Turn Away Study: Difference between revisions

From Abortion Risks
Jump to navigation Jump to search
No edit summary
Line 51: Line 51:
# With 68.1% of eligible women refusing to participate in the study at baseline, it is improper for the authors to suggest that their findings reflect the general experiences of most women.  There are numerous [[risk factors]] which have been identified as predicting which women will have the most severe post-abortion reactions.  One of these risk factors, for example, is ambivalence about having an abortion or carrying to term.  Another is the expectation that one will have more negative feelings about the abortion.  In a similar post-abortion interview study by [[Soderberg]], the author reported that in interviews with those declining to participate "the reason for non-participation seemed to be a sense of guilt and remorse that they did not wish to discuss. An answer often given was: ' Do do not want to talk about it. I just want to forget.'"
# With 68.1% of eligible women refusing to participate in the study at baseline, it is improper for the authors to suggest that their findings reflect the general experiences of most women.  There are numerous [[risk factors]] which have been identified as predicting which women will have the most severe post-abortion reactions.  One of these risk factors, for example, is ambivalence about having an abortion or carrying to term.  Another is the expectation that one will have more negative feelings about the abortion.  In a similar post-abortion interview study by [[Soderberg]], the author reported that in interviews with those declining to participate "the reason for non-participation seemed to be a sense of guilt and remorse that they did not wish to discuss. An answer often given was: ' Do do not want to talk about it. I just want to forget.'"
#It is very likely that the self-selected 31.9% of women participating at baseline were more highly confident of their decision to abort prior to their abortions and anticipated fewer negative outcomes. This concern about selection bias is highlighted by the study's own finding that "women feeling more relief and happiness at baseline were less likely to be lost [to follow-up]." Clearly, due to the large numbers of women choosing not to be questioned about their experience, and the large drop out of those who did agree, this sample is not representative of the national population of women having abortions.   
#It is very likely that the self-selected 31.9% of women participating at baseline were more highly confident of their decision to abort prior to their abortions and anticipated fewer negative outcomes. This concern about selection bias is highlighted by the study's own finding that "women feeling more relief and happiness at baseline were less likely to be lost [to follow-up]." Clearly, due to the large numbers of women choosing not to be questioned about their experience, and the large drop out of those who did agree, this sample is not representative of the national population of women having abortions.   
# There may have been additional selection bias on the part of the participating abortion clinics. According to the portion of study protocol that was published: "It is up to the clinic staff at each recruitment site to keep track of when to recruit abortion clients to match to the turnaways recruited."  In other words, the clinic staff exercised considerable leeway in deciding when to invite women to participate, and this leeway could have been exercised in ways to exclude women whom they may have anticipated were among the worst candidates for abortion.
# Despite the initial selection bias, 15% of those agreeing to be interviewed subsequently opted out of the baseline interview and another 31% opted out within the three year followup period.  This means that at the three year followup, only 27.0% of the eligible women were interviewed.  This continuing drop out rate suggests even among women who expected little or no negative reactions, the stress of participating in follow up interviews lead to a change of mind.  Previous research shows that [http://www.ncbi.nlm.nih.gov/pubmed/?term=10718164 women with a history of abortion feel more discomfort in answering questions about their reproductive history].
# Despite the initial selection bias, 15% of those agreeing to be interviewed subsequently opted out of the baseline interview and another 31% opted out within the three year followup period.  This means that at the three year followup, only 27.0% of the eligible women were interviewed.  This continuing drop out rate suggests even among women who expected little or no negative reactions, the stress of participating in follow up interviews lead to a change of mind.  Previous research shows that [http://www.ncbi.nlm.nih.gov/pubmed/?term=10718164 women with a history of abortion feel more discomfort in answering questions about their reproductive history].
#Another oddity, the authors report that in the final group analyzed, average age 25, 62% were raising children.  This would appear to be a very high rate that is not typical of national averages for women seeking abortion.
#Another oddity, the authors report that in the final group analyzed, average age 25, 62% were raising children.  This would appear to be a very high rate that is not typical of national averages for women seeking abortion.

Revision as of 16:43, 21 February 2017

The Turn Away Study is an ongoing study of women who had first and second trimester abortions compared to women who were "turned away" from late term abortions because they approached the clinics in their state after the gestational age limit for performing abortions. The Turn Away Study is conducted by the pro-abortion advocacy group Advancing New Standards in Reproductive Health (ANSIRH) which is a project of the Bixby Center for Global Reproductive Health at the University of California, San Francisco.

The Turn Away Study is seriously flawed by the non-representative selection of women used in the study. More detailed criticisms are further down this page. Here are some of the major points:

  • Of women approached to participate 62.5% declined. Another 15% dropped out before the baseline interview one week after their abortions. After the baseline interview, women continued to drop out at each six month followup period, with an additional 20% drop out at one year, 36% dropout by the third year, and 46% dropout by the fifth year. Despite the low retention rate, the authors mislead readers by declaring that 93% participated "in at least one" of the six month followups, implying high retention when in fact less than 17% of eligible participants, and 46% of those who initially agreed, actually participated in year five.
  • While not explicitly stated by the researchers, the numbers they do reveal indicate that only 27.0% of the eligible women were interviewed at the three year follow-up and only 17% participated at the five year mark. Of the 37% who agreed to participate (1132) only 84% participated at week 1, 66% at year one, 53% at year three, and 46% at year five.
  • There are well known risk factors which predict which women are most likely to have negative reactions to abortion, many of which would make women less likely to agree to participate in a follow up interviews . . . even if there was an offer to be paid. For example, from the APA list of risk factors:
  • perceived need for secrecy;
  • feelings of stigma;
  • use of avoidance and denial coping strategies;
  • low perceived ability to cope with the abortion;
  • perceived pressure from others to terminate a pregnancy.
  • The sample is disproportionately filled with women having late abortions. The sample used includes 413 women who had an abortion near the end of the second trimester compared to only 254 women having an abortion in the first trimester.
  • Women who had abortions due to suspected fetal anomalies were excluded. Probably because research shows high rates of psychological disruption after abortion in these types of cases, therefore excluding this segment of women was a way to reduce the effects associated with abortion. This is extremely misleading, of course, since this is a common reason for abortion . . . especially in the second and third trimester.
  • Demographically, the sample used is not representative of women having abortions. The average age at the time of the abortion was 25, of which 62% were raising children.
  • The comparison group, the Turn Away group (n=210), includes 50 women who later terminated at another facility or had a miscarriage. So 24% of this group, to which the researchers are comparing women who abort, actually includes women who experienced pregnancy losses. Yet the researchers barely disclose this fact, giving the false impression that their study is comparing women who had abortions to women who carried to term. In fact, they are comparing a group of women who had abortions to a group of women including those who (a) carried to term, (b) had abortions in a state other than where they first sought one, or (c) miscarried or had a still birth.




Decision Rightness with Regard to Abortion

Decision Rightness and Emotional Responses to Abortion in the United States: A Longitudinal Study Rocca CH, Kimport K, Roberts SC, Gould H, Neuhaus J, Foster DG. PLoS One. 2015 Jul 8;10(7):e0128832. doi: 10.1371/journal.pone.0128832. eCollection 2015.

Abstract

BACKGROUND: Arguments that abortion causes women emotional harm are used to regulate abortion, particularly later procedures, in the United States. However, existing research is inconclusive. We examined women's emotions and reports of whether the abortion decision was the right one for them over the three years after having an induced abortion.
METHODS: We recruited a cohort of women seeking abortions between 2008-2010 at 30 facilities across the United States, selected based on having the latest gestational age limit within 150 miles. Two groups of women (n=667) were followed prospectively for three years: women having first-trimester procedures and women terminating pregnancies within two weeks under facilities' gestational age limits at the same facilities. Participants completed semiannual phone surveys to assess whether they felt that having the abortion was the right decision for them; negative emotions (regret, anger, guilt, sadness) about the abortion; and positive emotions (relief, happiness). Multivariable mixed-effects models were used to examine changes in each outcome over time, to compare the two groups, and to identify associated factors.
RESULTS: The predicted probability of reporting that abortion was the right decision was over 99% at all time points over three years. Women with more planned pregnancies and who had more difficulty deciding to terminate the pregnancy had lower odds of reporting the abortion was the right decision (aOR=0.71 [0.60, 0.85] and 0.46 [0.36, 0.64], respectively). Both negative and positive emotions declined over time, with no differences between women having procedures near gestational age limits versus first-trimester abortions. Higher perceived community abortion stigma and lower social support were associated with more negative emotions (b=0.45 [0.31, 0.58] and b=-0.61 [-0.93, -0.29], respectively).
CONCLUSIONS: Women experienced decreasing emotional intensity over time, and the overwhelming majority of women felt that termination was the right decision for them over three years. Emotional support may be beneficial for women having abortions who report intended pregnancies or difficulty deciding.
Note: the conclusions should be reframed to note that the conclusions only apply to the 27% of eligible women on whom the researchers had data at the three year mark.

News Coverage

Hardly Any Women Regret Having an Abortion, a New Study Finds. Jenkins N. Time. Published July 14, 2015. 95 percent of women who’ve had an abortion say it was the right decision. Ingraham C. Washington Post. Published July 14, 2015.


Comments & Criticisms

Non-Representative Sample

  1. This study's findings and conclusions are overreaching in many regards, beginning with the fact that the sample of women is not representative of the national population of women having abortions due to high rates of self-exclusion plus high drop out rates. To quote from the study: "Overall, 37.5% of eligible women consented to participate, and 85% of those completed baseline interviews (n = 956). Among the Near-Limit and First-Trimester Abortion groups, 92% completed six-month interviews, and 69% were retained at three years; 93% completed at least one follow-up interview." This means 62.5% of women refused to participate in the study, at first request, and another 15% dropped out before or during the baseline interview, yielding a 31.9% participation rate at baseline.
  2. There are well known risk factors which predict which women are most likely to have negative reactions to abortion, many of which would make women less likely to agree to participate in a follow up interviews . . . even if there was an offer to be paid. For example, from the APA list of risk factors: perceived need for secrecy; feelings of stigma; use of avoidance and denial coping strategies; low perceived ability to cope with the abortion; perceived pressure from others to terminate a pregnancy.
  3. With 68.1% of eligible women refusing to participate in the study at baseline, it is improper for the authors to suggest that their findings reflect the general experiences of most women. There are numerous risk factors which have been identified as predicting which women will have the most severe post-abortion reactions. One of these risk factors, for example, is ambivalence about having an abortion or carrying to term. Another is the expectation that one will have more negative feelings about the abortion. In a similar post-abortion interview study by Soderberg, the author reported that in interviews with those declining to participate "the reason for non-participation seemed to be a sense of guilt and remorse that they did not wish to discuss. An answer often given was: ' Do do not want to talk about it. I just want to forget.'"
  4. It is very likely that the self-selected 31.9% of women participating at baseline were more highly confident of their decision to abort prior to their abortions and anticipated fewer negative outcomes. This concern about selection bias is highlighted by the study's own finding that "women feeling more relief and happiness at baseline were less likely to be lost [to follow-up]." Clearly, due to the large numbers of women choosing not to be questioned about their experience, and the large drop out of those who did agree, this sample is not representative of the national population of women having abortions.
  5. There may have been additional selection bias on the part of the participating abortion clinics. According to the portion of study protocol that was published: "It is up to the clinic staff at each recruitment site to keep track of when to recruit abortion clients to match to the turnaways recruited." In other words, the clinic staff exercised considerable leeway in deciding when to invite women to participate, and this leeway could have been exercised in ways to exclude women whom they may have anticipated were among the worst candidates for abortion.
  6. Despite the initial selection bias, 15% of those agreeing to be interviewed subsequently opted out of the baseline interview and another 31% opted out within the three year followup period. This means that at the three year followup, only 27.0% of the eligible women were interviewed. This continuing drop out rate suggests even among women who expected little or no negative reactions, the stress of participating in follow up interviews lead to a change of mind. Previous research shows that women with a history of abortion feel more discomfort in answering questions about their reproductive history.
  7. Another oddity, the authors report that in the final group analyzed, average age 25, 62% were raising children. This would appear to be a very high rate that is not typical of national averages for women seeking abortion.
  8. The study population is also non-representative of the women having abortion in that it included 413 women who had an abortion near the end of the second trimester compared to only 254 women having an abortion in the first trimester. This is totally disproportionate. It again shows that the authors should not be extending conclusions about this non-representative sample to the general population.
  9. The authors report that sample has an elevated number of low socioeconomic backgrounds. That, too, makes the sample non-representative. The offer of $50 per interview may also have created a participation bias.
  10. The comparison group, the Turn Away group (n=210), includes 50 women who later terminated at another facility or had a miscarriage. So 24% of this group, to which the researchers are comparing women who abort, actually includes women who experienced pregnancy losses. Yet the researchers barely disclose this fact, giving the false impression that their study is comparing women who had abortions to women who carried to term.
  11. Women who had abortions due to suspected fetal anomalies were excluded. Probably because research shows high rates of psychological disruption after abortion in these types of cases, therefore excluding this segment of women was a way to reduce the effects associated with abortion. This is extremely misleading, of course, since this is a common reason for abortion . . . especially in the second and third trimester.
Misrepresentation of Study Design

