and unpublished literature. The methods adopted have been reported in detail elsewhere (3).

THE TMS TEST

TMS is a simple blood test designed to test pregnant women of all ages. It measures three fetal biochemical “markers,” which in minute amounts spill into maternal serum throughout pregnancy. Abnormal levels of TMS markers are used primarily to identify those women at increased risk of carrying a fetus with Down syndrome or neural tube disorders.

Women testing TMS-positive are offered noninvasive and, in some cases, diagnostic ultrasound examination for detection of primarily structural abnormalities, some of which are associated with Down syndrome (9). Initial TMS-positive women who have not had dating ultrasound are asked to have the gestational age confirmed by scan. The TMS risk estimate must be recalculated if the ultrasound date estimate is greater than 10 days different from nonultrasound estimates.

TMS-positive women are offered diagnostic testing with amniocentesis. In contrast to TMS, amniocentesis is an invasive test, whereby amniotic fluid is obtained by inserting a needle (guided ultrasonographically) transabdominally into the amniotic sack. Fluid is aspirated and fetal cells are separated for culture. The fetal cells are analyzed for chromosomal patterns, a process which takes approximately 2–3 weeks.

The virtue of amniocentesis is its ability to discriminate between a true positive and a false positive TMS test for conditions such as Down syndrome. This power is, however, materially mitigated by significant numbers of abortions resulting from the procedure. The best currently available estimate of fetal loss comes from a randomized controlled trial of amniocentesis, which reported a procedure-related risk of 0.9 percent (mid-range estimate 15–75 percentile of 0.6–1.2) of pregnancies (29). This study was conducted in the 1980s and may overestimate loss. Techniques have since improved.

Using TMS before amniocentesis increases the percentage of affected fetuses detected among women undergoing amniocentesis and reduces unaffected fetal loss due to amniocentesis. TMS screening of pregnant women of all ages is proposed as a replacement for more limited amniocentesis testing of women 35 years of age and older (approximately 15–20 percent of women).

SOCIAL DETERMINANTS AND SOCIAL IMPACT: WHO IS AFFECTED?

Health policy-makers need to appreciate that those likely to be affected by the introduction of a prenatal testing program fall into two main groups of women, with their partners and families. The first group includes three distinct subsets of women already pregnant or contemplating pregnancy and, therefore, candidates for pre-natal screening:

• (i) The first set includes those women, who, given the option, will want TMS. This is likely to be the majority. Utilization of TMS by the majority of women has been found in Canadian jurisdictions (4–6;16) and suggested as likely in large population surveys conducted for the Royal Commission on Reproductive Technologies and reported in its summary:

“The vast majority of those surveyed would be prepared either to use PND (prenatal diagnosis) themselves (79 percent) or to allow others that option (81 percent). Approximately 18 percent were opposed to either personal use or wider availability of PND services.

A marked majority of those surveyed also support the availability of the option to terminate a pregnancy after PND, with only 16 percent opposed in all circumstances. The level of support depends on the severity of the disorder. For example, 73 percent of people surveyed strongly supported the availability of abortion if a disorder that is fatal early in life is diagnosed in the fetus, while approximately 60 percent supported the availability of abortion for disorders that make it almost certain that independent living will not be possible” (21).

• (ii) The group further includes women who, given the option, do not want the test. They may decline because (among other reasons) they are either willing to accept the possibility that their baby may be disabled, or unwilling to contemplate the possibility of follow-up tests and ultimately abortion if the fetus is found to be affected.
• (iii) This first group must also include women who would want the test, but who are not given the option. They or their primary-care provider may simply be unaware of the existence or applicability of TMS; or they may be denied access because they live outside the areas of BC where anything beyond the primary laboratory test is all but unavailable. Some of these women may believe that failure to provide the service and, therefore, the option of ending the pregnancy represents a violation of their rights, entitling them to pursue a case for damages in a “wrongful birth” lawsuit.

Of those women who subsequently have a baby with one of the relevant disorders, many will accept the challenges of raising their child, usually with the guidance of the one of the provincial support organizations. For this first group, however, TMS, accepted or declined, has the potential for a direct effect on the outcome of their pregnancy.

By contrast, the second main group is only indirectly affected by the test: it comprises those women and families who are already living with a child with disability. It may seem paradoxical to consider those for whom (except for subsequent children) the test would seem to serve no purpose. It is, however, fundamental to the ethical issues raised here that their voices be heard.

The principal concern of this group is that any testing procedure seeking to identify a disorder such as Down syndrome has a eugenic purpose, namely to cleanse the population of imperfection. This they regard as not merely bad for society, diminishing both its genetic and moral base, but also as perpetuating discriminatory attitudes toward individuals with Down syndrome as being unworthy to participate as equal citizens.

In raising these issues in relation to TMS, the clear expectation of this second group is that the ethical questions must be addressed not just by individuals, but at the policy-making level. They assert, for example, that it is unjust to channel public funds almost exclusively toward TMS and diagnostic services, instead of providing social support for affected individuals and families. The inadequacy of such support, they argue, is a major inducement for women to take the screening option.

