Dear Editor

Erin and Harris [1] suggest that we need a regulated market in live donor organs to make good the shortfall in organs available for transplantation. However, the example of the third world shows that to sell a kidney is virtually always an act of desperation when other options for raising money are exhausted, for example Goyal et al.[2] reported that 96% of participants in their survey of kidney sellers in India did it to repay debt. In these circumstances there would have to be grave concerns that any potential sellers could not be freely consenting, as they would effectively be being coerced by their social circumstances. In addition, recent scandals in the financial services industry should remind us that abuses can occur even in regulated, first world, markets.

References

(1) Erin CA, Harris J. An ethical market in organs J Med Ethics 2003;29:137-8.

(2) Goyal M, Mehta RL, Schneiderman LJ, Sehgal AR. (2002) The Economic and Health Consequences of selling a kidney in India. Journal of the American Medical Association (JAMA). 2nd Oct 2002. 288 (13) 1589 - 93.

Dear Editor,

I have two comments to make with regard to the article about the rhythm method. It is not true that the Catholic Church only approves the “rhythm” method of family planning. It approves all natural methods and especially the Billings Ovulation Method of natural fertility regulation that has a better than 99% effectiveness rate, as shown in numerous recent published studies. (www.woomb.org)

On a philosophical level L. Bovens’ argument is flawed in trying to equate natural miscarriages with induced abortion of a healthy embryo. A ten year old child would be able to see the moral distinction between nature taking its course as in the case of a miscarriage and deliberately taking action by swallowing a pill, or other means, to end the life of an embryo.

Justin and Sue Fryer

Dear Editor,

Bovens' argument is clever, but it misrepresents Alcorn, and pro-lifers in general, as being concerned about the absolute number of embryonic deaths without distinction as to the cause. But even a cursory reading of Alcorn (e.g. http://www.epm.org/articles/bcp5400.html) makes it clear that Alcorn's concern is not embryonic death in general, but induced embryonic death: Alcorn merely applies the customary ethical distinction between induced abortion and miscarriage to the situation of a fertilized embryo facing implantation.

Alcorn argues that the use of forms of contraception that create an environment "in utero" inhospitable to implantation (IUD, pill) in so doing induces the death of the embryo (which cannot survive long without implantation), making it ethically comparable to induced abortion. In response, Bovens argues that implantation failures can also conceivably occur when periodic abstinence is being used to avoid pregnancy. However, the convoluted probabilistic arguments Bovens uses to estimate how likely implantation failure might occur, arguments that make up the bulk of the paper, are quite beside the point: regardless of the likelihood of implantation failure under such circumstances, any such failure is spontaneous, not induced, and thus, for Alcorn, is a miscarriage, which he does not consider ethically culpable.

It is only at the very end of his paper that Bovens admits to the possibility that his probabilistic arguments about the likelihood of implantation failure can potentially be addressed by making the distinction between artificially inducing an environment "in utero" inhospitable to implantation, vs. its natural occurrence as part of the ovulatory cycle. But he fails to admit that Alcorn makes this distinction all along; Bovens merely states that he thinks the distinction asks "more from the action/omission doctrine than it can deliver", and leaves it at that. But if Bovens is serious about addressing Alcorn, it is here that his paper should start, not end.

Dear Editor,

Some of my critics draw a distinction between the rhythm method and natural family planning (NFP). I take the rhythm method to be any method that relies on abstinence around the time of ovulation. Of course there are various ways to determine when ovulation occurs, including the calendar method (Ogino Knauss), examining mucus (Billings) or checking basal temperature (STM). I do not take this method to cover the use of barrier methods during the fertile period, as some definitions of both ‘NFP’ and ‘the rhythm method’ seem to permit. I find reports on success rates for this method between 75% as the lowest number for typical use and 99.3% as the highest number for perfect use. Should one use numbers for perfect use or for typical use in moral arguments? On the one hand, one could say that a proponent of a method of contraception should not have to take responsibility for people failing to follow proper instructions. But on the other hand, recommendations are for real people and real people are not perfect users. To postulate a 90% success rate does not seem to be out of line with the available evidence. It is probably somewhat too low for perfect use and somewhat too high for typical use.

