SEARCH

SEARCH BY CITATION

Thumbnail image of

Prenatal screening and diagnosis aim to provide information to pregnant women regarding the risks of their pregnancy. In the case of fetal chromosomal abnormalities, this is offered as an optional component of prenatal care to give women the choice of continuing or terminating an affected pregnancy, but is often seen by the woman as reassurance that her fetus is at low risk or does not have Down syndrome. While maternal age is used as a basis for background risk calculation, the investigation of other markers, such as nuchal translucency (NT) thickness and maternal serum markers in the first or second trimester allow for individual risk to be calculated.

The implementation of screening programs relies on combinations of various methods of risk assessment. Performance of screening programs has been evaluated mainly within an epidemiological or economic context, with pregnant women's decision-making summarized as uptake rates. However, both the optional and individualized components of prenatal screening place women's decision-making at the heart of the analysis of screening programs1, 2.

Understanding women's decision-making and attitudes towards screening and diagnosis is therefore a crucial issue. Most studies have focused on informed choices3–6, with few addressing decision analysis frameworks as a means of providing insight into women's decision-making. It was my aim in this Editorial to explore the contribution of decision analysis in understanding individual decision-making in prenatal screening and invasive testing.

Information and decision-making

  1. Top of page
  2. Information and decision-making
  3. Outcomes and decision-making
  4. Probabilities and decision-making
  5. Sequential decision-making
  6. Conclusion
  7. References

Individual decision analysis frameworks are based on the hypothesis that information supplied about choice options makes individuals capable of expressing a preference. Choice options are characterized by their possible outcomes, with associated likelihoods generally expressed as probabilities. Based on this representation of decision problems, choice elicitation tools such as the ‘decision board’7, 8 have been designed as bedside instruments, used to decide between therapeutic alternatives9–11. More specifically, decision boards are designed as trees that segregate individual choices into three components: options, chances of outcomes and outcomes, with chances of outcomes being illustrated as pie charts and outcomes being described by means of ‘information cards’. Mainly used in oncology, decision boards appear to enhance patients' understanding of possible treatments as well as their awareness of individual choice and their capacity for informed decision-making12, 13.

Prenatal screening for fetal aneuploidies presents important issues concerning informed choice and published studies have mainly explored the degree of understanding resulting from information provided on test options and their results. However, decision boards have not been used in this context. Further studies should assess the practicality of using decision boards in prenatal screening and diagnosis and their impact on informed choice.

Outcomes and decision-making

  1. Top of page
  2. Information and decision-making
  3. Outcomes and decision-making
  4. Probabilities and decision-making
  5. Sequential decision-making
  6. Conclusion
  7. References

Considering that choice options are characterized by outcomes and their associated probabilities, assessing whether outcomes are poorly or highly valued by individuals is of interest for decision analysis. Methods are available for the assessment of so-called ‘utilities of outcomes’, which reflect the psychological values assigned to outcomes and are generally expressed as the desirability of the experience of outcomes.

Most studies involving utilities of pregnancy outcomes aim to elicit how pregnant women balance the risk of giving birth to an affected child with that of procedure-related miscarriage following amniocentesis or chorionic villus sampling14–19. These studies lead to similar findings, with women who favor or choose to carry out invasive prenatal testing assigning higher utilities to procedure-related miscarriage than those who do not. These women also assigned higher utilities to procedure-related miscarriage than they did to the birth of a child with Down syndrome, suggesting that the decision to undergo invasive testing is associated with relatively less concern about procedure-related miscarriage in comparison to giving birth to an affected child18.

These studies provide important findings, as pregnant women's trade-offs between risks can then be compared with those leading to cut-off risks for proposing invasive prenatal testing based on biochemical screening or NT measurement. The choice of cut-off risks is indeed commonly based on equalization of risks of occurrence of outcomes that are supposed to be similarly valued, assuming that women should not be offered prenatal invasive testing when the probability of procedure-related miscarriage exceeds the probability of having an affected child18–22.

The development of more discriminatory methods of screening, together with increasing evidence that utilities that women attach to the birth of a child with major chromosomal defects could differ greatly from those attached to procedure-related miscarriage, may challenge the hypothesis of equalization of risks for the definition of cut-offs. In every day medical practice, these findings also raise questions about possible conflicting attitudes between healthcare providers and pregnant women.

