A blog by the magazine's editors and contributors


Contraception and Abortion

File this with the other studies that confirm what most people would have predicted.

Freebirth controlled to greatly lower rates of abortions and births to teenagers, a large study concludes, offering strong evidence for how a bitterly contested Obama administration policy could benefit womens health. The two-year project tracked more than 9,000 women in St. Louis, many of them poor or uninsured, who were given their choice of a range of free contraceptives.

If there's anything surprising about the finding, it is perhaps the magnitude. The study group's abortion rate was less than half that of the control group[women in the St. Louis Metro area] and a third of the national rate.UPDATE: Here's a link to the actual study. I also updated the language in the post because the original wording inaccurately referred to a "control group." The study did not use one.


Commenting Guidelines

Larry -- you mean you just now got around to reading the actual study? Hmm. You still don't seem to comprehend the problem of selection bias. Without a way of accounting for selection bias, it isn't a solid study design. And if you don't understand that point by now, I'm not sure what else I can say except that reading several books on statistics and causal inference would be useful. Everyone else should be aware that the study was the exact equivalent of this: Suppose I want to do a study to find out about weight loss. I sign up thousands of people to go on a 500 calorie a day diet. But I don't have any actual control group with actual identifiable individuals who are being followed over time, and I don't even measure the dieters' starting weights. At the end of two or three years, I weigh the dieters, and then proudly announce that they weigh less than the rest of the city and country. The problem of selection bias would be obvious: the people who would sign up for such a study may be very different from everyone else, and the study design has zero ability to account for this. Indeed, I may have signed up people who were so motivated to lose weight that they would have done something to lose weight regardless. (Just so, this study probably signed up women who were so determined to avoid repeat abortions that they would have found some way of limiting pregnancy regardless, which is why one simply cannot claim that the rest of the city or country is a good idea for a control group.)Even in the field of nutrition, which isn't famed for its methodological rigor, that would be a laughable study. And this is true even though it is "common sense" that a radically calorie-restricted diet leads to losing weight -- it may well do so, but such a study is NOT good evidence of such a diet's effects. You have to learn to distinguish two points in your mind: 1) do I find the study's conclusion plausible? and 2) is the study itself actually solid evidence? It all goes back to logic: an argument might have a true conclusion even though the steps of the argument are invalid.

I don't doubt, by the way, that a program like this would have some effect (the progressive history of forcible sterilization undoubtedly had even more of an effect). But that is common sense, and a study isn't needed to show that there would probably be some effect. What a study would be useful for is answering the following question: exactly how big is the effect and under what conditions? But here's the rub: a study can answer that question only if it's good science, not junk science.

Pardon my double post, but my comment above includes:"Since the Spanish study suggests that intensive education encourages contraceptive use and decreases pregnancies and abortions, however, it seems appropriate to consider this..."The study itself suggests no such thing. The study authors, however, believe it to be the case and, I suspect, would cite other studies to support their belief.

Stuert the "scientist' wants a control group with no pills and he will count their abortions.PHEW

we are conducting one big social experiment with the future of our children.You are forgetting what I think is the most important social experiment, the one with the largest impact on the largest number of children: parental divorce.

In response to Claire:You are so correct concerning divorce and the disastrous effect it has on our children and the family. And once "no fault" divorce came on the scene, many couples no longer felt the need to work out their problems when they could just walk away convincing themselves along the way (when there were children involved) that it was the best thing for the kids! Is it any wonder so many young people are afraid of marriage and say they don't believe in it. They are afraid to trust anyone. Unfortunately, they can thank their parents for that.

WAY off original topic now, but even when parents make an effort to establish the detente that allows them to stay together for the sake of the children, the kids are still "swimming in the soup" of a society that sanctions easy divorce. Kids are quite quick, if they witness any marital discord, to say, "Why don't you just get a divorce?"One reason NOT to get a divorce were the number of "turtle" children I met at my kid's middle school while volunteering--kids who lived out of huge back packs packed full of toiletries and extra clothes as well as school work because they were shuttled to and from one parent to another. These were bright, affectionate kids for the most part, who simply accepted this as a way of life. But it broke my heart to see how much of their own care and organization they had to shoulder for themselves. I have to wonder how much this affects the ability of a child to focus, learn, and develop.

My parents were divorced when i was a baby and then a few years later another divorce, and another divorce again when I was in college. I can only imagine how different I would be now if I'd had two normal happily married parents, but I don't think people should stay together when they don't love each other - don't think that's good for kids either.

