| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
RESEARCH AND PRACTICE |
The authors are with L’Institut National de la Santé et de la Recherche Médicale (INSERM), Villejuif, France.
Correspondence: Requests for reprints should be sent to Babak Khoshnood, MD, PhD, INSERM U149, 16 avenue Paul Vaillant-Couturier, Villejuif 94807, France (e-mail: khoshnood{at}vjf.inserm.fr).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
Objectives. We sought to evaluate socioeconomic disparities in serum screening for Down syndrome and assess whether such disparities are more likely to reflect limits in access or information or, rather, informed decisionmaking.
Methods. A nationally representative sample of 12 869 French women completed interviews after giving birth.
Results. We found substantial disparities in the likelihood of (1) women not being offered screening, (2) screening not being offered as a result of late prenatal care, and (3) women not knowing whether or not they had undergone screening. Except in the case of nationality, there was essentially no evidence of differences in refusal of testing.
Conclusions. Rather than representing informed decisionmaking, socioeconomic disparities in screening for Down syndrome are mostly due to limits in access or to information.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
Thus, while obviously important, the objective of making the most effective and efficient use of the techniques available in order to detect the largest possible number of affected fetuses should ultimately be assigned less importance than women’s exercise of informed choice. It therefore becomes necessary, in the evaluation of prenatal testing programs, to examine not only the correlates of prenatal testing but also the reasons that might preclude its use. In particular, with regard to disparities in testing, it is important to assess the extent to which differences might reflect limits in access or information rather than an exercise of informed choice.
In France, prenatal diagnosis of Down syndrome has expanded considerably in recent years from a system essentially based on offering amniocentesis to women 38 years or older to a regulated system of universal access to both ultrasound and maternal serum screening.5,6 Ultrasound screening has increased steadily in use since 1996 and is now routinely offered for measurement of nuchal translucency. A January 1997 governmental decree stipulated that serum screening for Down syndrome be proposed to all pregnant women.7
The French national health insurance program reimburses patients for the costs of antenatal ultrasound and serum screening and also provides amniocentesis coverage for women who, after serum or ultrasound screening, are found to be at high risk (a risk of 1 in 250 or above) of carrying a fetus with Down syndrome. Health care providers are required by law to inform women of the purpose of serum screening and the implications of its results8; they must produce a written statement attesting to the fact that information was provided, and women must give their informed consent for the test to be done.
With increasing use of serum screening, it is important to study possible socioeconomic barriers to women making informed decisions about screening. Previous studies have shown socioeconomic disparities in the use of amniocentesis in several countries, including Australia,9,10 France,11,12 and the United States.13–15 In addition, a recent study suggested that knowledge of Down syndrome and screening might vary in women from different ethnic groups.16 However, in general, little information is available regarding the effects of socioeconomic factors on serum screening.
We assessed the role of several socioeconomic factors in determining women’s use of serum screening and, in particular, the extent to which socioeconomic differences in use of screening might reflect limits in access or information as opposed to informed decisionmaking. Specifically, using data from a nationally representative sample of women who gave birth in France in 1998, we examined socioeconomic differences in the following areas: likelihood of screening not being offered, likelihood of screening not being offered owing to late prenatal care, and likelihood of screening being refused. We also assessed whether women knew their screening status (i.e., whether they knew if they had undergone screening).
There is no consensus regarding the precise meaning of an informed decision about prenatal testing4; at a minimum, however, women need to be offered the test and know that the offer was made if they are to make an informed choice. Hence, any disparities in the likelihood of the test not being offered or of women not knowing whether the test was offered would reflect limits in access or information rather than informed decisionmaking. On the other hand, differences in rates of test refusal could be due to an exercise of informed choice.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
During the interview, both the questions and the response choices were read aloud to the women surveyed. The question regarding serum screening was as follows: "Did you have serum screening for finding out the risk of Down syndrome?" Possible responses were "yes" (screening category 1); "no, not offered" (screening category 2); "no, you refused" (screening category 3); "no, visit too late" (screening category 4); "no (no reason given)" (screening category 5); and "do not know" (screening category 6).
Socioeconomic variables assessed included maternal education, profession, nationality, and marital status (living with or not with a partner); paternal profession; health insurance coverage at the beginning of pregnancy; and whether women reported forgoing a medical test while they were pregnant because of financial reasons. In addition, maternal age and parity were considered potential confounders or effect modifiers. We estimated maternal age at the end of the first trimester using maternal age at childbirth and gestational age.
