Research Study: Lesser, Y., & Rabinowitz, J. (2001). Elective amniocentesis in low-risk pregnancies: decision making in the era of information and uncertainty. American Journal of Public Health, 91(4), 639 -641.

Summary of the Study: The purpose of the study was to apply the rational choice theory to explain women' s use of amniocentesis. Amniocentesis is a medical procedure that involves the sampling of amniotic fluid in order to screen for potential abnormalities in the fetus. The analytic sample of the present study consisted of 232 women who had low-risk pregnancies. About 41% or 96 of the 232 women had amniocentesis. Among the 96 women who had this procedure, 57 did so per physician' s recommendation and 39 had an elective amniocentesis, while 136 did not have this procedure. The main variables included knowledge about prenatal diagnostics, attitudes, and emotional preferences.

Study Questions

1. Which three groups are compared in this study? What statistical method(s) did they use in the analysis of the data and why?
2. Did the researchers of this study performed a post hoc analysis for the variables they considered in the analysis? If yes, what test did they use? Was it necessary to perform post hoc analysis for overall non-significant differences in the means for certain variables based on the ANOVA analysis?
3. Describe in a few sentences the most important findings.
4. Do the study findings have implications for health care? Provide a rationale for your answer.

Elective Amniocentesis in Low-Risk Pregnancies: Decision Making in the Era of Information and Uncertainty ABSTRACT Yael Lesser, PhD, and Jonathan Rabinowitz, PhD Objectives. Rational choice theory Knowing more about the fetus increases during weeks 28 to 34. About 41% (96 of 232) was applied to explain women's use of women's responsibility regarding their choices. had amniocentesis (57 did so as per a physi- amniocentesis. Variables included knowl- Studies of women's use of prenatal screening cian's recommendation, 39 had elective am- edge about prenatal diagnostics, atti- suggest that effective use involves knowledge niocentesis with no medical indication), and tudes, and emotional preferences and understanding of prenatal diagnostics pro- 136 did not have amniocentesis (3 against the Methods. Using structured instru- cedures, including probability of fetal abnor- physician's recommendation). Women were re- ments at 9 to 14 and at 29 to 34 weeks' mality in high maternal age and the risks of cruited from 5 maternal health clinics funded gestation, we interviewed 232 Israeli some tests. Studies have found that choosing by national health insurance and 5 private ob- women who had low-risk pregnancies. or avoiding preliminary tests such as ultrasound stetric practices. National health insurance pro- Results. Women who had elective and serum a-fetoprotein is related to awareness vides routine follow-up, including several ul- amniocentesis (n=39) were more knowl- and knowledge of prenatal screening possibile trasound screenings, serum a-fetoprotein test, edgeable about prenatal diagnostics, risks ities. Avoidance of prenatal tests has been re- and amniocentesis after age 34. Additional tests of invasive procedures, and probability of lated to high commitment to pregnancy or re- are provided whenever indicated. fetal abnormality in high maternal age; fusal to recognize the possibility of fetal Study variables and instruments are pre- had fewer children; and had less favor- aberration. "Factors associated with the use of sented in Table 1. The Bounded Rationality able attitudes toward parenthood than amniocentesis were higher education and oc- model was represented by variables of prenatal those who had medically indicated am- cupational level, more knowledge about pre- diagnostics information; understanding of use niocentesis (n=57) and those who did natal diagnostics procedures, "social pres- of tests, risk, and probability of abnormality; not have amniocentesis (n= 136). sure, spousal attitudes, " willingness to terminate and appreciation of technology. The Purposive Conclusions. The use and possible pregnancy based on test results, " and higher Action model was represented by commitment overuse of amniocentesis were associ- levels of anxiety and pessimistic thoughts about to pregnancy, anxiety levels, social pressure to ated with having more information about the anticipated child. "Previous studies thus perform tests, the wish for autonomy in decision prenatal diagnostics and definite emo- suggested that the use of prenatal diagnostics is making, attitudes toward parenthood, and tional preferences. (Am J Public Health. related to both rational variables, such as knowl- spousal consensus regarding tests and termi- 2001;91:639-641) edge, and emotional variables, such as social nation of pregnancy based on test results. pressure and levels of anxiety