Assignment: Counseling Clients Considering Abortion

Inadequate preabortion counseling and decision conflict as predictors of subsequent relationship difficulties and psychological stress in men and women.

Reference : Coyle, C. T., Coleman, P. K., & Rue, V. M. (2010). Inadequate preabortion counseling and decision conflict as predictors of subsequent relationship difficulties and psychological stress in men and women. Traumatology, 16(1), 16–30. https://doi-org.ezp.waldenulibrary.org/10.1177/1534765609347550

By: Catherine T. Coyle APART Inc., Madison, WI, USA; Priscilla K. Coleman Bowling Green State University, Bowling Green, OH, USA Vincent M. Rue Institute for Pregnancy Loss, Jacksonville, FL, USA

Acknowledgement: Catherine T. Coyle, RN, PhD, is a co-director of the Alliance for Post-Abortion Research & Training, Inc. Her research interests include the psychological effects of abortion on men and the psychology of forgiveness. Priscilla K. Coleman, PhD, is an Associate Professor of Human Development and Family Studies at Bowling Green State University. Her current research focuses on women's responses to induced abortion including mental health (anxiety, depression, suicide ideation), substance abuse, intimate partner relationship issues, and parenting. Vincent M. Rue, PhD, is the director of the Institute for Pregnancy Loss in Jacksonville, FL. For 35 years he has treated women and men who have experienced abortion as traumagenic and is an active litigation consultant. The authors declared no potential conflicts of interests with respect to the authorship and/or publication of this article. The authors received no financial support for the research and/or authorship of this article.

Introduction

Few contemporary social issues have evoked more controversy than elective abortion. The continuing debate over abortion and mental health has focused on the nature and frequency of adverse postabortion psychological sequelae. There is now consensus, however, that a significant percentage of women experience negative psychological reactions following abortion ( Bradshaw & Slade, 2003 ;  Coleman, Reardon, Strahan, & Cougle, 2005 ;  Wilmoth, deAlteriis, & Bussell, 1992 ). This study was designed to identify potentially key factors predictive of postabortion relationship problems and psychological stress in both women and men.

Women and Abortion

Recent studies have corrected methodological weaknesses of earlier studies and have revealed increased mental health risks associated with the experience of abortion. The most thoroughly researched adverse consequences include anxiety, depression, substance abuse, suicidal ideation, and suicide (Broen, Moum, Bodtker, & Ekeberg, 2004;  Coleman et al., 2005 ; Cougle, Reardon, & Coleman, 2003; Cougle, Reardon, Coleman, & Rue, 2005;  Coleman, Reardon, Rue, & Cougle, 2002 ;  Fergusson, Horwood, & Ridder, 2006 ; Gissler, Berg, Bouvier-Colle, & Buekens, 2005; Gissler, Hemminki, & Lonnqvist, 1996;  Pedersen, 2007 ,  2008 ;  Reardon & Cougle, 2002 ;  Reardon, Coleman, & Cougle, 2004 ;  Reardon et al., 2003 ;  Rees & Sabia, 2007 ;  Soderberg, Janzon, & Sojberg, 1998 ;  Thorp, Hartman, & Shadigan, 2003 ).

An estimated 43% of U.S. women will experience at least one anxiety disorder in their lifetime ( Breslau, Schultz, & Peterson, 1995 ). Posttraumatic stress disorder (PTSD) is a relatively common and particularly disabling anxiety disorder that may be caused by one or more profound stressors. Extensive research has documented how traumatic stress can significantly alter the quality of individuals' lives (Kapfhammer, Rothenhausler, Krauseneck, Stoll, & Schelling, 2004;  Marshall et al., 2001 ; Schnurr, Hayes, Lunney, McFall, & Uddo, 2006; Warshaw et al., 1993). In the United States, an estimated 13% of women develop PTSD in their lifetime ( Butterfield, Becker, & Marx, 2002 ). Systematic exploration of the role of elective abortion as a traumatic stressor associated with symptoms of PTSD has grown substantially in recent years ( American Psychological Association, 2008 ;  Bradshaw & Slade, 2003 ). Various clinicians have identified abortion as potentially traumagenic (Bagarozzi, 1993,  1994 ;  Burke & Reardon, 2002 ;  De Puy & Dovitch, 1997 ;  Speckhard, 1987 ;  Speckhard & Rue, 1993 ;  Torre-Bueno, 1996 ). Moreover, recent research has provided empirical evidence of this link between abortion and PTSD symptomatology (Kubany, Hill, & Owens, 2003; Mufel, Speckhard, & Sivuha, 2002;  Rue, Coleman, Rue, & Reardon, 2004 ;  Steinberg & Russo, 2008 ;  Suliman et al., 2007 ).  Rue et al. (2004)  and  Suliman et al. (2007)  reported that 12% to 18% of women met the full diagnostic criteria for PTSD after an abortion. An even greater number of women in these studies experienced subthreshold or partial PTSD symptoms following abortion ( Barnard, 1990 ;  Rue et al., 2004 ). The higher the number of these subthreshold symptoms present, the greater the risk of impairment, comorbidity, and suicidal ideation ( Marshall et al., 2001 ).

