点击显示 收起
It is useful to recall that reducing the prevalence of anemia in women of childbearing age by one-third was a goal of the World Summit for Children [Black M. World summit for children. Lancet 1990;336:1586]. This goal was subsequently endorsed at several other world conferences and has been largely agreed to by governments all over the world and by aid agencies, which are contributing to our efforts to tackle that problem. My understanding of that goal is not predicated on the assumption that it was a necessary prerequisite to reducing maternal mortality. Reducing anemia and reducing maternal mortality was inexplicit and has remained so. The implications of the World Summit for Children were that anemia in women was a significant problem in its own right and it was correct to take steps to reduce the problem not merely because it contributed to mortality. Subsequently, that goal was expanded to include tackling anemia in children. We might want to discuss the implication of yesterday's discussion for overall action to tackle anemia in women of childbearing age.
It may be also useful to recall that the World Summit for Children in 1990 elaborated a goal of reducing the prevalence of low birth weight. It was also not predicated on reducing the prevalence of low birth weight in certain groups of women or avoiding actions to reduce the prevalence of low birth weight in women who were <1.5 m tall, for example, which was suggested by Dr Rush to be something we should look at.
An area of yesterday's discussion focused on what is known about the relation between iron deficiency anemia and maternal mortality. Again, we must recognize that the justification for most iron supplementation programs has not primarily been maternal mortality reduction. It is true to say that attempts are being made, most notably in the paper commissioned by Jay Ross [internal United Nations Children's Fund report, unpublished observations, 1998], to calculate to what extent maternal mortality might be expected to be reduced if anemia were to be resolved. A figure of 20% has been suggested. Dr Ross's review was admittedly based on inadequate data and inadequate studies, although I think he reviewed some of the same studies that Dr Rush reviewed. Dr Rush's interpretation of the data may be slightly different from Dr Ross's interpretation.
There are 2 working hypotheses based on what both reviewers would likely agree to be inadequate data. Dr Rush described the enormous burden of severe anemia in some parts of the world; for example, in South Asia, 20–30% of women were said to be severely anemic (defined as hemoglobin <70 g/L). If we agree that this is a huge public health problem, is it not fairly certain that a substantial proportion of this anemia is caused by the consumption of diets that are very low in absorbable iron? That is an important cause but certainly not the only cause. Is there not reasonable evidence that this anemia contributes significantly to maternal morbidity, indicators of poor maternal health, and lack of energy important for self-care and for child care? Are the data not fairly strong relating the presence of anemia to work output? For example, recent studies in Indonesia showed that factory workers were simply less productive if they were anemic and that it was a good economic investment to tackle anemia if women were engaged in the work force [Scholz B, Gross R, Schultink W, Sastroamidjojo S. Anaemia is associated with reduced productivity of women workers even in less-physically-strenuous tasks. Br J Nutr 1997;77:47–57]. The same increases in productivity might take place in work, for example, in self-care or in child care.
Did the discussion yesterday have implications for the position that we should be making every effort to reduce the prevalence of iron deficiency anemia through a range of actions including diet improvement, fortification, and supplementation, in which several of our and many other agencies are already engaged? Many or most developing countries currently support programs to provide iron supplements to pregnant women. A few are going beyond that. A few are experimenting with giving iron supplements to all women of childbearing age. I think it would be fair to say that the agencies supporting these programs, of which the United Nations Children's Fund is one, would regard such programs to be far from perfect. Typically, 50% of women might be getting 50% of the iron supplements they need.
There have been extensive analyses of iron supplementation programs, and the reasons for lack of success or limited success include many factors that Dr Ladipo discussed [Ladipo I. Nutrition in pregnancy: mineral and vitamin supplements. Am J Clin Nutr 2000;72(suppl):280S–90S]. Compliance is an issue that is being hotly and actively debated. Simple lack of supply and logistics and the priority that is given to making iron supplements available is another constraint. We heard Dr van den Broek [van den Broek NR, Letsky EA. Etiology of anemia in pregnancy in south Malawi. Am J Clin Nutr 2000;72(suppl);247S–56S] say that for a whole year in Malawi there were no iron supplements available. The supply network did not work, which was the prime reason that people were not taking supplements.