If the method fails, then how does it fail? If the purpose is to avoid having sex around the time of ovulation, then the following seem to be reasonable answers. (1) The last time of having sex before the period of abstinence was too close to ovulation. (2) The first time of having sex after the period of abstinence was too close to ovulation. (3) Ovulation was atypically early or late during some cycle and though the users checked the markers for ovulation, they failed to determine its occurrence accurately. Or a combination of (1) and (3) or of (1) and (2) are also reasonable. Since we are talking about typical use, such failures could be due to self-deception and wishful thinking. (1) raises the problem of ageing sperm, (2) raises the problem of an ageing ovum and (3) raises the problem of an atypical cycle.

Now comes the main empirical point of contention. For my argument to work, it must be the case that the probability of viability given that a conception occurs with ageing sperm or ovum or during an atypical cycle is lower than the probability of viability given that a conception occurs with fresh sperm and a fresh ovum and during a typical cycle. Both Mark Witty and Anne Williams phrase the objection to this assumption very well. "There is no evidence that there is any variability of viability of the conceptus with time of fertilisation within this narrow window," writes Williams. “Any conception is as viable as the next, barring a fatal genetic or developmental defect; there is no truth to the 'old sperm' or 'old ovum' speculation...," writes Witty. So let us turn to the relevant empirical literature.

Tarin et al. (2000) review a fifty-year literature not only on the effect of ageing gametes on pre-menstruation embryonic loss, but also on fertilization, spontaneous abortions and the pathology of the offspring. A range of studies is reviewed of in vivo and in vitro fertilizations with ageing gametes, involving humans, non-human mammalians, and non-mammalians. These studies are not always univocal. Furthermore, we often have to extrapolate from non-human animal populations, in vitro contexts, and patients with a history of infertility to what might be happening in human populations of normal fertility in in vivo contexts. Sometimes the best we can do is to guess what would provide the best explanation for experimental results. Nonetheless, I do think that there is a trend present which supports the idea that viability given conception may vary with the age of the ovum, the age of the sperm and whether the cycle is or is not typical.

As to ageing ova, I quote: "It appears that ... post-ovulatory ageing of oocytes is associated with: (i) decreased potential of oocytes for fertilization and pre- and / or post-implantation embryo/fetus development." (Tarin et al., 2000: 544) Table 1 (Tarin et al., 2000: 533) contains a range of studies documenting the effect of the ageing of the ovum on embryo/fetus development and mortality. To pick one example, Wilcox et al. (1998) study ovulation, hCG levels and intercourse patterns of a cohort of women attempting pregnancy and find an increase in post- implantation embryonic loss for intercourse on the day of ovulation in human populations. Considering the time-lag between intercourse and fertilisation, these data support the hypothesis that post-ovulatory ageing of ova compromises embryonic survival. (I should add that Wilcox et al. (1998) do not record any conceptions from intercourse after the day of ovulation.) Wilcox et al. (1999) compare late implantations and early implantations. Late implantations have levels of embryonic loss that are radically higher (82% after day 11) than early implantations (13% up to day 9) . However, it is not known what causes these late implantations.