The authors frequently describe their study as a "prospective longitudinal cohort study." Actually, it is only a "case series study" of the remnent of women (27%) who came to a few abortion clinics who were willing to continue to participate in this study. But since they do not have data collected on the women prior to seeking abortion, much less becoming pregnant, they are not truly prospective cohort studies but rather case series, as clarified by Dekkers et al, (2012) "a cohort study, in principle, enables the calculation of an absolute risk or a rate for the outcome, such a calculation is not possible in a case series." and Song & Chung, 2010:

An important distinction lies between cohort studies and case-series. The distinguishing feature between these two types of studies is the presence of a control, or unexposed, group. Contrasting with epidemiological cohort studies, case-series are descriptive studies following one small group of subjects. In essence, they are extensions of case reports. Usually the cases are obtained from the authors' experiences, generally involve a small number of patients, and more importantly, lack a control group.12 There is often confusion in designating studies as “cohort studies” when only one group of subjects is examined. Yet, unless a second comparative group serving as a control is present, these studies are defined as case-series.

While it is true that the authors are attempting to claim that their sample of "women denied abortions" is the "unexposed group," this is clearly not true for three reasons:

(a) all the women were already exposed to a problem pregnancy,
(b) all the women have already had gone through the process of seeking an abortion...which itself may be all or a portion of the traumatic part of some abortion experiences...especially when they are subsequently raising a child whom they recall at one point having planned to abort (which can cause cognitive dissonance), and
(c) your "unexposed group" clearly includes women who actually have had multiple pregnancy experiences, including abortions and miscarriage, either before or after the index pregnancy, or both. (Indeed, I see that you controlled for parity, but not for prior or subsequent pregnancy losses, which is inconsistent.)


This is a very important distinction in that the description of these studies as prospective longitudinal cohort studies gives the false impression that they meet the criteria of such studies which are designed to follow a group of people before they are exposed to the subject of interest, in this case a pregnancy subject to abortion. In fact, this is a self-selected case series, with very high attrition rate, of women after they are already candidates for abortion.

This is another example of the authors efforts to present their findings as generalizable to the entire population of women when in fact there is no evidence, and every indication given the high refusal and drop out rate, that the findings of the remnent of women remaining in this study are applicable only to that remnant.

Dropouts and Refusers are at Higher Risk

The final sample was only 516 women, which is only 17% of the original 3,045 asked to participate in the study. Clearly, women experiencing the most post-abortion distress are more likely to refuse to participate or drop out. Indeed, the expectation of not coping well with an abortion is predictive of greater post-abortion distress and likely a major reason women would not choose to open themselves up to subsequent telephone interviews.

There is research evidence that low participation rates and / or high dropout rates distort the results of studies and lead to incorrect conclusions:


After a workplace disaster in Norway, 246 employees were required to participate in medical evaluations for PTSD (Weisaeth, 1989). At baseline pre-disaster, employees had a record of cooperation with the company medical officer. After the disaster, some were resistant and required repeated contacts; eventually participation reached 100%. The initial resistance was significantly associated with severity of PTSD at 7 months. The authors stated that if the initial refusals had been accepted, “the potential loss to the follow-up would have included 42% of the PTSD cases, and 64% of the severe PTSD cases would have fallen out, resulting in distorted prevalence rates of PTSD” (Weisaeth, 1989, p. 131). Additionally, “The initial resistance in many who later developed PTSD was found to relate to the psychological defenses such as avoidance which is seen both PTSD and acute post-traumatic stress syndrome” (Shuping, 2016, citing Weisaeth).