Informed consent and preprocedure counseling can benefit the patient's decision making and postprocedure emotional and physical adjustment ( Baker, Beresford, Halvorson-Boyd, & Garrity, 1999 ). The perceived adequacy of preabortion counseling may also play an important role in mitigating or increasing the amount of stress women feel following abortion. Preabortion counseling has been criticized as being too time limited, inadequate to address the ambivalence and the complexity inherent in the abortion decision, lacking in discussion of alternatives to abortion, deficient in assessing coercion or pressure to abort, provided by nonprofessionals who are biased, and not tailored to the needs of the individual patient ( Singer, 2004 ;  Steinberg, 1989 ; Stites, 1982).  The National Abortion Federation (2007)  advises that “there should be an opportunity for discussion of the patient's feelings about the abortion decision” (p. 3). However, there is no current standard of care in abortion clinics requiring individualized and thorough counseling regarding the patient's feelings and decision making. In a cross-cultural study,  Rue et al. (2004)  reported that only 29% of women in the U.S. sample received preabortion counseling, and 84% stated that it was inadequate.

Individual psychological responses to abortion have also been found to be related to the quality of preabortion decision making and, particularly, lack of partner support for the decision ( Bradshaw & Slade, 2003 ;  Coleman et al., 2005 ). Research has consistently identified ambivalence and absence of partner support as predictive of negative abortion outcomes ( Bracken, 1978 ;  Coleman et al., 2005 ;  Major & Cozzarelli, 1992 ;  Major et al., 1990 ;  Osofsky & Osofsky, 1972 ).  Payne, Kravitz, Notman, and Anderson (1976)  found that women electing abortion were significantly more angry and depressed afterward, if they were in conflict with their husband or lover over the abortion.  Rue et al. (2004)  reported that most women were unsure about their decision at the time of the abortion, and only 24% perceived their partners as supportive. Thus, the degree of perceived partner support and perceived quality of preabortion counseling are seemingly central factors in possible adverse psychological outcomes following elective abortion and they are addressed in this investigation.

Men and Abortion

Although few studies have addressed men's psychological responses to elective abortion ( Coyle, 2007 ), there are identifiable, recurring themes within the scientific literature. A number of reports have noted men's need and/or desire for counseling ( Gordon, 1978 ;  Lauzon, Roger-Achim, Achim, & Boyer, 2000 ; Myburgh, Gmeiner, & van Wyk, 2001a;  Rothstein, 1977a ; Shostak & McLouth, 1984). Most men who experience a partner's abortion do not perceive it to be a benign experience ( Blumberg, Golbus, & Hanson, 1975 ;  Gordon & Kilpatrick, 1977 ; Poggenpoel & Myburgh, 2002; Shostak, 1979,  1983 ;  White van-Mourik, Connor, & Ferguson-Smith, 1992 ) and specific emotions identified among men include anger, anxiety, guilt, grief, and powerlessness ( Gordon & Kilpatrick, 1977 ;  Holmes, 2004 ; Mattinson, 1985;  Speckhard & Rue, 1993 ). In studies of men dealing with therapeutic abortion following amniocentesis, 82% ( Blumberg et al., 1975 ), 50% ( Jones et al. 1984 ) and 47% ( White van-Mourik et al. 1992 ) of men have reported depression. Furthermore, clinicians have observed symptoms among postabortion men that are consistent with delayed or complicated grief reactions and PTSD (Mattinson, 1985;  Robson, 2002 ;  Speckhard & Rue, 1993 ). These clinical reports involved small numbers of men and, to date, no quantitative studies have looked at the potential for PTSD among men following a partner's abortion. In light of established comorbidity of PTSD with depression and other forms of anxiety ( Shalev, 2001 ), further investigation is warranted to determine the extent of risk of psychological trauma among men whose partners undergo elective abortion.