Brian Volck --Thank you for a very reasoned consideration of the problem. Only when we are willing to recognize all the evidence -- the possibly pro and the possibly con -- will the solution become possible.

I recommend Brian Volck's detailed and reasoned response to all.

To Dr. Volk,Perhaps you missed the point of the St. Louis study, but the purpose of the study was to see if *bullet-proof* contraceptive measures would reduce re-abortion rates (primary objective), overall abortion rates in the study population (secondary objective), and even reduce overall abortion rates in the St. Louis area, in comparison with those in a comparable metropolitan area (Kansas City) and in non-metropolitan Missouri (secondary objective). The study met all three objectives, in a fashion which can only be described as spectacular. Swedish and Spanish studies you quote did not study populations in which bullet-proof contraceptive measures were utilized and, thus, are not directly relevant to the current study.Both abortion and teen pregnancy rates have fallen dramatically in the USA since 1984. At one point, it was estimated that 25% of this effect may have been owing to abstinence education and 75% to contraception education (*I'll provide a 1998 reference at the end of this comment; I didn't make the effort to determine if there is an updated estimate). These reductions continue to persist up to the present time, despite an environment of continuing sexual permissiveness. The point is that there need not be a conflict between teaching moral behavior to those who are open to such teaching, and also making available effective contraception methods to those who wish to use them and who are actively participating in either "licit" and illicit" sex -- if most of us can agree that contraception is a far lesser evil than abortion, as also concluded by the large majority of the Magesterium and theologians on Pope Paul's Pontifical Commission on Birth Control. I want to repeat my point that the Pope himself acknowledged in his encyclical that lesser evils are preferred to greater evils and that this encyclical was written just before the world-wide emergence of what is essentially abortion on demand.* reference: Larry Weisenthal/Huntington Beach CA

"I dont think people should stay together when they dont love each other dont think thats good for kids either."Maybe it's not optimal, but if you can manage some mutual respect and common decency, I think that's better than shuffling your kid from house to house. Frankly, I think divorced couples should be required to leave the kids in a family home, with the parents doing the moving in and out when it's their custodial time. Why should the kids be messed up?

Hi Jean, Did you happen to catch yesterday's mass Gospel (Mark 10:2-16)? Although we can (in my opinion) argue contraception in good conscience, in the matter of divorce (particularly with children), there's simply no wiggle room. My parents were divorced when I was 12. 53 years older, I still haven't gotten over it and neither has my father, who's shortly turning 99 (my Mom died three years ago). Your story of the "turtle" children really is heartbreaking. I personally think that divorce in the case of a couple with children is infinitely worse than contraception, which is why I think that couples should strongly consider practicing contraception until both woman and man are certain that they'll be able to honor their marriage vows.- Larry W/HB

Crystal --Maybe in some cases divorce is better, but I certainly couldn't condone a woman staying with a man who beat her regularly in front of their children. But maybe I'm wrong. I've read more than once that there are studies showing that even into adulthood divorce is terribly painful for most children even when the divorce happens and the children are in their 20s and 30s. Be sure to read Larry's post directly above this one. I suspect the only solution is to help people not to marry the wrong person or not to marry at all. Years ago a French sociologist proposed that divorce be prohibited entirely. That would certainly discourage people from marrying for immature reasons. But . . .

Mr. Wiesenthal:Thank you for your kind response. No doubt I miss a great deal. Let me clarify: My comment did not critique the study done at Washington University (the institution from which I received my medical degree). I am familiar with what the study did and did not do and its level II evidence (presumably II-2) is worth some attention, though neither that evidence, nor the long term, user-independent contraceptive methods the study strongly encouraged the volunteers to choose, can be properly described as "bullet proof."Perhaps I did not make clear that I have prescribed a variety of contraceptive measures to young women, in much the spirit you suggest, as a lesser evil than abortion. Furthermore, I cited other studies to suggest that moving from a voluntary study (consistent with a paradigmatic approach) to governmental (or ecclesial) policy does not always have the population effect one hopes or imagines. For example, even the Guttmacher Institute's explanations for reduced teen pregnancy rates in the US require assumptions for which, to my knowledge, hard data don't exist. I'm confident available contraception plays a role, but I have my doubts that relationship is simple, straightforward, or entirely benign. That's all I meant to say. Pardon me if I seemed to suggest otherwise.