As we discuss program implications we need to remember that the International Nutritional Anemia Consultative Group produced a document endorsed by UNICEF and by the World Health Organization [Stoltzfus RJ, Dreyfuss ML. Guidelines for the use of iron supplements to prevent and treat iron deficiency anemia.Washington, DC: ILSI Press, 1998]. It is a very upbeat document and is an attempt at new guidelines for practical approaches to improving supplementation for women of childbearing age, pregnant women, and young children. It is UNICEF's belief that this is a useful and good piece of work, and we are looking forward to this document being published and used as part of the strategy of reenergizing, retooling, and strengthening programs. How does Dr Rush's document that we have before us relate to the INACG-WHO-UNICEF programming document? We need to be rather careful that we are not shooting ourselves in the foot by summing up 2 different documents, one saying that these are the new recommendations of what works in iron supplementation—let us roll up our sleeves and get this one going—and another document saying that there are many worries about iron supplementation.
I believe the INACG document produced by a concentration of 20 or so of the world's experts in iron supplementation did not consider the safety of iron supplements in pregnancy to be a major issue. So, it was hardly addressed at all in that draft document. Yesterday, discussion was postponed on whether safety should be a much more major concern. Programmatically, where do we take this? Is the INACG document still appropriate? Does it need to be revisited and greater caveats on safety put in? Is safety a major concern or is it a relatively minor concern?
Dr Heywood: I hope that we do not feel ourselves too bound by what other groups of people have done before. If we do that, we keep revisiting issues and will never have to do anything because many contradictory things have been said before. I do not think we should take that as the starting point. We should take the starting point as what is on the table and what is being said now.
Dr Beard: I see 2 issues in front of us. One is whether there is enough biological evidence to suggest linkages or relationship that we had not thought of before between either too little iron or too much iron for pregnancy outcomes. Do we have new evidence that was not available to other committees, for example, the Institute of Medicine committee that Dr Rush and I served on 5 y ago, that we need to reconsider in the biological relationship between iron and maternal health and perhaps, also, fetal growth and fetal outcomes?
The second issue is whether more data on the implementation of programs are needed to address this relationship that, as Dr Rush has done in his document, questions the effectiveness and perhaps efficacy of what we are doing currently. I suggest that we consider the first point—whether the biological evidence in front of us says that we need to consider whether the administration of iron in pregnancy is beneficial.
Regarding Dr Bothwell's [Bothwell TH. Iron requirements in pregnancy and strategies to meet them. Am J Clin Nutr 2000;72(suppl):257S–64S] presentation and our calculations of iron needs, I see no data that say that there is anything wrong with those calculations or the evidence. In all likelihood, many women are not going to be able to meet their iron needs in pregnancy with their usual diet.
Dr Maine: We should clarify the endpoint. Where, when, and how iron supplementation is appropriate constitute one meeting. Maternal mortality, the role of iron or nutrition, and the program implications of that role constitute another meeting. I think we ought to clarify which meeting we are at. May I say that the second meeting has never been held. The first has, many times.
Dr Johnston: We need to consider how we are viewing the nature of the problem. Is the problem anemia or is it iron deficiency? The 2 are linked, but focusing on iron is not going to solve the problem of anemia. I suggest that we focus on anemia first. Look at all the determinants and then where iron fits in relation to the whole picture of anemia.
Dr Caulfield: We need to ask what is too much or too little iron for. The mandate here was to discuss the relationship between anemia and mortality, to consider whether moderate and severe anemia are related to maternal mortality, and to consider whether the provision of iron can reduce or, as Dr Rush has suggested, cause mortality.
Dr Fleming: Regarding maternal mortality, one has to go back 30 or more years before there was rational blood transfusion. Going back those 30 y, the first reason for looking at the prevention of pregnancy anemia in Nigeria was to prevent maternal death. That was the starting point because we were then seeing about 20% maternal deaths due to severe anemia—not iron deficiency—and we must look at anemia. I think it is wrong for Dr Rush to say that we do not understand the pathophysiology of anemia. We do. The treatment of severe anemia is by blood transfusion, which undoubtedly saves lives. Today with the collapse of economies and other disasters affecting health, I am quite sure that in Africa, and probably in Asia as well, many women are still dying of profound anemia related to pregnancy. Figures do appear, and I have seen women out in the bush at the point of death from anemia.