Men tend to defer the abortion decision to their partners and suppress their own emotions and desires as they attempt to support their partners ( Gordon & Kilpatrick, 1977 ;  Robson, 2002 ; Shostak & McLouth, 1984), and men who disagree with their partners' abortion decisions may be more susceptible to intense anger (Naziri, 2007;  Reich & Brindis, 2006 ). Even men who agree with the abortion decision may suffer from ambivalence (Kero & Lalos, 2000,  2004 ; Kero, Lalos, Hogberg, & Jacobsson, 1999) and their relationships, both social and sexual, with their partners may be strained or come to an end ( Berger, 1994 ;  Coleman, Rue, Spence & Coyle, 2008 ; Myburgh, Gmeiner, & van Wyk, 2001b; Naziri, 2007;  Rothstein, 1977b ;  White van-Mourik et al., 1992 ).

Although little is known about the long-term effects on men,  M. Buchanan and Robbins (1990)  provided evidence that adolescent pregnancy resolution may have effects that last into adulthood. These authors found that adult men who experienced abortion during adolescence were more psychologically distressed than adult men who became fathers during adolescence.

Although men are involved with conception and abortion, they are not routinely offered abortion counseling. Despite the call for greater inclusion of and attention to males in abortion clinics (Shostak, 2007), little has changed. Most men who accompany women for abortion do not receive counseling and are left alone to wait.

Given that abortion is a highly personal and sensitive issue, an online investigation seems ideally suited to this topic. Participants may remain anonymous thereby increasing their comfort with self-disclosure. The very existence of an online survey concerning the emotional and relational aspects of abortion may serve to normalize respondents' experiences and encourage them to seek help if needed.

Web-Based Research

This investigation represents one of the first online studies pertaining to the topic of abortion and, in this section, established advantages and disadvantages of this contemporary data collection mode are examined. Use of the Internet to engage in data collection is time- and cost-efficient ( Duffy, 2000 ;  Wilson, 2003 ), effective in accessing difficult-to-reach populations ( Mangan & Reips, 2007 ; Yeaworth, 2001), and enhances respondents' comfort with the process and motivation to participate ( Adler & Zarchin, 2002 ;  Gosling, Vazire, Srivastava, & John, 2004 ). A review of Web-based studies published in the American Psychological Association journals between 2003 and 2004 (Skitka & Sargis, 2006) revealed that 21% of those journals had published at least one such study.

Gosling et al. (2004)  compared a large Internet sample with 510 traditional samples and found that Internet samples “are generally more diverse than samples published in a highly selective psychology journal” (p. 99). Similarly, Mathy, Schillace, Coleman, and Berquist (2002) reported their Internet sample as being more representative in terms of education, income, and ethnic diversity than that of a large sample obtained through random digit dialing. Still others have argued that Internet samples are at least as representative as the ubiquitous college-student samples ( Gosling et al., 2004 ;  Smith & Leigh, 1997 ).

Because data collected through Web-based surveys are often obtained from self-selected, convenience samples, generalization must be approached with caution. However, the voluntary nature of such samples offers considerable benefits ( Buchanan & Smith, 1999 ; Reips, 2000) such as superior responses in terms of clarity and completeness ( Petit, 2002 ;  Walsh, Kiesler, Sproull, & Hesse, 1992 ), and responses that are less likely to be contaminated by social desirability ( Richman, Kiesler, Weisb, & Drasgow, 1999 ). Furthermore, research indicates that data collected online appears to be equivalent to that collected via more traditional methods ( Ballard & Prine, 2002 ;  Hewson & Charlton, 2005 ;  Knapp & Kirk, 2003 ; Robie & Brown, 2006) and  Meyerson and Tryon (2003)  concluded that “data collection on the Web is (1) reliable, (2) valid, (3) reasonably representative, (4) cost effective, and (5) efficient” (p. 614).

