On being at higher risk: A qualitative study of prenatal screening for chromosomal anomalies

Abstract

This paper explores the meaning of higher risk status to women undergoing prenatal maternal screening for chromosomal anomalies. Quotations from lightly structured interviews and transcripts of pre-screening consultations in suburban London are used to illustrate pregnant women's diverse responses to the offer of screening, and to entering, living with and exiting from higher risk status. Some women reject screening in order to avoid the psychosocial and medical risks associated with higher risk status, or because they rule out pregnancy termination. They may question the risk selection implicitly built into the provision of preventative systems for some health problems but not others. Women who screen at higher risk may challenge this designation by questioning the system-specific probability used to separate them from the lower risk population. However, some experience distress even when they appreciate the precautionary basis on which their higher risk designation is based. They may find disengagement from higher risk status difficult after a diagnostic test has ruled out chromosomal anomalies. The findings highlight the complexity of communicating risk information to pregnant women and other screened populations, and emphasise the need to support those living with higher risk status and the benefits of keeping the time lived with this status as short as possible.

Introduction

This paper will explore women's understandings of higher risk health status, drawing upon a study of prenatal screening for chromosomal anomalies. Higher risk status is constituted by screening systems operating at the limits of technological innovation where non-invasive and individually accurate tests are not yet available. At this frontier, non-invasive screening systems which divide populations into categories with lower and higher probabilities of experiencing a health problem may be developed. Those who screen above a defined risk threshold can then be offered accurate but invasive diagnostic tests. Higher risk status, although not the conditions tested for, is entirely constituted by the screening system. Screened individuals may realise that the results will change ‘their’ probability of encountering the selected problem, but cannot know the direction of this change.

Ethical issues concerning abortion and disability (Williams, Alderson & Farsides, 2002) would remain even if an affordable, non-invasive and accurate diagnostic test for chromosomal anomalies were available. Screening offers a second-best, probabilistic view of the unknown chromosomal status of the fetus, derived from empirical associations, usually neither causal or understood, between diverse markers and the screened condition. Since each marker differentiates populations with a greater prevalence of the condition, combining markers will improve prediction providing that the markers are not themselves closely associated. Probabilistic induction from populations to individuals requires heuristic acceptance of the ecological fallacy that aggregate properties of a category appertain to its members (Robinson, 1950; Greenland & Robbins, 1994; Heyman, Henriksen, & Maughan, 1998).

Heuristics offer simple rules concerning search, search termination and decision-making in complex environments (Gigerenzer & Selten, 2001, p. 8). Heuristics mostly work well enough, but will not always provide a sound guide to action, the price paid for simplification. The probability heuristic will assign some individuals to the higher risk group who do not have the health problem in question and vice versa. The proportions of cases located in these two undesirable categories will depend upon the accuracy of the screening indicators used and the cut-off probability employed for differentiating lower from higher risk cases. With infrequent events, even a low percentage of higher risk cases leads to a large proportion of cases being located in this category despite not having the index condition (Gigerenzer, 2003). Table 1 illustrates this problem with respect to first trimester screening for Down's syndrome and other less common chromosomal anomalies (Bindra et al., 2002).

Table 1 shows that the positive predictive value, the proportion of true cases among those classified as being at higher risk as about 12% (75+54/1096) for this system. Although it is more accurate than second trimester screening (e.g. Canini et al., 2002), about 90% of women who screen at higher risk will not be carrying a fetus with chromosomal anomalies. Because maternal age is included in the risk calculation, older women face a much lower positive predictive value than do younger women, but are less likely to screen at lower risk if carrying a fetus with chromosomal anomalies (Spencer, 2001).

Suggestions for improving the predictive power of this screening system, for example by assessing structural heart defects (Fredouille et al., 2002), are frequently put forward. However, increasing the number of markers may reduce the overall reliability of risk estimation as each marker may be affected by measurement error (Seth & Ellis, 1994). Combining first and second trimester screening, as recommended by Wald, Watt, and Hackshaw (1999), would generate more accurate predictions but leave women waiting several weeks longer for their results. Prediction can also be improved by taking into account covariates of the screening markers, for example a previous higher risk classification in the absence of chromosomal anomalies, and a history of maternal smoking (Aitken, Crossley, & Spencer, 2002). Because of the atheoretical, weakly correlational nature of this form of knowledge, new candidate markers and covariates will be continually identified, subjecting established screening systems to constant pressure to change.

The social status of being at higher risk is constituted entirely by screening provision. Its costs are psychosocial and medical. A recent systematic review of the extensive available literature (Green, Hewison, Bekker, Bryant, & Cuckle, 2004) concluded that women, particularly younger women, experience an increase in anxiety after screening at higher risk which is not necessarily assuaged by a negative diagnosis. One possible explanation for the latter finding is that screening at higher risk for chromosomal anomalies makes the risk of other adverse outcomes more salient. Conflicting research conclusions have been drawn as to whether serious psychological reactions are common (Leithner et al., 2004) or uncommon (Goel, Glazier, Summers, & Holzapfel, 1998). The longer-term impact of living temporarily with higher risk status is not known. Detection of higher risk through a scan rather than chemical markers may cause more anxiety (Weinans et al., 2004), perhaps because it is associated with viewing the fetus. Green et al. (2004) hedge their conclusions with methodological cautions, for example that more anxious women may exclude themselves from studies by declining to complete questionnaires. Moreover, they reference average responses. Qualitative data, including our own, discussed below, documents the crucial mediating role of women's interpretations of being at higher risk.

Available accurate but invasive diagnostic tests, chorionic villus sampling (CVS) and amniocentesis, are associated with medical risks, including that of spontaneous abortion (Tabor et al., 1986). The identified increase in risk is small, from 0.7% to 1.7% in the above large, randomised controlled trial of second trimester amniocentesis versus ultrasound screening, but is faced by all women who undergo diagnostic tests. Most will not be carrying a fetus with chromosomal anomalies. The chromosomal screening and testing system operates as an upwards risk escalator, since an initial risk assessment can lead to further interventions which in turn generate new psychosocial and medical risks (Heyman, 2005).

The present paper will consider women's perspectives on becoming candidates for higher risk status, living with this status, and exiting from it when chromosomal anomalies are ruled out.