Mr Alnwick mentioned the question of work capacity of moderately anemic women. I think we have looked at that in women, and it has a major effect on the other children and the whole family. Hypoxia is a word that we do not hear very much about but that describes the condition we are discussing—not lowered hemoglobin, but hypoxia. I quite agree that there are 2 issues, but I do not think that we really should be too concerned about whether iron is toxic. The case has not been made. You should not talk about a toxic level of hemoglobin. There have been studies with high hemoglobin in pregnancy resulting from low plasma volume. It is only when this polycythemia is in the range of a hematocrit of 10 or 10% higher that you get hyperviscosity.
There is really no evidence that increasing birth weight has been followed by any obstetric complications. The dangers have not been shown.
Dr Labbok: Dr Yip's comments [Yip R. Significance of an abnormally low or high hemoglobin concentration during pregnancy: special consideration of iron nutrition. Am J Clin Nutr 2000;72(suppl):272S–9S] summed up information that anemia and iron levels are proxies, often markers, for the actual cause. If we are talking about maternal mortality, I do not think we can just look at iron or at anemia. Much research has done just that. Some of the discussion must be on causation or what anemia is a marker of, not that anemia causes hemorrhagic death, which we do not really know. If we are talking about maternal mortality, I agree that we have to look at anemia as a marker of other things going on that might be involved in the totality and not just iron deficiency.
Participant: I am not convinced that there are program implications from yesterday's discussions or that we know something so definite that we would change the INACG publication's program recommendations. We have many research recommendations, whether categorized as Dr Beard's first or second issue, and a lot of new data are needed, not just a review of the evidence. I do not see a large conflict between Dr Rush's review and the INACG document.
As Dr Heywood says, let us just take the topics that we are considering today and the very provocative questions from yesterday. We do not need to issue a document from this meeting saying that everything is now up in the air, do anything programmatically, or say that what we have been doing for years may be wrong. Let us just go ahead and generate the new research; there are more questions in the research area now than in programming.
Mr Alnwick: My response to Dr Maine's comment is that this is a meeting addressing the narrower issue of iron supplementation, iron deficiency anemia, and maternal mortality and relating these to programmatic issues. Are we comfortable with the point that 2 things can run concurrently, but we accept that anemia in developing countries—at 40–60%, which may be attributable to iron deficiency—is undesirable and that the programs to tackle anemia including in pregnancy are rather useful? The question we are asked to discuss is to what extent will those programs—besides making women feel better—reduce some of the sequelae of anemia in those women. To what extent will—or can—these programs be expected to contribute to reducing the terrible toll on maternal mortality?
Dr Rush: Dr Fleming said that plasma volume expansion accounts for the association between high hematocrit or hemoglobin levels in late pregnancy and adverse outcome. This is a plausible hypothesis without direct data to either support or refute it. The strong association observed by Murphy and colleagues [Murphy JF, O'Riordan J, Newcombe RG, Coles EC, Pearson JF. Relation of haemoglobin levels in first and second trimesters to outcome of pregnancy. Lancet 1986;1:992–5] between high hemoglobin concentrations early in pregnancy, before any appreciable expansion of plasma volume, and later hypertension is more consistent with a toxic effect of hemoglobin level per se.
There are 2 distinct issues about anemia and maternal mortality. First, should special provision be made for the severely anemic pregnant woman? Given the facts, I cannot believe any of us would disagree that there should be. I believe we can all agree that it should be. Dr Fleming has pointed out that there is an effective therapeutic response—transfusion—for very severe anemia [hemoglobin <40 g/L]. However, the infrastructure for delivering transfusion therapy is nonexistent in much of the world. Can we change this sad fact, and in the interim, are there alternatives? Second, if there is no relationship between mild or moderate anemia and adverse pregnancy outcome, including maternal mortality, should the current policy of universal supplementation be reconsidered? About this, some of us disagree. If maternal mortality and most fetal problems are associated only with severe anemia, exclusive reliance on the current policy of routine, universal, and unmonitored supplementation is very unlikely to address the problem effectively.