Potential risks of online survey administration such as inaccurate responses, failure to respond, and the influence of phrasing and ordering of questions are applicable to traditional survey administration methods as well. The risks of multiple survey submissions and nonserious responses ( Buchanan & Smith, 1999 ;  Schmidt 1997 ) may be avoided by using Internet protocol numbers to identify surveys coming from the same respondent ( Birnbaum, 2004 ;  Gosling et al., 2004 ). Furthermore, the anonymity afforded by the Internet facilitates honest disclosure ( Levine, Ancill, & Roberts, 1989 ;  Locke & Gilbert, 1995 ;  Mangan & Reips, 2007 ).

Ethical considerations in Web-based research are the same as those for other research forms. Consent to participate may be defined as and verified by submission of an online survey. The risk of psychological harm in online surveys has been deemed to be no greater than that of offline surveys ( Kraut et al., 2004 ) if initial instructions include a clear statement respecting the participant's freedom to withdraw from the study at any time. For studies involving sensitive subjects, information concerning referrals for counseling or support may be provided.

Objectives and Hypotheses

Based on the literature reviewed here, it appears that preabortion counseling for women may be limited, whereas for men, it is nonexistent. In addition, men and women may be arriving at abortion decisions that are made without adequate communication and candor between them thus resulting in decisions that are less than satisfactory to one or both parties. Consequent to both the crisis of pregnancy resolution and insufficient communication, relationships may be strained ( Rue et al., 2004 ;  Speckhard & Rue, 1993 ) and psychological stress increased (Bagarozzi, 1994;  Coleman & Nelson, 1998 ;  Fergusson et al., 2006 ).

Both inadequate preabortion counseling and the incongruence of partner abortion decision making may therefore predict postabortion relationship difficulties and/or psychological trauma. Given that some studies on women have found factors such as prior mental health ( Major et al., 2000 ), religious beliefs ( Adler et al., 1990 ;  Major, Richards, Cooper, Cozzarelli, & Zubek, 1998 ), opinions or attitudes about abortion ( Soderberg et al., 1998 ; Zolese & Blacker, 1992), number of abortions ( Rue et al., 2004 ), and various sociodemographic characteristics (Zavodny, 2001) are likely to influence the decision to abort and/or postabortion adjustment, these factors were used as control variables in this study. In addition, history of physical or sexual abuse during childhood or adulthood may be a confounding variable in terms of postabortion mental health given the evidence that such abuse may contribute to emotional problems ( Fergusson, Horwood, & Lynskey, 1996 ;  Schilling, Aseltine, & Gore, 2007 ). Therefore, controls were also implemented for various forms of childhood and adulthood victimization.

The primary objective of this study was to investigate the extent to which perceived inadequacy of preabortion counseling and partner incongruence in abortion decision making predicted postabortion relationship problems and psychological stress. The following hypotheses were tested: Hypothesis 1: Men and women who do not perceive preabortion counseling as having been adequate will be at significantly greater risk for abortion-related anger, relationship problems, and sexual problems after controlling for sociodemographic and personal history variables. Hypothesis 2: Men and women who do not perceive preabortion counseling as adequate will report significantly higher abortion-related stress as evidenced by symptoms of intrusion, avoidance, and hyperarousal, and they will be at significantly greater risk of meeting the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV) diagnostic criteria for PTSD after controlling for sociodemographic and personal history variables. Hypothesis 3: Men and women who were not in agreement with their partners regarding the decision to abort will be at significantly greater risk for abortion-related anger, relationship problems, and sexual problems after controlling for sociodemographic and personal history variables. Hypothesis 4: Men and women who were not in agreement with their partners regarding the decision to abort will report significantly higher abortion-related stress as evidenced by symptoms of intrusion, avoidance, and hyperarousal, and they will be at significantly greater risk of meeting the DSM-IV diagnostic criteria for PTSD after controlling for sociodemographic and personal history variables.