Most previous studies were conducted in single centers with convenience samples of Results women of high maternal age (37 years or older). In the current study, we examined Table 1 compares the means of the 3 study women's choices regarding amniocentesis groups (elective [n=39], medical [n=57], and among younger women with low-risk preg- no amniocentesis [n=136]) with multivariate nancies and applied 2 complementary models analysis of variance. The F test indicated a sig- of rational choice. The first model, Bounded nificant difference on all variables except trait Rationality, suggests that choices are based and state anxiety and autonomy of decisions, on knowledge and understanding of relevant which was almost significant. The elective information. The second model, Purposive Ac- group, compared with the no-amniocentesis tion, " suggests that internalized values and group, was significantly older, had more edu- emotional priorities within a culture influence cation and higher occupational level, had more how women obtain knowledge and assign at- prenatal diagnostics information, had greater tributes of cost or benefit to the perceived out- technology appreciation, was autonomous, and come of their choices. On the basis of the lit- had more social pressure and a higher degree erature and in-depth interviews conducted with of consensus with spouse, but the number of 12 pregnant women and 3 prominent obstetric pregnancies and children and value placed on cians in teaching hospitals, we hypothesized parenthood were significantly lower. The elec- that women's choices regarding the use of am- tive group, compared with the medical am- niocentesis would be associated with variables niocentesis group, was younger, had fewer chil- from both theoretic models. dren and pregnancies, had less favorable Methods The authors are with the School of Social Work, Bar Ilan University, Ramat Gan, Israel. All 272 recruited women (aged 18 to 43) Requests for reprints should be sent to Yael Les- ser, PhD, School of Social Work, Bar Ilan Univer- were interviewed during weeks 9 to 14 of ges- sity, Ramat Gan, Israel (e-mail: lessey(@ inter.net.il). ation, and 232 completed a second interview This brief was accepted July 21, 2000.TABLE 1-Comparison of Means Between 3 Groups: Elective Amniocentesis, Medically Indicated Amniocentesis, and No Amniocentesis ANOVA Comparison of Means Scheffe Test of (1) Elective 2) Medical (3) No Amniocen- Group Mean Differences n=39) (n=57) tosis (n= 136) Signif- (1) & (2) (2) & (3) (1) & (3) Variable (Range) Mean SD Mean SD Mean SD F Test icance Significance Background variables Age, y (18-43) 30.2 3.1 33.5 4.6 27.8 4.2 34.4 .00 .001 .000 010 Education, y (9-21) 15.9 1.7 14.5 3.2 13.2 2.3 19.5 Do0 .007 094 000 Occupation (10-100) 30.8 16.0 69.5 24.9 62.7 25.4 8.9 .000 .089 -285 001 No. of children (0-9) 0.4 0.6 1.4 1.1 1.0 1.2 8.8 .000 .000 027 047 No. of pregnancies (1-11) 1.5 0.6 3.1 18 2.4 1.7 11.5 .000 .000 003 023 Bounded Rationality model Prenatal diagnostics 26.9 5.9 23.2 7.0 20.3 7.2 14.3 -000 038 018 information (1-38) Use, risk, and probability (0-8) 4.9 1.7 4.4 1.6 3.8 1.7 7.2 .001 361 .111 1003 Technology appreciation (0-4) 3.2 1.1 9.2 2.5 8.8 -000 989 006 013 Purposive Action model Commitment to pregnancy" (1-3) 2.9 0.1 2.8 0.4 2.8 0.2 3.1 .047 072 .131 687 Trait anxiety"(20-80) 36.1 7.3 37.2 8.8 35.9 7.5 0.5 557 .853 -638 .984 State anxiety' (20-80) 41.1 7.8 41.8 10.2 38.9 8.7 2.4 093 .930 .160 468 Social pressure" (0-4) 1.3 1.1 1.4 1.2 0.8 1.0 6.8 001 .384 .005 029 Autonomy of decision" (0-3) 1.6 0.6 1.5 0.6 1.4 0.6 2.8 .058 608 321 052 Parental attitudes' (27-135) 101.0 8.4 106.0 8.5 108.0 9.3 10.6 .000 020 .437 Spousal consensus (1-11) 1.1 8.2 16 6.8 2.1 22.0 000 .350 .000 .000 Note. ANOVA= analysis of variance. Degrees of freedom for all F tests =2, 230. "Marteau et al. 1988, 10 items, Cronbach a =.83. 'Developed for current study, 7 items, Cronbach a=.75. "Developed for current study, 3 items, Cronbach a=.72. "Developed for current study, 4 items. 'Spielberger et al. 1970, 20 items, Cronbach a=.78. Spielberger et al. 1970, " 20 items, Cronbach a=.92. "Developed for current study, 4 items. 