Additional evidence of selection bias problems are reported by Soderberg

Inappropriate Measures & Study Design Flaws

  1. The focus of this report in on women's persistent satisfaction with their abortion decisions, "decision rightness," as measured by a single question of whether or not the "abortion was right for them." Women were asked to answer this question "yes", "no" or "uncertain." This measure is flaws in several ways:
    1. A better research approach would have been to have this question rated on a numeric scale (1 to 10, for example) in order to better identify any shift in attitudes.
    2. There is no report of tests to verify the validity and reliability of the questions used. In other words, the authors do not report on any efforts made to evaluate whether the question(s) used provide reliable consistency...or are even understood by women in the same way. For example, do all women interpret the question the "Was my abortion right?" in the same way it is being interpreted by researchers? Or might it be interpreted in numerous ways by different people? Was it "right" as in "moral"? Was it right as in "the best choice I could make at the time?" Was it right in that it was the best choice any person could make? Was it right meaning one would make the same choice if one became pregnant again? Was it right in that "It made my life far better"? Or was it right only in the sense that "What's done is done, and I'm moving my life forward doing the best I can so that is my focus so I guess it was right . . . or at least what I have to work with."
    3. A feeling that a person made the right decision due to circumstances at that time is not the same as an assertion that it was the right decision regardless of circumstances or "if I knew what I knew now."
    4. The decisions assessment were conducted through telephone interviews approximately every six months. A well known problem with interview based studies is that many participants will try to please the interviewer by giving the answer they believe is expected of them. Similarly, some interviewers may be more prone to elicit certain types of response. The investigators did not report on any tests made to verify that such influences were not at play.
  2. As a general rule, questions regarding decision satisfaction (even about things such as the purchase of a purse) may produce reaction formation and therefore defensive answers affirming the rightness of a decision even if there are actually unresolved anxieties or other issues. (To voice dissatisfaction may invite anxiety provoking thoughts. Responding the way one is expect to respond, avoids reflection).
  3. Rather than rely on a single question about the "rightness" of the abortion decision, additional questions should have been asked to better gauge the subjects thoughts. For example, in the Soderberg study, including a one year post-abortion interview of 847 women (after a 33% self-exclusion rate), 80% of the women reported they were satisfied with their decision to abort but at the same time 76% also stated that they would never abort again if faced with an unwanted pregnancy. In this case, the second question offers a great deal of additional insight. A woman expressing unwillingness to not have another abortion may be telling us more than her abortion experience than she is when she says that a past decision was "right."
  4. Another difficulty raised by the researcher's methodology is that their interviews apparently did not inquire about any steps women took to resolve negative emotions. It is necessary to know if women who had negative feelings sought any help to deal with those feelings, perhaps with a therapist, a pastor, or family or friends. The increase in the number of women participating in post-abortion programs should, for example, help to reduce the longevity of negative reactions to abortion. But if this is the case, the conclusion of the authors that negative reactions to abortion naturally diminish over time may be wrong if, in fact, the decrease is due to women receiving post-abortion psychological or spiritual counseling. In other words, if the decline in negative reactions is real (and not due to denial, repression, or just a desire to rush through the phone interview to collect the $50 gift card) it is important to understand the reason for this. Is it due to support given to those having negative feelings, or is it "natural" and permanent?
  5. The authors did not use any validated measures of psychological illness, as has been done in many other studies. Instead the assessment of psychological health is all inferred from a two scales created from six questions in which rated six emotions associated with their abortion women rated each emotion on a five point scale from "not at all" to "extremely." The six emotions were: relief, happiness, regret, guilt, sadness and anger. From these six self-assessments, reported by telephone to an interviewer, the scores for the four negative emotions were combined for a single scale and the two positive emotion scores were combined for a positive emotion scale. These scales were not tested for validity or reliability. Nor were they tested as a measure of overall psychological health or, conversely, psychological illness.
  6. The comparison group, the Turn Away group (n=210), includes 50 women who later terminated at another facility or had a miscarriage. So 24% of this group to which the researchers are comparing women who abort actually includes women who experienced pregnancy losses. Yet the researchers barely disclose this fact, giving the false impression that their study is comparing women who had abortions to women who carried to term.

Inconsistency With Prior Research Findings

  1. As mentioned above, a similarly designed followup study by Soderberg study reported that 80% of the women reported they were satisfied with their decision to abort but at the same time 76% also stated that they would never abort again if faced with an unwanted pregnancy. Soderberg also found that even though many women reported satisfaction with their decision they also experienced negative psychological outcomes, with 50-60% of women undergoing induced abortion experienced some measure of emotional distress, classified as severe in 30% of cases.
  2. Notably, the claim of declining regret and declining negative reactions is at odds with Brenda Major's two year longitudinal study, which also had high drop out rates, which found that there was a trend in decline in relief and increase in negative emotions over the two year period among those who did not drop out of her study. (See Major B, et al. Psychological responses of women after first-trimester abortion. Archives of General Psychiatry. 2000: 57(8), 777-84.)
  3. Their claim that there is no evidence of mental health problems after abortion is not consistent with the findings of studies utilizing validated measures of mental health, including a five year longitudinal study nor with the Christchurch Health and Development Study, a lifetime longitudinal study.
  4. Their findings are not consistent with findings of record linkage studies which have shown an elevated risk of psychiatric admissions following abortion, an elevated rates of suicide and elevated rates of sleep disorders.
  5. Their findings are not consistent with the results of meta-analyses comparing a large number of studies, including reviews by Bellini, Fergusson, and Coleman.

Unreported Details

  1. According to an infographic about the study published by the research group, the followup interviews were actually continued every six months for five years, not just three. Why then did this report limit itself to three years rather than cover the full five years covered by the study?
  2. The study population included 413 women who had an abortion near the end of the second trimester and only 254 women having an abortion in the first trimester. Overall, only 31% participated at the baseline interview (35.7% agreed to be interviewed, but 15% of those dropped out before or during the baseline interview). The authors should report the drop out rate for each of the two groups: first trimester and second trimester.