Method

Procedure

Surveys were posted at  www.abortionresearch.net  from April, 2005 through August, 2008. The surveys consisted of questions concerning sociodemographics, meaningfulness of religious affiliation, abortion history, reasons for abortion, perceived adequacy of preabortion counseling, agreement in abortion decision making, opinion regarding abortion at time of procedure, relationship status with partner postabortion, mental health history, abuse history, trauma symptoms related to abortion, abortion-related anger, relationship problems, sexual problems, and general stress attributed to abortion. The introduction to the survey clarified that submission of the survey would qualify as consent to participate and that respondents could withdraw from participation at any time. Links were provided for those respondents who desired support or counseling. Participants were recruited through e-mail requests to crisis pregnancy centers across the United States and to a few other organizations that offer postabortion counseling. Potential participants could also find the survey via search engines using phrases such as “men and abortion,” “women and abortion,” or “abortion research.”

Sample

Surveys were completed by 374 women and 198 men. U.S. citizens comprised 81% of the female sample and 78% of the male sample. Citizens from England (6.5% male and 4% female surveys), Canada (4.5% male and 6.4% female surveys), and Australia (2.5% male and 2.7% female surveys) contributed the next largest number of surveys. Respondents also identified the following as country of citizenship: France, Ireland, Norway, Romania, Czechoslovakia, Germany, Sweden, New Zealand, South Africa, Kenya, Mexico, Nicaragua, Brazil, Nepal, and South Korea. The average age of both male and female respondents was 38 years (SD = 12.8 for males and 11.1 for females). Religious affiliation of women was as follows: 81.6% Christian, 0.3% Jewish, 9.5% Other, and 8.6% None. Religious affiliation of males was 82% Christian, 0.5% Jewish, 0.5% Islam, 7.2% Other, and 9.8% None. Females reported an average of 15 years (SD = 11.8) had elapsed since abortion and males reported a mean of 14.7 years (SD = 12) had passed since abortion occurred. Approximately half of the respondents endorsed liberal views prior to abortion with 21% of males and 24% of females agreeing that abortion “should be legal for any reason at any time during pregnancy” and 27% of males and 36% of females agreeing that abortion “should be legal for any reason during the first trimester of pregnancy.” Additional demographic information can be found in  Table 1 .

Measures

Perceived adequacy of preabortion counseling was assessed via a single item question, “Do you think the counseling you received at the abortion clinic was adequate?” to which respondents indicated “yes” or “no.” Agreement regarding abortion decision making was determined by respondents' endorsement of agreement or disagreement with their partners about the decision to abort.

Relationship quality was assessed with single item variables indicating the presence or absence of abortion-related relationship problems, abortion-related anger, and abortion-related sexual problems. These items had dichotomous (yes/no) responses.

Psychological stress was assessed using the PTSD Checklist–Civilian Version (PCL-C). The entire PCL-C was contained within the online survey. The PCL is composed of 17 items that measure the severity of PTSD symptoms. The PCL yields a total score of 17 to 85 and assesses three symptom clusters: arousal, avoidance of, and re-experiencing of the traumatic event. The response format of the PCL is a 5-point Likert-type scale with higher scores indicative of greater traumatic stress. The diagnosis of PTSD was determined using DSM-IV criteria: (a) one or more endorsements of re-experience symptoms; (b) three or more endorsements of avoidance symptoms; and (c) two or more endorsements of hyperarousal symptoms not present prior to the abortion. Reliability and validity of the PCL have been established ( Weathers, Litz, Herman, Huska, & Keane, 1993 ). With the current sample, internal consistency reliability estimates for the full scale and for the arousal, avoidance, and re-experiencing subscales were equal to .89, .77, .78, .80, and .92, .82, .80, .82 using the women's and men's data, respectively.

Results

Table 1  provides frequency data for the independent variables, sociodemographic and personal history control variables, and dependent variables separately for men and women. To test the first and third hypotheses, which predicted that perceptions of inadequate preabortion counseling and disagreement with one's partner regarding the decision to abort would be associated with increased risk for abortion-related anger, relationship, and sexual problems after employing various controls, three sets of logistic regression analyses were conducted separately for males and females in the sample. In the first set, perceptions of counseling inadequacy and partner disagreement operated as the independent variables with abortion-related anger problems functioning as the dependent variable. A similar logistic regression analysis was then conducted incorporating the control variables listed in  Table 1 . In the second set of two logistic regression analyses, the analyses were structured similarly to the first set except relationship problems functioned as the dependent variable. Finally, in the third set of logistic regressions employing a similar structure to the preceding analyses, sexual problems operated as the dependent variable.