HOW ARE POPULATION GROUPS AFFECTED?

Most studies of women's experiences with maternal serum screening use standardized scales to assess the psychological impact of these tests. Some research focuses on explaining variations in emotional and psychological responses to maternal serum screening in terms of women's lack of proper information about it. A few studies have examined women's experiences of maternal serum screening along multiple dimensions, including attitudes toward disability, feelings of being at risk, and attitudes toward abortion. While such assessments may explore the psychological-emotional dimensions of maternal serum screening in some detail, the entire experience in which women make sense of and negotiate prenatal genetic technology is intricate, and certainly more complex than simply assessing what women know about serum screening or about detectable conditions.

Asking “what women know” addresses whether women understand the medical meanings and rationale for the test. By contrast, asking “how women make serum screening meaningful” reminds us that women make sense of new technologies in ways that reflect existing cultural frameworks, social relationships of family and community, financial circumstances, and individual histories.

Ethnographic studies of women's experiences with prenatal screening and diagnosis include maternal serum alpha-fetoprotein in California (17;18), amniocentesis in New York City (19;20), amniocentesis in Manitoba (30), and ultrasound fetal imaging in Quebec (14). Several studies examine women's experiences, for example, with the detection of Down syndrome in BC (11), with ultrasound-detected anomalies in BC (28), and with amniocentesis in Quebec (12).

These studies specifically address how women's experiences are shaped by the social and cultural configurations of their lives, and how women come to undergo, or to refuse, prenatal diagnosis. Unfortunately, while richly detailed, most of these ethnographies examine prenatal screening in the United States rather than in Canada, and none specifically addresses TMS.

INFORMED CHOICE: INDIVIDUAL AND COLLECTIVE

The principle of “individual informed choice and consent” is well established in medicine, law, and ethics. Yet to facilitate informed choice about TMS by as many as 30,000 individual women a year is a very considerable challenge in itself, a task that would in all probability fall to primary-care givers (almost exclusively general practitioners and obstetricians, whose formal training in genetic knowledge and counseling varies widely).

A second relevant principle, however, is that of “collective informed choice,” which in relation to prenatal screening, means a population-wide support for a genetic testing program that could lead to selective abortion. This notion remains far less developed and without an institutional, professional, or political home. The question to be resolved is: Does society support a universal prenatal genetic screening option, and if so, for which conditions, and with what testing accuracy?

The literature contains several well-described examples of this aspect of “collective informed choice.” The most often cited instance is support for genetic screening for Tay-Sachs disease, a rare condition resulting in early infant death. Although not a test during pregnancy but a test of prospective parents for recessive gene carrier status, it provides an example of how several Jewish communities in large American cities collectively agreed that funding of this genetic testing program was warranted (13).

The Canadian Royal Commission on New Reproductive Technologies provided evidence of broad public-support across Canada for the availability of prenatal testing to detect conditions such as Down syndrome and neural tube disorders: “A substantial majority (approximately three-quarters) say that if the fetus has a severe anomaly, the parents should have the option to terminate the pregnancy” (22).

Knowing that women would seek this service or think that this service ought to be made available is merely the starting point for consideration, however. Another question might consider whether women would rather allocate scare resources to other programs, known to prevent congenital disorders (for example, programs to reduce neural tube disorders through folic acid supplements or to reduce toxic injuries to the fetus by supporting drug and alcohol addiction programs for women)?

At this interface between the broad social issues and resource allocation, significant questions emerge as to where TMS and requests for program support might fit within government. The issues are not only of spending priorities, important though these are, but also of how TMS is to be evaluated and prioritized and who its champions should be.

The public-policy process that has set current prenatal testing limits has been particularly difficult to discern. The present authors examined the more common cutoff levels that have been applied, such as age 35 as the age of amniocentesis eligibility (5). However, very little has been established about other prenatal tests such as routine ultrasound, its distribution, justification, or relationship to genetic counseling and informed choice, although the latter has been studied in some jurisdictions (2).

CHANGING DEFINITIONS OF DISABILITY

A further critical issue that arises in the social context is the expanding power of the technologies to identify conditions that may be collectively assembled under the heading of “disabilities.” Provincial policy-makers face an era of rapid technological improvement in maternal serum screening innovations already proven accurate (32) or alternate prenatal programs with ultrasound as the primary prenatal screening maneuver (10). It is probable that prenatal programs may change appreciably in the near future, with increasing technical capability to detect a wide range of conditions during pregnancy. The challenge for policy-makers is how to set funding limits within and alongside changing definitions of disability.

Although voiced primarily by disability-rights groups, the issues go beyond narrow sectional interest. In fact, the principal argument is one of accelerating importance within what we properly regard as a free society. At base, it asserts that, where the interest of society as a whole are affected, simply because the technology exists to enable a choice does not mean a person should necessarily be obliged, or even entitled, to make it.