Overreaching Conclusions

  1. While the report and accompanying press release claim that this study proved there is "no evidence of widespread 'post-abortion trauma syndrome,' in fact it did not use any standard scales for assessment of psychological well being.
  2. The bias of the research team is made clear in press releases and a infographic purporting to summarize the study. In these "summaries" the research group conceals the details regarding the high non-participation rate and boldly claims "95% of women who had abortions felt it was the right decision, both immediately and over 3 years," omitting the fact that 62.5% refused to answer the question at the time of their abortion and of those interviewed at the time 31% were out of the study by the third year. Notably, the problem of high non-participation and drop out rates is not mentioned in the abstract, press release, or other summarizing materials published by the authors. To the contrary, they consistently imply that their results apply to the entire population of women having abortions.
  3. The authors make much of the claim that 93% of the participants "completed at least one follow up interview" which the media outlets incorrectly reported as meaning "Only 7% of the participants dropped out of the study during follow-up." It would have been far more accurate to state that of the "Only 37.5% of eligible women agreed to be interviewed, of whom 85% actually did complete the complete baseline interviews (n = 956). Of this group, only 7% refused to do at least one followup interview."
  4. From the observation that the scale created from four negative reactions showed a modest decline in negative reactions over three years, the authors they draw the very broad conclusion that there is no evidence of widespread negative psychological reactions to abortion. As indicated above, this conclusion is contradicted by better designed studies. Moreover, this conclusion ignores the fact that many psychological problems are characterized by denial and repression of negative emotions. There is, in fact, clear evidence from other studies that many women experience symptoms of post-traumatic stress disorder which includes symptoms of denial and avoidance behavior. In a study by Rue, for example, among women reporting intrusive memories or thoughts related to their abortion, only half denied that these thoughts were attributed (caused) by their abortions. In other words, it is not always easy for women to recognize which feelings may be attributable to their abortions. For example, it is only when in post-abortion counseling that many women may attribute increased feelings of anger after their abortions to unresolved feelings over the abortion which they were projecting onto other people and situations. This is all fairly basic psychology. Negative emotions often crop up in other parts of our lives because we have trouble dealing with them at the source. Therefore, women reporting less "anger" relative to their abortion may in fact have more feelings of anger in their lives than before their abortion but are simply attributing it to other issues. This demonstrates the difficulty in trying to judge the post-abortion emotional adjustment of women based on just six oversimplified questions about six basic emotions.
  5. The Turn Away Study hangs its claim to uniqueness on the fact that it utilizes as it group for comparison only women who initially sought to terminate a pregnancy but were denied abortions because they were beyond the gestation age cut off in various states and then carried to term. But this group of women giving birth is a very small and distinct sample, numbering only a few thousand women per year in the United States. Most importantly, this distinction should be clearly applied to all of the discussion and conclusions offered by the Turn Away Study authors, but it is not. Rather than frame their conclusions as applying to the very small women in the US each year who seek an abortion at or after the gestation date limit on legal abortions in various states, they seek to apply their conclusions to all women having abortions and all women carrying unplanned pregnancies to term.
But even the above clarification limiting the findings to women denied abortions would not be inaccurate because the "turn away" group is not made up only of women who subsequently carried to term. Instead, it includes who found an abortion elsewhere and those who miscarried, making up approximately 24% of the total "turn away" group.
So, to be completely accurate, the Turn Away Study's sloppy methodology mixes different experience and outcomes into the "turn away group" in a way that obscures rather than clarifies the differences between women who (a) have late term abortions and (b) those who carry to term or have late term abortions elsewhere.
Notably, if the Turn Away Study abstracts, conclusions, and press releases were actually rewritten to accurately describe the makeup of the "turn away group" the conclusions drawn from these studies would be so narrow as to be almost meaningless. On the other hand, because the authors generally mention those limitations only once in the methods section of their studies, and then in the conclusion section, abstract, and press releases make it appear that their findings apply to the general population of women having abortions and those who carry unintended pregnancies to term, they are clearly overreaching what their data actually shows. They are merely using their weak data as an excuse to make general pronouncements about "safe abortion" without actually having meaningful data to support those broad claims.

Refusal to Share Details and Data

The authors have refused to share or publish the complete questionnaires used to collect data. They have also refused to share details of their analyses or any of their data for reanalysis by others.


Letter from David Reardon to PLOS One

Dear PLOS One Editors,

I am writing to register a formal complaint against the authors of a PLOS ONE article who I believe have made disingenuous representations to PLOS ONE in order to improperly withhold data.

I have previously been a reviewer for another article submitted to another journal by this team of researchers and in that case also they refused to provide additional requested information, including a refusal to be provided with a blank copy of their survey form so I could review the exact wording of their questions.

Specifically, the article is Decision Rightness and Emotional Responses to Abortion in the United States: A Longitudinal Study by Corrine H. Rocca, et al.

When I emailed Dr. Rocca to request access to the data repository for reanalysis, she responded:

To excuse themselves from providing any data the authors state:

The Data Availability Statement for the paper is on the first page of the publication: “The authors are not able to provide any data beyond what is presented in the manuscript due to restrictions that study participants agreed to when they signed the consent form, which was approved by the UCSF IRB. The authors have included sufficient details in the Methods section of the manuscript for others to replicate the analysis in a similar setting, using a similar study population.”

Regarding the first sentence of this claim, while clearly it would be appropriate to guarantee to participants that no identifying information would be released to others, what possible restrictions would participants be required to agree to that would preclude sharing non-identifying data with other researchers?

So I emailed Dr. Rocca the following request: "Would you please provide a blank copy of the consent form that the study participants signed, where I assume the restrictions on data sharing are described?"