The results of these tests are provided in  Table 2  for the female respondents and in  Table 3  for the male respondents. As indicated by the data presented in  Table 2 , prior to inclusion of the control variables, both independent variables (disagreement regarding the abortion decision and perceptions of preabortion counseling as inadequate) were significant predictors of abortion-related anger, relationship, and sexual problems. However, once the controls were entered into the analyses, only the inadequate preabortion counseling variable significantly predicted postabortion-related anger, relationship, and sexual problems in the women sampled. More specifically, the inadequate counseling variable was associated with a 592%, 831%, and 340% increased risk for anger, relationship, and sexual problems, respectively, among the females.

A different pattern of results emerged with the male data. As indicated in  Table 3 , both independent variables were significant predictors of postabortion-related anger, relationship, and sexual problems after statistically controlling for the wide range of sociodemographic and personal situational variables. Inadequate counseling was specifically associated with a 1,797% increased risk of postabortion anger, a 1,421% increased risk of postabortion relationship problems, and a 407% increased risk of postabortion-related sexual problems. In addition, disagreement with one's partner regarding the abortion decision was associated with a 4,248%, 469%, and a 331% increased risk of postabortion-related anger, relationship problems, and sexual problems, respectively.

To test the first part of the second and fourth hypotheses, two sets (one for males and one for females) of analyses of variance were conducted. In each test, the independent variables of partner disagreement on the decision and preabortion counseling inadequacy served as the independent variables with scores on the single item measure of abortion-related stress serving as the dependent variable. Higher scores on the stress measure are indicative of greater stress. One analysis in each set incorporated controls and one did not. Using the female data, without controls employed, the main effect for counseling inadequacy was significant, F(1, 334) = 71.92, p < .0001, as was the main effect for partner disagreement, F(1, 334) = 71.92, p < .0001, and the interaction was significant as well, F(1, 334) = 20.58, p < .0001. Then, with the controls instituted, the results were similar—counseling inadequacy: F(1, 218) = 36.31, p < .0001; partner disagreement: F(1, 218) = 12.23, p < .0001; interaction: F(1, 334) = 5.45, p < .0001. Means were as follows—no agreement, counseling inadequate: 8.80 (SE = .21); no agreement, counseling adequate: 8.26 (SE = .23); agreement, counseling inadequate: 6.78 (SE = .77); agreement, counseling adequate: 3.96 (SE = .56).

Using the male data, without controls employed, only the main effect for partner disagreement was significant, F(1,152) = 10.99, p < .001. Then, with the controls instituted, partner disagreement remained significant, F(1, 95) = 8.24, p = .005, and the interaction effect was likewise significant, F(1, 95) = 4.00, p = .048. Adjusted means were as follows—no agreement, counseling inadequate: 7.81 (SE = .36); no agreement, counseling adequate: 8.28 (SE = 1.77); agreement, counseling inadequate: 6.94 (SE = .38); agreement, counseling adequate: 3.49 (SE = .73).

To test the second part of the second and fourth hypotheses, which predicted that inadequate preabortion counseling and partner disagreement on the abortion decision would be associated with higher risk for experiencing intrusion, avoidance, hyperarousal, and with meeting diagnostic criteria for PTSD after employing controls, four sets of logistic regression analyses were conducted separately for males and females in the sample. The dependent variable in each of the four sets of two analyses was different (intrusion criteria, avoidance criteria, hyperarousal criteria, and general PTSD criteria met) and as in the previous set of logistic regressions performed to test the first and third hypotheses, there were separate tests conducted with and without the controls.  Table 4  provides these results for women, and  Table 5  provides these results for men.