'Developed for current study, 3 items. Shemer Verbin after Hoffman and Hoffman 1973,"27 items, Cronbach o=.82. Developed for current study, 9 items, Cronbach a =.58. attitudes toward parenthood, and had more pre- have amniocentesis. We found similar differ- might have attracted women who felt positive natal diagnostics information. The no-amnio- ences between those who had medically indi- about their pregnancy, whereas ambivalent centesis group, compared with both the elec- cated amniocentesis and those who did not women avoided participating, which would ex- tive and the medical amniocentesis groups, was have amniocentesis. plain the high commitment to pregnancy across significantly younger, had less prenatal diag- The results suggest that the use and pos- groups. The decision to use unnecessary tests nostics information, felt less social pressure, sible overuse of amniocentesis were associated may be an expression of vulnerability that re- and relied less on spousal consensus. with having more information about prenatal sults in constant alertness and the taking of In comparison to the elective group, the no- diagnostics and definite emotional preferences. every possible precaution against all eventual- amniocentesis group had significantly more fa- Although most women who had elective am- ities. Also, the mere existence of various pre- vorable attitudes toward parenthood. Commit- niocentesis reported the wish to rule out Down natal screening tests may make the unknown ment to pregnancy was high in all 3 groups, with syndrome and to reduce uncertainty as the and uncertain intolerable stimuli that can only no significant difference between the groups. main reasons for having the test, the associa- be eliminated by actually doing these tests. About 77% in the elective group (30 of 39) noted ion with lower age, fewer children and preg- Paying for additional tests not covered by in- the wish to rule out Down syndrome and reduce nancies, lower value placed on parenthood, and surance may provide an illusion of "buying uncertainty as reasons for having amniocentesis. higher value placed on consensus with spouse certainty." may suggest that elective amniocentesis is a Results of the current study suggested that decision of the couple, involving considera- women use neither a purely rational nor an emo- Discussion tions of social status and shared values of par- tional model but, rather, cognitive elements and enthood and child rearing. Thus, both rational emotional priorities in choosing amniocentesis. When applying variables from rational variables, such as knowledge and understand- Although increased access to information is an choice theory, we found that women who had ing of medical procedures, and emotional vari- objective in public health, emotional preference elective amniocentesis were more knowl- ables, such as attitudes toward parenthood and in prenatal care should be closely evaluated. edgeable about prenatal diagnostics, risk, and spousal consensus, are related to women's use probability of abnormality; had fewer children; of amniocentesis. and had less favorable attitudes toward par- The major limitations of the study stem Contributors enthood than those who had medically indi- from sampling method. The appeal to women Y. Lesser planned the study, conducted preliminary cated amniocentesis and those who did not with low-risk pregnancies on a voluntary basis interviews, prepared questionnaires, supervised col-TABLE 1-Comparison of Means Between 3 Groups: Elective Amniocentesis, Medically Indicated Amniocentesis, and No Amniocentesis ANOVA Comparison of Means Scheffe Test of (1) Elective (2) Medical (3) No Amniocen- Group Mean Differences In=39 (n=57) tesis (n =136) Signif- (1) & (2) (2) & (3) (1) & (3) Variable (Range) Mean SD Mean SD Mean SD F Test icance Significance Background variables Age, y (18-43) 30.2 3.1 33.5 4.6 27.8 4.2 34.4 .000 .001 Education, y (9-21) 000 010 15.9 1.7 14.5 3.2 13.2 2.3 19.5 .000 007 094 000 Occupation (10-100) 80.8 16.0 69.5 24.9 62.7 25.4 8.9 .000 .089 285 No. of children (0-9) 001 0.4 0.6 1.4 1.1 1.0 1.2 8.8 .000 .000 No. of pregnancies (1-11) 027 047 1.5 0.6 3.1 1.8 2.4 1.7 11.5 .000 .000 .003 023 Bounded Rationality model Prenatal diagnostics 26.9 5.9 23.2 7.0 20.3 7.2 14.3 .000 .038 018 000 information (1-38) Use, risk, and probability (0-8) 4.9 1.7 4.4 1.6 3.8 1.7 7.2 .001 .361 111 003 Technology appreciation" (0-4) 3.2 1.1 3.2 1.0 2.5 1.3 8.8 .000 .989 .006 .013 Purposive Action model Commitment to pregnancy (1-3) 2.9 0.1 2.8 0.4 2.8 0.2 3.1 .047 .072 .131 Trait anxiety (20-80) 687 36.1 7.3 37.2 8.8 35.9 7.5 0.5 .557 853 638 984 State anxiety' (20-80) 41.1 7.8 41.8 10.2 38.9 8.7 2.4 .093 .930 160 468 Social pressures (0-4) 1.3 1.1 1.4 1.2 0.8 1.0 6.8 .001 384 005 029 Autonomy of decision" (0-3) 1.6 0.6 1.5 0.6 1.4 0.6 2.8 .058 608 321 .052 Parental attitudes' (27-135) 101.0 8.4 106.0 8.5 108.0 9.3 10.6 .000 .020 437 000 Spousal consensus' (1-11) 8.8 1.1 8.2 1.6 6.8 2.1 22.0 000 .350 .000 000 Note. ANOVA =analysis of variance. Degrees of freedom for all F tests =2, 230. Marteau et al. 1988,' 10 items, Cronbach a=.83. "Developed for current study, 7 items, Cronbach o= .75. "Developed for current study, 3 items, Cronbach a= .72. "Developed for current study, 4 items. *Spielberger et al. 1970, 20 items, Cronbach a=.78. *Spielberger et al. 1970,"20 items, Cronbach a=.92. "Developed for current study, 4 items. "Developed for current study, 3 items. 'Shemer Verbin after Hoffman and Hoffman 1973," 27 items, Cronbach of =.82. 'Developed for current study, 9 items, Cronbach a= .58