She refused to reply.

Therefore, I am specifically requesting that PLOS ONE require Dr. Rocca to provide a copy of the consent form which the participants signed so that the claim that the non-identifying data cannot be made available based on promises made to the participants may be verified.

If Dr. Rocca should refuse to provide documentation supporting her claims, the journal should retract the paper due to her clear effort to evade the data availability requirements of the PLOS journals

I would note that the Turnaway Study data set, on which this PLOS ONE article is based, has been employed in numerous published articles authored by scores of authors. It is unreasonable to expect that the participants were promised that only a specific list of researchers would be allowed to analyze the non-personal data.

Regarding the claim that "The authors have included sufficient details in the Methods section of the manuscript for others to replicate the analysis in a similar setting, using a similar study population," this is another bogus assertion. As mentioned above, the authors have refused to share even the blank survey instruments used to collect the data so specific questions cannot be replicated.

Furthermore, the ANSIRH team collecting the data is closely aligned with abortion advocacy which is the only reason they were provided access to abortion patients at 30 abortion clinics.

Obviously, abortion is a very contentious issue both politically and academically. Clearly, researchers who are critical of the claim that abortion has no mental health effects are not allowed the access to abortion patients which has been granted to ANSIRH. Therefore, it is impossible for large segments of the research community to "replicate the analysis in a similar setting," as Rocca asserts. Indeed, it is my clear impression, based on Rocca's refusal to provide any additional information even to reviewers, is that she and her team are seeking to limit access to the data and their study methodology precisely to prevent any reanalyzes which may undermine their own preferred spin on the data they collected.

More importantly, the PLOS journals requirements for data sharing exist precisely to alleviate the high cost of replicating data collection and to facilitate reanalyzes of existing data sets.

Please investigate the concerns outlined above, beginning with a request for documentation regarding precisely what was promised to the Turnaway Study participants.

Thank you.

Sincerely yours,

David C. Reardon, Ph.D. Elliot Institute

  • PLOS One declined the request to ask Dr. Rocca to provide any evidence that the consent form did indeed bar sharing non-personal data with other researchers.

Critique by Priscilla Coleman

The following is reprinted with permission from WECARE's website where it is titled The Turnaway Study Analyzed by WECARE Director: The Latest Attempt to Reverse Evidence-based, Women-Centered Advances in Abortion Policy

The PLoS ONE study titled “Decision Rightness and Emotional Responses to Abortion in the United States: A Longitudinal Study” is riddled with serious design flaws that render the results meaningless. The problematic issues are described in detail below followed by evidence that the true motivation for publishing the study is likely political. In recent years, credible science has informed policy with 26 states, now requiring information regarding mental health effects be shared with women considering abortion. This study is a poor attempt to provide counter “evidence” and obscure the reality of women’s suffering, reminiscent of the highly flawed research from the 70s and 80s.

Methodological Issues:

1) As reported by the authors, the consent to participate rate is only 37.5%. This is unacceptable, as the missing 62.5% who were approached and declined were likely the women who had the most adverse psychological reactions to their abortions. With sensitive topic research, securing a high initial consent rate is vitally important and in order to approach being representative, a minimum of 70% should be retained.

2) The authors note that the sample was comprised of a high concentration of women from low socioeconomic backgrounds, rendering the sample not representative of US women undergoing abortion today. There is an ethical concern here as a well, since providing $350 to participate is coercive, as it would be difficult for most of the women to turn down the money.

3) The authors fail to reveal the specific consent to participate rates for each group. Because prior research has demonstrated that second trimester abortions are potentially more traumatizing than first trimester procedures, it is likely that a significantly higher percentage of women in the first-trimester group consented to participate; and the percentage of willing to participate, second trimester participants was likely well under 37.5%. If the rates were comparable, why not report this? Failure to report critical information increases suspicion that this “near limit’ group is in no way representative.

4) In the Turnaway Study, women who secured abortions near the gestational limits included women for whom the legal cut off ranged from 10 weeks through the end of the second trimester. There is a wealth of data indicating that women’s reasons for choosing abortion and their emotional responses to the procedure differ significantly at varying points of pregnancy. Women aborting at such widely different points should therefore not be lumped together, particularly when gestational age information is available in the data.

5) No information is provided regarding how the sites were actually chosen. What type of sampling plan was employed? Why were only those identified with the National Abortion Federation used? What cities were included? Which areas of the country were sampled?

6) The majority of the outcome measures are single items, and this is problematic given the many psychometrically sound multiple item instruments available in the literature for the variables examined. Well-trained behavioral science researchers should not attempt to measure complex human emotions in such a superficial manner; and ethically responsible scientists would not extrapolate from such minimalistic assessments to women’s emotional reactions to one of life’s more challenging decisions.

Bias issues:

7) The authors’ uneasiness with recent litigation is stated in the opening paragraph: “Arguments about emotional harms from induced abortion—including decision regret and increasing negative emotions over time—have been leveraged to support abortion regulation in the United States. To uphold a 2007 law banning a later abortions, Justice Kennedy of the Supreme Court stated: “While we find no reliable data to measure the phenomenon, it seems unexceptionable to conclude some women come to regret their choice to abort...” In support of a state-level ban, a researcher testified that abortion “carries greater risk of emotional harm than childbirth.” Arguments about emotional harm have been used to forward parental consent, mandatory ultrasound viewing, and waiting period legislation as well.” This is a rather odd way to open a supposed scientific investigation and the authors’ unapologetic decision to do so reveals their rather transparent political motivation (i.e., to provide counter results no matter what the scientific cost).