With the female data, both independent variables were associated with increased risk for meeting the DSM-IV criteria for intrusion (202% and 2,383% for the partner disagreement and inadequate counseling variables, respectively) and full PTSD diagnostic criteria after the controls were applied (89% and 283% for the partner disagreement and inadequate counseling variables, respectively.) However, only the inadequate counseling variable was a significant predictor after the controls were included on the avoidance subscale (559% increased risk) and on the hyperarousal subscale (425% increased risk). Using the male data, both independent variables were associated with increased risk of meeting the DSM-IV criteria on the intrusion subscale (925% and 1,737% for the partner disagreement and inadequate counseling variables, respectively). However, only the inadequate counseling variable was associated with increased risk for meeting the DSM-IV criteria for the avoidance subscale (1,005%) after controls were applied. Only partner disagreement over the abortion decision increased risk for experiencing hyperarousal symptoms (384%) and for meeting the full diagnostic criteria for PTSD (210%).

Discussion

The purpose of this study was to explore associations between two independent variables (perceptions of preabortion counseling adequacy and partner abortion decision congruence) and two sets of dependent variables (postabortion relationship problems and psychological stress). Perceptions of inadequate preabortion counseling significantly predicted all the dependent relationship variables for both men and women with utilization of control variables. Although other research has found abortion in itself to be associated with abortion-related anger (Kero, Hogberg, & Lalos, 2004; Naziri, 2007), relationship difficulties ( Barnett, Freudenberg, & Wille, 1992 ;  Lauzon et al., 2000 ;  Rue et al., 2004 ), and sexual dysfunction ( Bradshaw & Slade, 2003 ;  Rue et al., 2004 ), no studies had previously investigated the association between preabortion counseling and postabortion relationship challenges. The inclusion of participants' perceptions of counseling adequacy is therefore an important contribution of the current study.

For women, perceived inadequate counseling also predicted all trauma subscale scores (i.e., intrusion, avoidance, hyperarousal) and predicted meeting diagnostic criteria for PTSD. For men, only intrusion and avoidance scores were predicted by perceptions of inadequate counseling. Similarly,  Peters, Issakidis, Slade, and Andrews (2006)  observed that whereas women were significantly more likely to report arousal symptoms, men were significantly more likely to report avoidance symptoms particularly the symptom of detachment. Both biological ( Bryant & Harvey, 2003 ) and sociocultural (Gavranidou & Rosner, 2003) explanations have been proposed to explain these observed differences between men's and women's endorsement of specific PTSD symptoms. From a biological perspective, males and females may have innate predispositions that differentiate their responses to trauma. Alternatively, culturally prescribed gender roles may influence which trauma symptoms men and women are likely to endorse depending on whether symptoms are perceived as being gender appropriate.

Sex differences in the association between perceived counseling inadequacy and meeting full diagnostic criteria for PTSD may be related to women's direct participation in the abortion procedure, which could predispose them to greater trauma and an increased likelihood of developing PTSD regardless of the quality of counseling. Nonetheless, a large majority of both women and men (85.8% and 86.6%, respectively) in this study indicated that they did not perceive preabortion counseling to be adequate. Because abortion is the legal right of females in the United States and continues to be viewed as an exclusively women's issue, there are no requirements or incentives to offer counseling to male partners. If men receive any counseling at all, it is likely to occur informally if and when they accompany their partners for preabortion clinic visits.

When unplanned pregnancy is experienced as a crisis situation for one or both partners, the individuals tend to use more primitive coping skills and to be psychologically vulnerable as they struggle to solve the problem and regain equilibrium ( Caplan, 1961 ). The emotional strain of the crisis and the lack of effectiveness of one's usual coping mechanisms may result in anxiety and an inability to function ( Caplan, 1961 ). Thus, men and women facing a crisis pregnancy may need considerably more counseling than is currently being offered.

With control variables applied, incongruence of abortion decision significantly predicted trauma symptoms of intrusion and meeting diagnostic criteria for PTSD for both men and women. Contrary to the findings concerning counseling adequacy, disagreement about the abortion decision predicted hyperarousal in men but not in women. Furthermore, decision incongruence predicted abortion-related anger, relationship problems, and sexual difficulties for men only. The inherent inequality of abortion decisions may explain these differential associations.

Numerous studies ( Bracken, Hachamovitch, & Grossman, 1974 ;  Major, Zubek, Cooper, Cozzarelli, & Richards, 1997 ;  Moseley, Follingstad, Harley, & Heckel, 1981 ;  Payne et al., 1976 ) have identified conflict with one's partner and lack of partner support for abortion as predictors of women's postabortion distress. In contrast, very few studies, with the exception of work by Shostak and McLouth (1984) and Naziri (2007), have examined the male's reaction to an abortion that occurs against his wishes. Our findings suggest that disagreement about abortion decisions may be a more robust predictor of traumatic stress among men compared with women.