After exclusion of women with missing screening data (4.2% of the study population), the study population consisted of 12 346 women younger than 38 years and 523 women 38 years or older. In general, missing data were more frequent among women who were members of lower socioeconomic groups. The socioeconomic factors that were most strongly associated with higher odds of unavailable screening data were low maternal education level, nationality other than French, and lack of health insurance coverage. Marital status (living with or not with a partner), maternal profession, and paternal profession had no significant effect on the odds of missing data other than in instances in which the mother or father was unemployed or, in particular, information on paternal profession was missing.
We limited the analysis of the effects of socioeconomic factors on screening to women who were younger than 38 years for 2 reasons. First, on the basis of the maternal age criterion alone, 38 years is the threshold in France for reimbursed amniocentesis. Thus, there are likely to be different reasons for and implications of socioeconomic disparities in screening for women younger than 38 years and women 38 years or older, the majority of whom use amniocentesis with or without serum screening. Second, the number of women 38 years and older was not sufficient for the intended analyses.
We used 
2 tests in conducting univariate analyses and multinomial logit18 models in conducting multivariate analyses. In the multinomial models, the reference category for screening was category 1 (an affirmative response to the question regarding whether screening had been completed), and the reference group for each socioeconomic factor represented the largest group (or one of the largest groups) associated with that particular factor. We report multinomial results in terms of relative risk ratios (RRRs), which corresponded to the exponentiated value of a coefficient for a given socioeconomic factor and category of response (hence, relative risk ratios represented the likelihood, for example, of one socioeconomic group versus another not being offered screening as compared with undergoing screening).
In the context of the multinomial models, the relative probability of a response in a given outcome category (e.g., category 2) relative to the reference outcome response (e.g., category 1) can be expressed as follows:
 | (1) |
where X and ß(2) are vectors of data and coefficients for outcome category 2, respectively. The ratio of this risk for a 1-unit change in variable i is the relative risk ratio, which can be represented as
 | (2) |
where 
is the coefficient for the variable i and outcome category 2. STATA software19 was used in conducting all statistical analyses.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
). Screening status varied significantly between women younger than 38 years and those 38 years or older (2 test; P < .001); in particular, a substantially higher percentage of women 38 years and older were not offered screening.
|
and 3 present data on the associations between socioeconomic factors and screening. Table 2 shows the percentage of women in each screening category according to socioeconomic group, whereas Table 3 shows multinomial relative risk ratios comparing the likelihood of responses in screening categories 2 (screening not offered), 3 (screening refused), 4 (screening not offered owing to late visit), 5 (no test, reason not given), and 6 (unawareness of screening status) with the likelihood of the reference response category (completion of screening). For example, the relative risk ratio of 2.9 (95% confidence interval [CI] = 2.2, 3.9) for unskilled workers in screening category 2 (test not offered) suggests that the likelihood of the test not being offered (as compared with testing having been completed) was about 3 times higher in unskilled workers than in women who were employed in the professional category (Table 3).
|
|
and 3). Proxies of low socioeconomic status were associated with a lower likelihood of screening and higher frequencies of responses from categories 2 (not offered), 4 (visit too late), 5 (no test, reason not given), and, in particular, 6 (unawareness of screening status). In contrast, low socioeconomic status was not associated in general with a higher probability of refusing screening. The only exception was nationality, in which case the relative risk ratios for refusal were significantly higher among women who were not of French nationality (Table 3). In general, the effects of socioeconomic factors on reasons for not undergoing screening varied considerably across different factors. Women’s level of education was most strongly associated with their knowledge of whether or not they had been screened (screening category 6). Education was observed to be associated with screening categories 2 (test not offered), 4 (no test owing to late visit), and 5 (no test, reason not given). In the case of all of these response categories, and in particular category 6, there was a "dose–response" relationship between maternal education level and screening. Differences were substantial in magnitude, especially in the case of women who were not offered screening and women who did not know whether they had undergone screening. On the other hand, there was little evidence that education level predictably affected women’s likelihood of refusing the test.
Maternal profession and paternal profession both affected screening status, and relationships were generally similar in that, among both men and women, the employment categories most likely to be positively associated with completion of serum screening were professional, intermediate, public administration, and commerce. Unemployed women, unskilled workers, and women who were farmers were least likely to have undergone screening. Among women who reported the father’s profession as farming, however, there were high rates of screening, low frequencies of lack of screening (for any reason), and low frequencies of lack of knowledge regarding screening status.