In all probability, therefore, questions on the utilization of any given test turn to questions over which conditions should be tested for at all. Asche (1) discusses this in the US context: “Given that more than 50 million people in the U.S. population have disabling traits and that prenatal tests may become increasingly available to detect more of them, we are confronting the fact that tests may soon be available for characteristics that we have until now considered inevitable facts of human life, such as heart disease.”

In Asche's argument, the growth of prenatal testing and selective abortion should be accompanied by a public policy obligation to support those individuals born with disabilities: “[O]ur clinical and policy establishments must communicate that it is as acceptable to live with a disability as it is to live without one and that society will support and appreciate everyone with the inevitable variety of traits…. When our professions can envision such communication and the reality of incorporation and appreciation of people with disabilities, prenatal technology can help people to make decisions without implying that only one decision is right” (1).

The emphasis on heredity and individual disease, however, risks overwhelming equally important and largely overlooked environmental and social factors with potential to affect population composition. For example, a strong public policy framework is in fact already in place to limit prenatal testing for socially unacceptable reasons such as sex selection.

Even so, simply to limit certain types of prenatal testing is unlikely to achieve total control. So, for example, to ban sex selection and selective abortion of female fetuses would not necessarily prevent alternatives such as infanticide, or social conditions discriminatory against female children.

DISABILITY RIGHTS AND THE BC DOWN SYNDROME COMMUNITY

The HTA authors actively solicited the perspectives of mothers of children with Down syndrome, and professionals who work with affected families. The investigators explored personal experiences, asking how members of these groups confront a technology designed selectively to abort a fetus with the very condition that is an integral part of their lives (11).

The two major support groups in the province have, in summary, asserted that reproductive technologies are predominantly a woman's issue; that they can only be considered positive if they support and enhance women's right to control their own bodies and make meaningful choices about when and whether to give birth; and that in general, reproductive technologies do not support and enhance the equality of either nondisabled women or of disabled people.

Significantly however, the Down syndrome and disability rights communities recognize that TMS will inevitably continue, with or without public funding. Therefore, they oppose licensed private funding, believing that this would give them less opportunity to influence educational messages provided to physicians and women.

COST IMPACT

The HTA project provided health policy-makers with costing estimates for various TMS options, and these are summarized in Table 2. These estimates were to some extent based on current data and to a further extent substantially speculative. The first problem facing policy-makers, however, is not the level of accuracy of costing but whether it is material to policy considerations at all.

This question was first asked in the province almost 20 years ago (23;24), and, although largely side-stepped since, it is still central to the matters under review. Simply put, if the utility of TMS is to be measured only by financial savings, then a full program should be introduced without delay. Allowing for even a broad degree of uncertainty in costing estimates, introduction of TMS with its corollary, the termination of affected pregnancies, will undoubtedly secure substantial savings in public expenditure.

Several generally accepted reports have estimated the more quantifiable health, as opposed to social, costs associated with Down syndrome. One of the more widely cited appraisals, derived in part from primary data from children in BC with Down syndrome, estimated that, in 1997 dollars, the excess cost of health care for a person born with Down syndrome is $350,000 (assuming 75 percent inflation since 1981; 24). A similar estimate, after adjusting for inflation, puts the net present value (in 1987) of the excess cost to society of a child with Down syndrome at around $300,000 (8).

These estimates of excess health-care costs for children with Down syndrome, although crude, support the generally accepted conclusion that even for younger women at lowest pretest risk, prenatal screening for Down syndrome is cost-beneficial to society. The maximum estimate for the cost of TMS per detected fetus with Down syndrome in the younger age groups is $100,000, resulting in a minimum cost-savings ratio of 3:1.

Even the much lower lifetime excess health costs associated with a child with neural tube disorders are estimated at $150,000 (assuming 75 percent inflation since 1983; 23). A cost-benefit analysis of prenatal detection of Down syndrome and neural tube defects in older mothers is sufficient alone to make TMS screening a cost-saving exercise.

Given these figures, why was a comprehensive screening policy not adopted 20 years ago? It is, simply, because the value of a life, disabled or otherwise, is not to be calculated in the columns of credit and debit. The ethical questions, that is, the human questions, are not usefully illuminated by bare arithmetic.

In summary, if policy-makers had only to consider whether to meet the financial cost of extending these facilities throughout the province, it would be an easy decision to make. But the ethical dilemma is far less easy to accommodate and should be expected to have fundamental importance in establishing policy.

It is important to note also that the costing issues relate to cost of program, not cost of “caring.” As Sheldon and Simpson argue (26), cost-benefit analyses can help if a health jurisdiction decides to provide prenatal screening for a condition such as Down syndrome, in that such analyses assign a cost figure to evaluate which prenatal program might provide a cost-benefit versus alternatives. However, cost-benefit analyses provide no assistance with the more difficult judgments of which conditions to test for, or what constitutes adequate testing accuracy.