A notable feature of this study is that it is the first to explore the association between preabortion counseling and postabortion relationship problems and postabortion psychological stress. Employment of numerous control variables, including prior mental health, which has been found to be a determinant of both postabortion adjustment ( Major et al., 2000 ) and PTSD ( Brewin, Andrews, & Valentine, 2000 ), is a major strength of this investigation. Also included as control variables were other known risk factors for the development of PTSD, including history of childhood sexual abuse (Astin, Lawrence, & Foy, 1993), childhood physical abuse (Bremner, Southwick, Johnson, Yehuda, & Charney, 1993;  O'Keefe, 1998 ), physical abuse during adulthood ( Breslau, Davis, Andreski, & Peterson, 1991 ), and sexual abuse during adulthood ( Kessler, Sonnega, Bromet, Hughes, & Nelson, 1995 ).

The use of the Internet is another asset of this study as it enabled acquisition of an international sample and offered anonymity for those who may have been hesitant to participate in research concerning such a sensitive topic. The anonymity afforded by an online survey may augment respondents' perceptions of safety and enhance honest disclosure. Finally, the inclusion of men as well as women is an essential strength of this investigation as the research pertaining to men's responses to abortion is severely inadequate.

A limitation of this study is the fact that the sample was self-selected. Although self-selection may bring some benefits such as a high level of motivation and a genuine desire to contribute to science, self-selected samples limit generalizability of findings. Moreover, the high rate of PTSD among respondents (54.9% of women, 43.4% of men) is indicative of a traumatized sample. By comparison, prevalence of PTSD among women with a history of assault has been reported as 21% ( Breslau et al., 1991 ), among rape survivors, 50% (Foa, 1997), and among Vietnam veterans with high combat exposure, 31% (Kulka et al., 1988).

On the other hand, this highly traumatized sample may represent those who drop out of other studies concerning abortion. In a review of 17 such studies ( Adler, 1976 ), the attrition rate was found to be from 13% to 86% leading to the conclusion that those women who do not participate in follow-up assessments tend to be the most stressed by abortion. As a result, follow-up studies may underestimate negative responses to abortion. Conceivably, anonymous surveys conducted online may be an effective means to reach these traumatized individuals and to gather information from them concerning their abortion experience.

Abortion is one of the most common surgical procedures among women aged 15 to 44 years ( Owings & Kozak, 1996 ) and from 1973 through 2005, more than 45 million elective abortions were performed in the United States ( Guttmacher Institute, 2008 ). If even a small percentage of the men and women involved in abortion are severely traumatized, this may represent a large absolute number of individuals who need psychological support. In addition, the increase in suicidal ideation among those with PTSD ( Sareen, Houlahan, Cox, & Asmundson, 2005 ) and with subthreshold PTSD symptoms ( Marshall et al., 2001 ) raises serious public health concerns if these individuals are not identified and offered help.

In this study, perceptions of preabortion counseling inadequacy were associated with more negative postabortion outcomes in both women and men. Future research should seek to identify the specific elements of counseling that need to be changed or added to achieve better satisfaction with the content and process. Aspects of preabortion counseling to explore further might include the following: (a) the sufficiency of time allotted for counseling, (b) the nature and quality of training of counselors, (c) the inclusion of men in the preabortion counseling process, (d) whether it is better for men and women to be counseled separately or together, and (e) the comprehensiveness and accuracy of information provided.

Findings reported herein provide preliminary evidence that perceptions of inadequate preabortion counseling and abortion decision incongruence may contribute to relationship challenges between partners and to individual psychological stress. Future research to investigate factors that improve the quality and comprehensiveness of preabortion counseling as well as factors that contribute to decision congruence could do much to improve men's and women's postabortion adjustment. In-depth interviews with men and women prior to and after abortion might reveal specific counseling needs that could be incorporated into preabortion counseling protocols. Qualitative studies are needed to delve more deeply into the processes of decision making between women and men facing crisis pregnancies to further our understanding of both intraindividual factors and interpersonal dynamics that may affect the quality and congruence of abortion decisions.