Women of French nationality exhibited the highest rates of screening, and they produced the fewest responses in screening categories 2 (test not offered), 4 (no test owing to late visit), 5 (no test, reason not given), and 6 (unawareness of screening status). Women of other European, North African, and other African nationalities exhibited substantially lower rates of screening and considerably higher proportions of responses from categories 2, 3 (refusal of test), 4, 5, and, particularly, 6 (unawareness of screening status).
Women who lived without a partner, those who did not have health insurance coverage, and those who reported financial difficulty in procuring medical examinations during their pregnancy were less likely to have undergone screening, more likely to have forgone screening for any reason, and more likely to not know whether or not they had undergone screening. In particular, women without health insurance were about 8 to 9 times more likely to produce a response in category 4 (visit too late), 5 (no reason provided for lack of testing), or 6 (unawareness of screening status) than they were to report that they had undergone screening (Table 3).
Table 4 shows results of the multinomial analysis in which the effect of each socioeconomic factor was adjusted for maternal age and parity as well as the study’s other socioeconomic variables. Results showed that the adjusted effects of each of the socioeconomic factors remained significant overall (likelihood ratio tests; P < .001) but were of a substantially lower magnitude than the unadjusted effects; in addition, the adjusted effects of the subgroups were, in many cases, not statistically significant.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
The greatest socioeconomic disparities involved the likelihood of women knowing their screening status. However, overall, and particularly among women from disadvantaged socioeconomic groups, the most important reason for not undergoing screening was that the test was not offered, a situation that occurred in the case of 1 out of every 6 women. On the other hand, except in the case of nationality, socioeconomic groups did not systematically differ in terms of their likelihood of refusing screening. Taken together, these findings suggest that socioeconomic disparities in serum screening are for the most part related to limits in access or information rather than to an exercise of informed choice.
Prenatal testing should include an explanation of the significance and implications of test results. It has been suggested that the information provided before antenatal screening should also include, at a minimum, detailed descriptions of (1) the biological, cognitive, or psychological impairments associated with specific disabilities; (2) the implications of these disabilities for day-to-day functioning; (3) the laws governing education, entitlements to family support services, access to buildings and transportation, and financial assistance to disabled children and their families; and (4) literature produced by disabled people and their families.20
Previous research focusing on screening4,16,21 suggests that a wide gap remains between the reality of testing and the provision of such comprehensive information as just described. Our findings further imply that the information barrier is particularly large among disadvantaged socioeconomic groups, especially women at lower levels of education.
In addition, and notwithstanding the debate regarding the definition of an informed decision,4 it can be argued that our screening categories 1 (completion of testing) and 3 (refusal of testing) would indicate informed decisions only if adequate and appropriate information was provided in a nondirective manner and was understood and used as the basis for women’s decisions. The overwhelming majority (96%) of women who had undergone screening reported that they had received a prescreening explanation.15
However, previous studies suggest that, in the case of a substantial proportion of women, such information is not adequate or is not completely understood.8 Our findings and the findings of others16 indicate that the adverse effects of lack of information and communication on women’s ability to make informed decisions regarding serum screening might be more pronounced among those in lower socioeconomic groups, and particularly those at lower levels of education.
In general, socioeconomic factors can be related to screening through several different mechanisms, including those associated with women themselves, their families, the health care system, and society at large. These mechanisms can affect (1) access to services; (2) information conveyed or understood; (3) religious and cultural preferences, particularly those related to abortion; and (4) motivations or incentives for antenatal screening, which might be in part related to the opportunity cost of time for the mother22 and social views and policies vis-à-vis individuals with congenital anomalies and other disabilities.20
Socioeconomic factors are likely to affect the use of screening through a distinct combination of mechanisms. Indeed, our results suggest that the influence of these factors on screening cannot be adequately summarized by considering the effects of any single factor. Furthermore, within the same socioeconomic categories, maternal factors can have effects that are distinct from those of paternal factors. For example, we found a reduced likelihood of serum screening among women whose profession was farming, but high rates of screening were observed when the father’s profession was listed as farming. More broadly, an understanding of the effects of socioeconomic factors on use of serum screening and on the reasons that might preclude its use can inform policies that might be considered in addressing screening disparities.