As to old spermatozoa, I quote from Tarin et al., 2000: 544: "Likewise the ageing of spermatozoa in ... the female reproductive tract ... is associated with decreased ... potential for fertilization and pre- and/or post-implantation embryo/fetus development." Table 2 contains entries with articles documenting the effect of in vivo ageing of spermatozoa in the female genital tract and increased embryo/fetal mortality with ageing spermatozoa. Tarin et al. (2000: 542) write: "This notion is supported by the high mortality rate observed in embryos/fetuses derived from sperm aged in stagnant environments, e.g in ... the] female reproductive tract." Parkening and Soderwall, in a study of golden hamsters, write that their data "indicate that inseminated spermatozoa are capable of penetrating the zona pellucida and fertilizing some ova after residing 14 to 16 h within the female reproductive tract, but that the viability of ova fertilized in this manner is greatly reduced." (1975: 627-8) There are supporting results about golden hamsters in Bell and Shaver (1982). Vishwanath and Shannon find that after storage, bull sperm may retain its "oocyte penetrating ability", yet "mitochondrial DNA damage and chromosomal abnormalities" could "compromise the viability of the resulting conceptus." (1997: 321-32)

One might conjecture that if there is a higher rate of early embryonic loss in conceptions involving ageing gametes, then there would also be a higher rate of spontaneous abortions. Gray et al. (1995: 1568) cite a number of studies that confirm this finding. However, in his own study, he does not find a difference in the spontaneous abortion rate for conceptions resulting from intercourse on the day before or on the day of ovulation on the one hand and from intercourse outside this narrow window on the other hand within the total population. There is however a significant difference in the spontaneous abortion rate in the subpopulation of women who have had spontaneous abortions before. There is no conclusive interpretation of these results.

Let us now turn to atypical cycles. There are many open questions here and much of my evidence comes from consulting with researchers in the field. Implantation rates are dependent on the nature of the embryo and the development of the endometrium. In IVF (in vitro fertilisation), ovulation is timed carefully by adjusting drug dosages so that it comes neither too early nor too late. This might indicate that the quality of the ovum could be a function of ovulation time and since we can control for fertilisation in IVF, the concern is a concern about the viability of the embryo. In IUI (in utero insemination), some doctors will refrain from inseminating in case of an early ovulation. This might indicate that the quality of the ovum is compromised when there is early ovulation or that a short follicular phase may compromise the development of the endometrium. Early ovulation is believed to correlate with higher rates of oocyte aneuploidy leading to more pre- and post-implantation embryonic loss due to chromosal abnormalities. About late ovulations, Tarin et al. (2000: 535) write that "...under particular circumstances, including long follicular phases ... fully grown GV oocytes may undergo pre-ovulatory ageing before the occurrence of the endogenous LH surge ..." (2000: 535) He also presents a list of studies documenting the effect of pre-ovulatory ageing on fertilization, the development of the embryo and the offspring. (2000: 534)

Whitty objects to my use of the figure of 50% of embryonic loss, claims that such high figures are based on old and questionable studies (from 1956 and 1975), and that animal studies give single-digit percentages. A standard source for embryonic mortality is Edmonds (1982). Edmonds assesses embryonic loss by the appearance and disappearance of hCG in the urine at the time of implantation and presents a figure of 62%. This underestimates the actual percentage since it does not count pre-implantation loss. In Wilcox et al.'s study (1999) the pre-menstruation embryonic loss plus miscarriages is at 33%. Why is there this discrepancy? The only explanation that I can see is that Edmonds samples from a normal population, whereas Wilcox restricts his population to couples without previous fertility problems. In any case, considering that this does not measure pre-implantation loss, a figure of 50% for normal populations does not seem outlandish in the face of these data.

So far I have tried to defend the assumptions that come into the model. But of course, it is another thing to check whether the predictions of the model hold true in the real world. Clearly we know too little to fill in the values of all the parameters. It may be the case that, say, the viability given conception is variable, but this variability is not sufficiently great to obtain meaningful differences between embryonic death rates for condom users and NFP users. Or there can be hidden variables that are not included in the model. In principle, it is possible to do the following empirical test. One could compare post -implantation embryonic loss rates between condom users and NFP users by checking for hCG in urine tests. However, in practice, this would require quite a large number of subjects, because in any given cycle, there would only be a relatively small number of conceptions. This would be one step forward, but it would still not yield information about pre-implantation embryonic loss, because hCG does not register in the woman’s blood or urine before implantation.