For those searching for the latest Guttmacher data ( released 2010 with data through 2006), see: the data is about as hard as one will find, the report includes this under the heading "interpreting the data":Because health department abortion statistics are incomplete or nonexistent in many states, care should be used in interpreting the teenage abortion and pregnancy data. For the states with no information on the age of women having abortions, the rate of abortion among teenagers was estimated. Similarly, error is introduced by the assumption that teenagers have abortions out of state in the same proportions as older women. Therefore, one cannot draw inferences about the effects of parental involvement requirements on the number of abortions obtained by minors.

Hi Dr. Volck,When I used the term "Bullet-proof," I was referring to contraceptive methods with a 99% reliability rate -- namely the IUD and implants. I did not label the study itself as being "bullet-proof," although I believe that the study, in the context of the totality of existing knowledge, was both credible and impressive. It certainly rises well above the level of being "worthless" and "not worth publishing," which was my original criticism of an earlier comment made by someone else (not by you).I'm a medical oncologist myself, and there are many situations where prospective, randomized trials are required (for example, to show a 5% improvement in 10 year survival with the addition of a taxane to cyclophosphamide/doxorubicin in stage II breast cancer). There are other situations where a prospective, randomized trial is not required. For example, here is a single institution experience using historical controls, in chronic myelogenous leukemia (not even concurrent, cohort controls, as in the presently described study): "The median survival was 15% before 1983, 42%-65% from 1983-2000, and 87% since 2001." These improvements were owing to (a) bone marrow transplantation (1983 - 2000) and (b) the introduction of a new drug, Gleevec (since 2001). In the latter situation, not only is a randomized trial not required, but even a concurrent, cohort control study is not required (and, indeed, would be unethical). Reference to the leukemia study:, in the present study, the authors reported the following: "The rate of teenage birth within the CHOICE cohort (50% African-American) was 6.3 per 1,000, compared with the U.S. rate of 34.3 per 1,000." This represents a hazard ratio of 0.18! And the raw number (6.3 per 1,000) would be impressive in any group of subjects, including white middle class adult women of child-rearing age.You brought up levels of evidence, stating that the present study represented Level II (which is, you'll agree with me, well above "worthless"). Let me ask you to answer a question, once you've reviewed the following (from Wikipedia), which is a related topic, namely "Categories of Recommendation:"Categories of recommendationsIn guidelines and other publications, recommendation for a clinical service is classified by the balance of risk versus benefit of the service and the level of evidence on which this information is based. The U.S. Preventive Services Task Force uses:[19] Level A: Good scientific evidence suggests that the benefits of the clinical service substantially outweigh the potential risks. Clinicians should discuss the service with eligible patients. Level B: At least fair scientific evidence suggests that the benefits of the clinical service outweighs the potential risks. Clinicians should discuss the service with eligible patients. Level C: At least fair scientific evidence suggests that there are benefits provided by the clinical service, but the balance between benefits and risks are too close for making general recommendations. Clinicians need not offer it unless there are individual considerations. Level D: At least fair scientific evidence suggests that the risks of the clinical service outweighs potential benefits. Clinicians should not routinely offer the service to asymptomatic patients. Level I: Scientific evidence is lacking, of poor quality, or conflicting, such that the risk versus benefit balance cannot be assessed. Clinicians should help patients understand the uncertainty surrounding the clinical service.Now, my question to you (in your capacity as a pediatrician) is the following: For a teenage girl, who is either sexually active or who intends to become sexually active within 6 months (despite whatever counseling efforts have been made), what would be your "category of recommendation" regarding the totality of available information (including the present study), regarding the advisability (or not) of the teenager being offered the same range of contraceptive choices as were offered to participants in the St. Louis study?A, B, C, D, or I ?- Larry Weisenthal/Huntington Beach CA