Our results suggest that maternal education is the most important predictor of knowledge and communication barriers regarding informed screening decisions. In addition, we found that education was independently related to effective access to screening. Maternal profession and paternal profession had independent effects related to access and information but not preferences. Nationality was related to preferences regarding prenatal testing, but important barriers to access and information, which might in part be related to difficulties with language and interactions with the health system, also exist for women of foreign nationalities.
Finally, even though screening was fully reimbursable, pregnant women who had scarce financial means to pay for medical examinations and those not covered by the national health insurance program were more likely to lack effective access to screening. In this context, the higher likelihood of "refusal" of the test among women with financial difficulties could also indicate lack of effective access rather than preferences of women, since screening was reimbursable but not free.
In France, pregnant women have strong financial incentives to begin prenatal care early and obtain coverage through the national health insurance program.23 While insurance coverage is generally provided, access to coverage might be more difficult among women in disadvantaged socioeconomic groups, particularly women who are not legal residents of France. Overall, 98.5% of the women in our study population were covered by national health insurance. Hence, lack of health insurance did not constitute a major barrier to screening at the population level. Similarly, fewer than 2% of women reported financial difficulty in procuring medical examinations during their pregnancy.17
Overall, fewer than 3% of women did not undergo screening because of late prenatal care. It has been reported in previous studies that 96% of pregnant women in France benefit from early and adequate prenatal care.24,25 In an era of increasing availability of prenatal screening and diagnostic services, it becomes more important to encourage women to begin prenatal care early during their pregnancy. Several European countries provide strong incentives for women to begin prenatal care early and maintain a regular schedule of visits,23 and indeed rates of late or inadequate prenatal care are significantly lower in these European countries than in the United States.24,25
An important caveat that needs to be considered in interpreting our results is that women’s actual screening status might have differed from their reports. Also, women who were less likely to know their status might have misreported it in a systematic way (e.g., responding that the test "was not offered"). Therefore, the socioeconomic differences we found may not accurately reflect disparities in actual use of screening. However, any such "recall bias" might represent, at least in part, lack of informed decisionmaking, because it is plausible that if women had made an informed decision about the test, they would have recalled their status more accurately. In addition, the frequency of screening in our data, approximately 67%, is consistent with the results of a study of screening rates in Paris maternity units conducted by De Vigan and colleagues.26
A conceptually legitimate critique of the increasingly widespread use of prenatal testing has been that it is based on a socially discriminatory view of disability and that widespread testing might reinforce this discriminatory view and practices that stem from it.20 Another form of social discrimination related to prenatal testing, however, might result from lower rates of use of antenatal screening owing to barriers to informed decisionmaking among members of certain socioeconomic groups.
Indeed, our results suggest that socioeconomic disparities in the use of screening for Down syndrome, rather than reflecting preferences and an exercise of informed choice, are for the most part due to limits in access and information. Unless these disparities are addressed, along with the increasing use of serum screening as a strategy for lowering the birth prevalence of Down syndrome, the care of infants with this condition is likely to disproportionately focus on families with fewer resources. Moreover, it is likely that the disparities described here also apply to other congenital anomalies that are subject to antenatal screening.
|
Acknowledgments |
|---|
We wish to thank the Protection Maternelle et Infantile services without whose help the survey could not have been conducted. We also thank the heads of the maternity units who agreed that the survey be conducted in their units. Finally, we wish to gratefully acknowledge the work of the interviewers and the women who participated in the survey.
Human Participant Protection
The national perinatal surveys, in their general form, received an authorization from La Commission Nationale de l’Informatique et des Libertés. Informed consent was obtained from all participants.
|
Footnotes |
|---|
Accepted for publication February 5, 2003.
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
2. Gilbert RE, Augood C, Gupta R, et al. Screening for Down’s syndrome: effects, safety, and cost effectiveness of first and second trimester strategies. BMJ. 2001;323:423–425.
3. Cheffins T, Chan A, Haan EA, et al. The impact of maternal serum screening on the birth prevalence of Down’s syndrome and the use of amniocentesis and chorionic villus sampling in South Australia. Br J Obstet Gynecol. 2000;107:1453–1459.