Let us turn to the philosophical objections. My critics claim that even if I were right about the empirical data, there would still be a moral divide between NFP and contraceptive pills, because the embryonic deaths occurring on NFP would be "natural deaths" whereas the deaths occurring on contraceptive pills would be "induced deaths", in Ira Winter’s words. Justin and Sue Fryer talk about "letting nature take its way" versus "deliberately taking action by swallowing a pill, or other means, to end the life of an embryo". And Mark Whitty points to the relevance of the doctrine of double effect.

These are difficult issues that require a much more careful analysis than what I can offer at this point. Let us clearly bracket the empirical issues from the philosophical issues. My critics say that even if I were right about the empirical details, then we could still invoke the action/omission doctrine or the doctrine of double effect to draw a distinction between NFP and contraceptive pills. So let us not quibble about empirical data from here on and suppose—to make things simple—that a particular implementation of NFP and a particular contraceptive pill, would lead to the same number n of conceptions, lead to the same number of embryonic deaths d, and lead to the same number of live births m. And let us suppose that condoms lead to the same number m of live births but with fewer conceptions and hence fewer embryonic deaths. A final alternative is abstinence with carefully targeted pregnancies, leading to m life births with even fewer conceptions and even fewer embryonic deaths.

Let us do the following admittedly fanciful thought experiment. Suppose that conceptions always lead to live births. However, all these contraceptive techniques would have the same mechanisms and consequences as they now have, substituting neonatal death for embryonic death. So a contraceptive pill user would have n conceptions, n live births, but d deaths of newborns and hence only m children that survive for longer than, say, one week. Similarly for NFP, condom usage and abstinence with targeted pregnancies. Would there not be an outcry about any method that would have an excessive neonatal death rate? Could we defend NFP on grounds that the deaths of these newborns is just nature’s way or on grounds that these deaths are not intended? I don’t think so. If embryos have the same moral status as newborns, then why would we think that we can put up such defences for NFP involving embryonic death?

Now I am a friend of the action/omission doctrine and the doctrine of double effect. For example, as to the action/omission doctrine, I do agree that one has to have better reasons—i.e. the prognosis has to be worse—before one is justified to withdraw treatment (action) rather than withhold treatment (omission). As to the doctrine of double effect, suppose that I can advance the war effort to the same extent by target bombing, say, by bombing a depot, while expecting to kill n innocent civilians living around the depot as collateral damage, as by terror bombing, say, by bombing a neighbourhood with the intention of killing n innocent civilians. I do agree that terror bombing is worse than target bombing. But just to say that bad effects did not come about due to direct agency or were not intended is not sufficient to make a practice permissible on the action/omission doctrine or on the doctrine of double effect. The bad effects also have to be outweighed by the good effects and furthermore, there must be no other practice (in which the bad effects do not come about due to direct agency or were not intended) that achieves the same good effects while incurring fewer bad effects.

Further to the action/omission doctrine: There is a continuum between agency (say, a direct killing of an embryo) and an omission (say, not providing a drug that would prevent a miscarriage). Now one might argue that, on this continuum, "taking a pill so that fertilisation is unlikely to occur or so that the endometrium is inhospitable for incoming embryos" is more on the action side than "organizing your sex life so that fertilisation is unlikely to occur or so that embryos are either non-viable or come in at times when the endometrium is not hospitable". But is the distance on this continuum sufficiently great to make the former qualify as a grave sin and the latter as a morally permissible method of birth control? I am not convinced. Maybe it is worse to remove edible plants and animals from an island to make it inhospitable and then to drop off someone on it than to carefully pick a time of the year to drop off someone when you know the island not to be hospitable for human habitation. But really, is it that much worse? And furthermore, even if, given our assumptions about the numbers, contraceptive pills are worse than NFP on action/omission grounds, then would one not want to reverse this judgment if embryonic death rates for NFP would come to exceed embryonic death rates for contraceptive pills to a sufficient degree?