8) The authors’ effort to draw sweeping conclusions from this single, seriously compromised study is evident in their remarks regarding the implications of the study: “Results from this study suggest that claims that many women experience abortion decision regret are likely unfounded.” As scientists we never make such sweeping conclusions based on a single study, particularly when there is an abundant literature comprised of hundreds of sophisticated studies wherein the conclusions are quite discrepant. Courts throughout the US have concluded that women should be appraised of the risks before consenting to abortion; it almost seems silly that these researchers hope to shift the tide based on this study alone.

9) Funding was secured from the David and Lucille Packard Foundation among other sources with a political agenda. As described on their website, “Our work in the United States seeks to advance reproductive health and rights for women and young people by improving access to quality comprehensive sexuality education, family planning and safe abortion care.”

Effect of abortion vs. carrying to term on a woman's relationship with the man involved in the pregnancy

Effect of abortion vs. carrying to term on a woman's relationship with the man involved in the pregnancy. Mauldon J, Foster DG, Roberts SC. Perspect Sex Reprod Health. 2015 Mar;47(1):11-8. doi: 10.1363/47e2315. Epub 2014 Sep 8.


CONTEXT:When a woman who seeks an abortion cannot obtain one, having a child may reshape her relationship with the man involved in the pregnancy. No research has compared how relationship trajectories are affected by different outcomes of an unwanted pregnancy.
METHODS:Data from the Turnaway Study, a prospective longitudinal study of women who sought abortion in 2008-2010 at one of 30 U.S. facilities, are used to assess relationships over two years among 862 women who had abortions or were denied them because they had passed the facility's gestational age limit. Mixed-effects models analyze effects of abortion or birth on women's relationships with the men involved.
RESULTS: At conception, most women (80%) were in romantic relationships with the men involved. One week after seeking abortion, 61% were; two years later, 37% were. Compared with women who obtained an abortion near the facility's gestational age limit, women who gave birth had greater odds of having ongoing contact with the man (odds ratio at two years, 1.7). The odds of romantic involvement at two years did not differ by group; however, the decline in romantic involvement was initially slower among those giving birth. Relationship quality did not differ between groups.
CONCLUSIONS: Giving birth temporarily prolonged romantic relationships of women in this study; most romantic relationships ended soon, whether or not the woman had an abortion. However, giving birth increased the odds of nonromantic contact between women and the men involved throughout the ensuing two years.

Claimed Examination of Physical Health Effects

Side Effects, Physical Health Consequences, and Mortality Associated with Abortion and Birth after an Unwanted Pregnancy Gerdts C, Dobkin L, Foster DG, Schwarz EB. Womens Health Issues. 2016 Jan-Feb;26(1):55-9. doi: 10.1016/j.whi.2015.10.001

INTRODUCTION: The safety of abortion in the United States has been documented extensively. In the context of unwanted pregnancy, however, there are few data comparing the health consequences of having an abortion versus carrying an unwanted pregnancy to term.
METHODS: We examine and compare the self-reported physical health consequences after birth and abortion among participants of the Turnaway Study, which recruited women seeking abortions at 30 clinics across the United States. We also investigate and report maternal mortality among all women enrolled in the study.
RESULTS: In our study sample, women who gave birth reported potentially life-threatening complications, such as eclampsia and postpartum hemorrhage, whereas those having abortions did not. Women who gave birth reported the need to limit physical activity for a period of time three times longer than that reported by women who received abortions. Among all women enrolled in the Turnaway Study, one maternal death was identified-one woman who had been denied an abortion died from a condition that confers a higher risk of death among pregnant women.
CONCLUSION: These results reinforce the existing data on the safety of induced abortion when compared with childbirth, and highlight the risk of serious morbidity and mortality associated with childbirth after unwanted pregnancy.

Note: In addition to the usual problems of this study in regard to its use of a non-representative sample, the measures used are inaccurate. Moreover, the claimed assessment of physical health was based on just two questions (with no examination of actual medical records): 1) “Did you experience any side effects or health problems from your [birth/abortion]?” and 2) “Was there a period after your [birth/abortion] when you were physically unable to do daily activities such as walking, climbing steps or doing errands?”

Abortion related deaths are defined by the Centers for Disease Control (CDC) in the United States as any death due to "1) a direct complication of an abortion, 2) an indirect complication caused by the chain of events initiated by the abortion, or 3) an aggravation of a preexisting condition by the physiologic or psychologic effects of the abortion, regardless of the amount of time between the abortion and the death" (Bartlett, L. a, Berg, C. J., Shulman, H. B., Zane, S. B., Green, C. a, Whitehead, S., & Atrash, H. K. (2004). Risk factors for legal induced abortion-related mortality in the United States. Obstetrics and Gynecology, 103(4), 729–737.) But in the Turnaway Study, the researchers excluded examination of deaths beyond 42 days . . . and, of course, ignored all the record linkage studies showing higher mortality rates after abortion.

Substance Use

Receiving versus being denied an abortion and subsequent drug use.Roberts SC, Rocca CH, Foster DG. Drug Alcohol Depend. 2014 Jan 1;134:63-70. doi: 10.1016/j.drugalcdep.2013.09.013. Epub 2013 Sep 23.

BACKGROUND: Some research finds that women receiving abortions are at increased risk of subsequent drug use and drug use disorders. This literature is rife with methodological problems, particularly inappropriate comparison groups.
METHODS: This study used data from the Turnaway Study, a prospective, longitudinal study of women who sought abortions at 30 sites across the U.S. Participants included women presenting just prior to an abortion facility's gestational age limit who received abortions (Near Limit Abortion Group, n=452), just beyond the gestational limit who were denied abortions (Turnaways, n=231), and who received first trimester abortions (First Trimester Abortion Group, n=273). This study examined the relationship between receiving versus being denied an abortion and subsequent drug use over two years. Trajectories of drug use were compared using multivariate mixed effects regression.
RESULTS: Any drug use, frequency of drug use, and marijuana use did not change over time among women in any group. There were no differential changes over time in any drug use, frequency of drug use, or marijuana use between groups. However, Turnaways who ultimately gave birth increased use of drugs other than marijuana compared to women in the Near Limit Abortion Group (p=.041), who did not increase use.
CONCLUSION: Women receiving abortions did not increase drug use over two years or have higher levels of drug use than women denied abortions. Assertions that abortion leads women to use drugs to cope with the stress of abortion are not supported.