4. Marteau TM. Towards informed decisions about prenatal testing: a review. Prenat Diagn. 1995;15:1215–1226.[ISI][Medline]
5. Ayme S, Morichon N, Goujard J, Nisand I. Prenatal diagnosis in France. Eur J Hum Genet. 1997;5(suppl 1):26–31.
6. De Vigan C, Vodovar V, Verite V, Dehe S, Goujard J. Current French practices for prenatal diagnosis of trisomy 21: a population-based study in Paris, 1992–97. Prenat Diagn. 1999;19:1113–1118.[ISI][Medline]
7. Journal Officiel de la République Française. Decree of January 1997. 23-1-1997.
8. Seror V, Costet N, Ayme S. Prenatal screening for trisomy 21 with maternal serum markers: information for decision-making in pregnant women [in French]. J Gynecol Obstet Biol Reprod (Paris). 2000;29:492–500.[Medline]
9. Halliday J, Lumley J, Watson L. Comparison of women who do and do not have amniocentesis or chorionic villus sampling. Lancet. 1995;345:704–709.[ISI][Medline]
10. Halliday JL, Warren R, McDonald G, et al. Prenatal diagnosis for women aged 37 years and over: to have or not to have. Prenat Diagn. 2001;21:842–847.[ISI][Medline]
11. Moatti JP, Le Gales C, Julian C, et al. Socio-cultural inequities in access to prenatal diagnosis: the role of insurance coverage and regulatory policies. Prenat Diagn. 1990;10:313–325.[ISI][Medline]
12. Julian-Reynier C, Macquart-Moulin G, Moatti JP, et al. Reasons for women’s non-uptake of amniocentesis. Prenat Diagn. 1994;14:859–864.[ISI][Medline]
13. Kuppermann M, Gates E, Washington AE. Racial-ethnic differences in prenatal diagnostic test use and outcomes: preferences, socioeconomics, or patient knowledge? Obstet Gynecol. 1996;87:675–682.[Abstract]
14. Bishop J, Huether CA, Torfs C, Lorey F, Deddens J. Epidemiologic study of Down syndrome in a racially diverse California population, 1989–1991. Am J Epidemiol. 1997;145:134–147.
15. Khoshnood B, Pryde P, Wall S, et al. Ethnic differences in the impact of advanced maternal age on birth prevalence of Down syndrome. Am J Public Health. 2000;90:1778–1781.
16. Chilaka VN, Konje JC, Stewart CR, Narayan H, Taylor DJ. Knowledge of Down syndrome in pregnant women from different ethnic groups. Prenat Diagn. 2001;21:159–164.[ISI][Medline]
17. Blondel B, Norton J, du Mazaubrun C, Bréart G. Trends in main indicators of perinatal health in France between 1995 and 1998: results from the National Perinatal Surveys [in French]. J Gynecol Obstet Biol Reprod (Paris). 2001;30:552–564.[Medline]
18. Hosmer D, Lemeshow S. Applied Logistic Regression. New York, NY: John Wiley & Sons Inc; 1989.
19. Stata Statistical Software. College Station, Tex: Stata Corp; 2001.
20. Asch A. Prenatal diagnosis and selective abortion: a challenge to practice and policy. Am J Public Health. 1999;89:1649–1657.
21. Seror V, Ayme S. Information disclosure when screening for trisomy 21: a decision-making aid for pregnant women? [in French]. J Gynecol Obstet Biol Reprod (Paris). 2001;30(suppl 1):80–84.[Medline]
22. Becker GS. A Treatise on the Family. Cambridge, Mass: Harvard University Press; 1991.
23. McQuide PA, Delvaux T, Buekens P, and the Study Group on Barriers and Incentives to Prenatal Care in Europe. Prenatal care incentives in Europe. J Public Health Policy. 1998;19:331–349.[ISI][Medline]
24. Buekens P, Kotelchuck M, Blondel B, et al. A comparison of prenatal care use in the United States and Europe. Am J Public Health. 1993;83:31–36.
25. Delvaux T, Buekens P, and the Study Group on Barriers and Incentives to Prenatal Care in Europe. Disparity in prenatal care in Europe. Eur J Obstet Gynecol Reprod Biol. 1999;83:185–190.[ISI][Medline]
26. De Vigan C, Vodovar V, Goujard J, et al. Mothers’ knowledge of screening for trisomy 21 in 1999: a survey in Paris maternity units. Eur J Obstet Gynecol. 2002;104:14–20.[ISI][Medline]
This article has been cited by other articles:
|
|
 |
|
  B. Khoshnood, C. De Vigan, V. Vodovar, G. Breart, F. Goffinet, and B. Blondel Advances in Medical Technology and Creation of Disparities: The Case of Down Syndrome Am J Public Health, December 1, 2006; 96(12): 2139 - 2144. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