To bring the current discussion around to a relevant and timely topic, extensively discussed on Commonweal in the recent past, many conservative critics of ObamaCare object to the mandate to cover contraceptive counseling and services. These critics state that contraceptives are either free or inexpensive, and that it is an unnecessary assault on religious freedom to have a regulation that insurance companies are required to offer private, third party contracts to employees for a rider to their employer-provided policies to pay for these contraceptive services.However, the current study confirms reasonable expectations that "bullet-proof" forms of contraception are dramatically effective in reducing both teenage pregnancy and abortion rates. It should be recalled that this contraception mandate was not a nefarious assault on religious liberties hatched by the evil Sibelius/Obama duo, but was rather the unanimous recommendation of the doctors on the evaluation committee tasked by the highly-respected (and independent/non-partisan) Institute of Medicine to carry out this evaluation. The Obama administration was simply following this recommendation.The "bullet-proof" contraception methods do not have trivial costs, which is why the Institute of Medicine recommended that coverage for contraception be included.Those who are very concerned about the high incidence of abortion in this country should seriously consider the claim that prohibiting employers from preventing their employees from entering into private third party contracts is really a serious infringement on religious liberty. They should then consider the relative degree of evil associated with contraception on one hand and abortion on the other hand.- Larry Weisenthal/Huntington Beach CA

Larry Weisenthal -- Thank you for your effort and careful detail in shedding light on this important topic. An ACOG discussion of Levels of Evidence I, II-1,2,3 is included in "Reading the Medical Literature - Applying Evidence to Practice" and may be useful. The article under discussion says Level II.

Dr. Wiesenthal:I regret that my comments above apparently lead you to believe that I am your adversary. What can I do to persuade you otherwise? While I am well aware of the use-effectiveness of non-user dependent methods of contraception, describing them -- or, indeed, any therapeutic technology -- as "bullet proof" is new to me. Forgive me for stumbling over this novelty in terminology. I agree with you that preventing pregnancies with contraception is preferred to ending pregnancies with abortion. Shall I write for a third time in this thread that I prescribe contraception to adolescents and will do so again? I teach residents how to evaluate evidence AND engage in the terribly inexact process of applying that evidence to the patient before them. No doubt I have much to learn. My reference to Level II evidence above was drawn from the study's abstract, using United States Preventive Services Task Force classification. At no time did I write that the study's evidence was worthless. If you believe I was misusing the very classification scheme referenced in the study, I regret that any imprecision on my part. My point all along -- at which, judging from your response, I have failed -- has been to suggest that applying studies to policy is never simple and almost always have unanticipated, unintended consequences. My comments have not been about evidence, but application. Perhaps we are talking past each other. Perhaps my experiences in Native American health, US poverty medicine, and global health have made me overly suspicious of government programs designed to help. In any case, there's not much more for me to add.

However, the current study confirms reasonable expectations that bullet-proof forms of contraception are dramatically effective in reducing both teenage pregnancy and abortion rates. Larry, the question isn't whether particular types of contraception work, but what are the society-wide effects of a social policy of handing out those types of contraception. The study does nothing to resolve that question. As you cannot dispute, selection bias contaminates the study from top to bottom. This means, for example, that the women who signed up for such a study may be quite different from everyone else, and perhaps they would have found a way to avoid pregnancy regardless. It would be intellectually dishonest to suggest that this study tells us anything about women who were not study participants.

Larry -- I just realized that the fundamental confusion here may be that you're thinking about the question, "Do IUDs and implants work in preventing pregnancy," to which the answer is obviously yes, whereas I'm thinking about the question, "What exactly will happen if there is a social policy of handing out free IUDs and implants," which is a very different question, as it depends entirely on exactly how many people are enticed to sign up and how seriously those people would have tried to avoid pregnancy even in the absence of the program. The latter question, which is more relevant to the HHS mandate, is not answered by this study. That much is indisputable.

Hi Stuart,No, what I was "thinking about" was the following: In a group of women who are motivated to ask their doctors to prescribe "bullet proof" contraception, what is the impact, with regard to teen pregnancy and abortion, of providing free contraceptives? Isn't this precisely the question most relevant to the likely impact of the contraception-coverage mandate?Under ObamaCare, coverage for contraceptive services is not automatic. Rather, the individual woman must enter into a private contract with her health insurer to add a no-cost rider to her policy to cover payment for contraceptive services. Women who are not proactive in obtaining this rider do not receive coverage for contraceptive services.Thus, the study is directly relevant to the question at hand. The study conclusively shows that when poor minority women who want to utilize bullet-proof contraceptives are provided with these contraceptives at no cost to them, the observed rates of teenage pregnancy and abortion are extremely low -- far lower than ever previously reported in these demographic groups and lower than presently reported in the USA in more favorable demographic groups.These data provide powerful support for the unanimous recommendation by the physician members of the (independent, non-partisan) Institute of Medicine panel that coverage for contraceptive services be a mandated component of the Affordable Care Act. - Larry Weisenthal/Huntington Beach CA