Women’s Mental Health andWell-being 5 Years After Receiving or Being Denied an Abortion

Women’s Mental Health and Well-being 5 Years After Receiving or Being Denied an Abortion: A Prospective, Longitudinal Cohort Study. Biggs MA, Upadhyay UD, McCulloch CE, Foster DG. JAMA psychiatry. December 2016. doi:10.1001/jamapsychiatry.2016.3478.

Objective To assess women's psychological well-being 5 years after receiving or being denied an abortion. Design, Setting, and Participants This study presents data from the Turnaway Study, a prospective longitudinal study with a quasi-experimental design. Women were recruited from January 1, 2008, to December 31, 2010, from 30 abortion facilities in 21 states throughout the United States, interviewed via telephone 1 week after seeking an abortion, and then interviewed semiannually for 5 years, totaling 11 interview waves. Interviews were completed January 31, 2016. We examined the psychological trajectories of women who received abortions just under the facility's gestational limit (near-limit group) and compared them with women who sought but were denied an abortion because they were just beyond the facility gestational limit (turnaway group, which includes the turnaway-birth and turnaway-no-birth groups). We used mixed effects linear and logistic regression analyses to assess whether psychological trajectories differed by study group.
Main Outcomes and Measures We included 6 measures of mental health and well-being: 2 measures of depression and 2 measures of anxiety assessed using the Brief Symptom Inventory, as well as self-esteem, and life satisfaction. Results Of the 956 women (mean [SD] age, 24.9 [5.8] years) in the study, at 1 week after seeking an abortion, compared with the near-limit group, women denied an abortion reported more anxiety symptoms (turnaway-births, 0.57; 95% CI, 0.01 to 1.13; turnaway-no-births, 2.29; 95% CI, 1.39 to 3.18), lower self-esteem (turnaway-births, -0.33; 95% CI, -0.56 to -0.09; turnaway-no-births, -0.40; 95% CI, -0.78 to -0.02), lower life satisfaction (turnaway-births, -0.16; 95% CI, -0.38 to 0.06; turnaway-no-births, -0.41; 95% CI, -0.77 to -0.06), and similar levels of depression (turnaway-births, 0.13; 95% CI, -0.46 to 0.72; turnaway-no-births, 0.44; 95% CI, -0.50 to 1.39).
Conclusions and Relevance In this study, compared with having an abortion, being denied an abortion may be associated with greater risk of initially experiencing adverse psychological outcomes. Psychological well-being improved over time so that both groups of women eventually converged. These findings do not support policies that restrict women's access to abortion on the basis that abortion harms women's mental health.

NOTE: The authors hide the fact that only 11% of the eligible women participated in this study thru the fifth year, making it impossible to generalize any findings from this highly self-selected sample, especially when women having negative reactions would be most likely to drop out. Moreover, the only elevated risk persisted for only a few weeks after "being denied" an abortion, while women are still under stress and trying to sort out their lives. Most importantly, the study actually found NO negative mental health effects after the child was born, either in the first year or over the five years examined. All the general problems discussed at the top of this page also apply. It is also worth noting that at one week after the abortion, those denied abortion had an average depression score of .13 compared to .44 for women who had an abortion . . . in other words, while turnaways had higher anxiety they had less depression than those who had aborted. The difference in depression scores was not statistically significant due to the low power of this study, but it demonstrates the authors' tendency to overgeneralize findings in away that minimizes effects of abortion and magnifies the effects of being denied an abortion.

Media Coverage

Abortion Is Found to Have Little Effect on Women’s Mental Health New York Times Dec 14, 2016 No Evidence Abortion Leads to Long-Term Depression and Anxiety Newsweek 12/14/16 Worse Psychological Outcomes for Women Denied Abortion Psychiatry Advisor Abortion isn’t linked with mental illness, study shows — but being denied one might be Salon More Mental Health Issues Among Women Denied Abortions MedPage Today Women Denied Abortions Report Worse Mental Health Outcomes AJMC Managed Markets Network Women Denied Abortion Endure Mental Health Toll WebMD. Dec. 14, 2016

Abortion Study 2016: Most Women Didn't Struggle With Decision To Terminate Medical Daily Oct 2016 Denying Abortion Access May Harm Women's Mental Health Dec 15, 2016 Abortion Doesn't Negatively Affect Women's Mental Health: Study Time. Dec 14, 2016 Study: Abortion Doesn't Harm Women's Mental Health, but Denying One Does. The Daily Beast. Dec 14, 2016 New Longitudinal Study Confirms That Women Who Get Abortions Do Not Suffer Psychological Harm Slate Dec 14, 2016

Critique by Priscilla Coleman

Other Criticisms

Still Trying to Disprove Post-Abortion Trauma Syndrome

Flawed, Biased Turnaway Study Now Claims 95 Percent of Women Happy After Abortion

Hardly Any Women Regret Having an Abortion -- Only Millions of Us!

Takeaways from the UCSF Abortion "Turnaway" Study (Series from NRL News Today): Part I: Set up for a Spin

Part II: Finding What They Looked For

Part III: Spinning the Consequences of Abortion

IV: Research Team with an Agenda

Part V: How Bias Can Tilt Results