Clearly we carry some responsibility for the consequences of our omissions. If embryonic deaths are a bad thing, then should we not try to minimize them? Indeed, we should take due care, also in our omissions. And, given my assumptions, there are alternatives available, viz. condoms and abstinence with targeted conceptions, that come at a lower embryonic death rate.

I have two concerns with invoking the doctrine of double effect. Clearly it is not enough to say that an NFP user does not intend the embryonic death. This would get a pill user off the hook. She could just say that she only intends to block ovulation by taking this pill and furthermore she intends this as a means not to become pregnant. A proponent of the doctrine of double effect would just say to the pill user that if embryonic death were not to occur she would shoot short of her target – there would simply be more pregnancies. Embryonic death is integral to reaching her goal while, say, the deaths of innocent civilians in target bombing are not. But the same response holds for the NFP users. Without the embryonic deaths, there would be more pregnancies.

Secondly, target bombing would be no longer permissible if there were another way to gain the same military advantages at a lower cost in the lives of innocent civilians. But is this not what we are facing in the case of NFP? There are alternative ways to gain the same advantages at a lower cost of embryonic deaths, e.g. by condom use. Now I realize that there may be independent grounds for opposing condom use as a malum in se, but then there is the possibility of abstinence with targeted conceptions. Granted, the gains of sex are lost in abstinence, but do they weigh up against embryonic deaths?

Clearly there are many open questions in this area. And I have not even touched on questions about the status of the embryo. These are not only questions in the context of the pro-life versus pro-choice debate. For instance, some critics have pointed out that many of my embryonic deaths are deaths of defective embryos and defective embryos are not possible persons. Other critics have claimed that embryonic deaths as such are not a bad thing, but that killing an embryo is nonetheless a wrong action. Many such questions remain unexplored.

My intention here was to point out that an appeal to the action/omission doctrine or the doctrine of double effect does not straightforwardly block the argument. Maybe it can be blocked. To make headway on this issue will require some serious thinking and constructive dialogue.

Let me take this occasion to thank my critics in JME, e-mail correspondents and bloggers for their thoughtful comments. I am also grateful for helpful (e-)conversations with Roger Gosden, Kathy Hoeger, Kathy King, Michael Otsuka, Alexander Pruss, Danny Schust, Alan Thornhill, and Alex Voorhoeve.

References

Bell, C.L. and Shaver E.L. (1982) Analysis of preimplantation golden hamster conceptuses resulting from spermatozoa aged in utero. Gamete Res., 6: 199-207.

Edmonds, D.K., Lindsay, K.S., Miller, J.F. Williamson, E. and Wood, P.J. (1982) Early embryonic mortality in women. Fertil. Steril., 38, 447- 53.

Gray, R.H., Simpson, J.L., Kambic, R.T., Queenan, J.T., Mena, P., Perez, A. and Barbato, M. (1995) Timing of conception and the risk of spontaneous abortions occurring during the use of natural family planning. Am. J. Obstet. Gynecol., 172, 1567-72.

Parkening, T.A. and Soderwall, A.L. (1975) Delayed fertilization and preimplantation loss in senescent golden hamsters. Biol. Reprod., 12, 618- 31.

Tarin, J.J., Perez-Albala, S. and Cano, A. (2000) Consequences on offspring of abnormal function in ageing gametes. Hum. Reprod. Update, 6, 532-49.

Vishwanath, R. and Shannon, P. (1997) Do sperm cells age? A review of the physiological changes in sperm during storage at ambient temperature. Reproduction, Fertility and Development, 9, 321-32.

Wilcox, A.J., Weinberg, C.R. and Baird, D.D. (1998) Post-ovulatory ageing of the human oocyte and embryo failure. Human. Reprod., 13, 394-7.

Wilcox, A.J., Baird, D.D. and Weinberg C.R. (1999) Time of implantation of the conceptus and loss of pregnancy. N. Engl. J. Med., 340, 1796-99.