Probabilities and decision-making

  1. Top of page
  2. Information and decision-making
  3. Outcomes and decision-making
  4. Probabilities and decision-making
  5. Sequential decision-making
  6. Conclusion
  7. References

The decision analysis framework is also able to provide a deeper understanding of decision-making by connecting choices to utilities assigned to outcomes. The relationship between decision-making and outcomes relies on the probabilities associated with those outcomes. However, recent developments in decision theories have considered the possibility of subjective transformation of probabilities. Several experimental studies have shown that people tend to overweight low probabilities and underweight high probabilities23–25. For example, the overweighting of low probabilities may contribute to the popularity of lotteries or subscribing to household insurance. Other experimental studies have shown that losses matter more than do gains of the same magnitude, which results in a tendency to overweight outcomes that are perceived as being losses relative to outcomes that are perceived as being gains (loss aversion)24. Within the context of prenatal diagnosis, the overweighting of a low probability of giving birth to an affected child may contribute to decision-making in favor of invasive testing, considering the frequently observed risk aversion in choices between probable (non-invasive) and sure (invasive) testing gains, i.e. the birth of an unaffected child.

Eliciting healthcare providers' and pregnant women's attitudes towards the probabilities at stake in decision-making (probability of giving birth to an affected child or of miscarriage due to invasive testing) could shed light on the physician–patient relationship. To date, only one exploratory study has been conducted14 that focused on consistency of observed choices in prenatal testing with those derived from theoretical models of choice. This study showed that both subjective transformation of probabilities and loss aversion are essential components of women's choices. Considering these encouraging findings, further investigations should be undertaken to provide consistent results regarding the determinants of decision-making.

Sequential decision-making

  1. Top of page
  2. Information and decision-making
  3. Outcomes and decision-making
  4. Probabilities and decision-making
  5. Sequential decision-making
  6. Conclusion
  7. References

Down syndrome screening may involve successive decision-making. Ultrasound screening using NT measurement is usually combined with or followed by biochemical screening in the first or second trimester. Invasive testing can then be undertaken by either chorionic villus sampling or amniocentesis. Finally, there is the option to terminate an affected pregnancy once the results have been disclosed. In successive decision-making, the sequence of decisions itself has to be accounted for, examining each step and its implications. Some studies eliciting women's preferences about biochemical screening in the first or second trimester have accounted for the decisional implications of screening: elicitation of preferences involved the performance of the screening tests, the time when results are disclosed and also the risk of miscarriage due to invasive testing26–28. However, these studies were not designed to elicit the women's preferences over the sequence of possible decision-making from screening to termination of the pregnancy.

This approach to decision-making is conceptualized within the dynamic choice framework and would be particularly relevant to the analysis of decision-making in prenatal screening29. In this model, a decision-maker qualifies as ‘naïve’ or ‘myopic’ when lacking anticipation or awareness about the implications of the current decision-making. When being able to grasp the whole sequence of possible choices, the decision-maker can design a course of action that relies on her preferences about the choice options involved in the sequence and on her anticipated ability to stick to the defined strategy. Within the context of Down syndrome screening, pregnant women who have decided to continue with the pregnancy regardless of the results may adopt different strategies. They may decline screening or invasive testing, because the information provided by testing has no particular value to them or because they want to avoid being faced with the choice of continuing or terminating an affected pregnancy. On the contrary, they may opt for screening and invasive testing to get as much information as possible about their pregnancy because information (even though highly valued) would be unlikely to affect their decision to continue with the pregnancy. While non-myopic decision-making about prenatal screening and diagnosis should be promoted, further investigation within the dynamic choice framework should be undertaken to identify and analyze both pregnant women's and healthcare providers' behaviors and attitudes over the sequence of decision-making possibly involved in Down syndrome screening.

Conclusion

  1. Top of page
  2. Information and decision-making
  3. Outcomes and decision-making
  4. Probabilities and decision-making
  5. Sequential decision-making
  6. Conclusion
  7. References

Societal attitudes towards screening as well as the expected outcomes of screening programs have been studied extensively. However, pregnant women's preferences and decision-making need to be explored further, considering the rapid development of screening methods. Decision analysis is particularly suited to prenatal screening because of the probabilistic nature of the results of screening, and such analysis can help to provide insight into the difficult decisions faced by pregnant women.

References

  1. Top of page
  2. Information and decision-making
  3. Outcomes and decision-making
  4. Probabilities and decision-making
  5. Sequential decision-making
  6. Conclusion
  7. References