Dr Fleming said that there is no evidence that danger to the mother or infant increases with increased birth weight. I do not agree. Although there is no evidence that increased birth weight induced by nutritional supplementation causes increased danger, this may be because no one has directly explored this issue. There is very strong evidence that increasing birth weight is associated with increasing risk of obstructed labor, which I presented yesterday [Rush D. Nutrition and maternal mortality in the developing world. Am J Clin Nutr 2000; 72(suppl):212S–40S]. With larger fetal size, the chances of trauma to both the mother and fetus increase. It is imperative to demonstrate that programs aimed to increase fetal growth are safe.
Mr Alnwick: Can we reach a consensus to move forward? Many of us agree that much more needs to be done regarding severe anemia, which is a problem, and routine supplementation programs are unlikely and unproved to be effective in preventing severe anemia. The continuum of mild to moderate to severe anemia is a gray area. Programmatically, we are saying that more needs to be done and we need to have more effective ways of tackling severe anemia.
Dr Rush: If we limit ourselves to considering the problem of severe anemia, we not only would be addressing the rationale for this meeting, but we might pave the way toward a major advance in public health. New resources will be necessary both to do the needed research and implement new and refined programs.
Mr Alnwick: You mean that we should maintain the status quo of routine supplementation programs, admitting this is a gray area—that data are not sufficient for us to be very clear about the effectiveness of routine supplementation programs in preventing maternal mortality—but that we should accept that there may be other benefits from routine supplementation programs. You are also saying, "Let us not get into those," as was Dr Maine, so we should put aside all the other morbidity-related benefits from routine supplementation programs, work capacity, and so on, and say that those have been debated before and guidelines exist for them. Let us leave them alone for the moment and let us concentrate on the mortality.
Dr Labbok: I got 3 things out of the meeting yesterday vis-a-vis maternal mortality. First, there are hints that combining things, such as vitamin A, with iron is associated with changes in maternal mortality. Thus, there is a need to consider a multiple vitamin and mineral supplement. Second, there is clearly a need to consider taking it outside pregnancy, whether it is supplementation or fortification throughout a lifetime. Third, which I think is the most difficult, is that neither anemia nor iron deficiency causes the mortality per se. We need to try to look very directly at the immediate causes of mortality and the things that come before. What is the preexisting nutritional status associated with these different causes of anemia? We know, for example, that we are not properly defining obstructed labor. It could be that certain nutritional deficiencies are associated in some way with problems that lead to the immediate causes of death. Anemia does not cause death but it is highly associated with death from hemorrhage because you are already depleted. The immediate cause of death is hemorrhage. In this way we could probably tease out more directly the prevention of the conditions that support the direct causes. Nutrition is preventive. Generally speaking, when we talk about vitamin and mineral supplementation and fortification, we are talking about prevention, keeping people healthy. We are not talking about running in there and stopping the hemorrhage. If we are talking about prevention, and if we are talking about nutrition, research is definitely needed directly on the causes of maternal mortality and the nutritional situations associated with it.
Dr Caulfield: The papers we have presented here, and the whole structure of the conference has, by definition, focused on a particular aspect of the whole series of questions related to iron and maternal mortality—whether iron supplementation can prevent moderate and perhaps severe anemia and its relation with mortality. I do not think there is any disagreement that severe anemia causes mortality. The question, rightly put, is, What are the causes of severe anemia and is nutritional intervention (meaning iron supplementation) going to prevent that? Some evidence from Africa shows that the causes of severe anemia, including iron deficiency, play a role but that we also need to prevent malaria. We need to deworm women. There are lots of interventions to attack severe anemia, if that is the only part of anemia that kills women during childbirth. The issue is whether moderate anemia [hemoglobin 80 or 90 g/L] along with blood loss puts women in a position to die regardless of how they deliver their infants. Is iron supplementation going to move them up a little bit on the hemoglobin ladder?
Dr Rush has said that we could move them up theoretically although it is possible that such an increase in hemoglobin is not effective. He also argues that we could move women up to 130 g/L and cause other problems during labor and delivery, which is why it is necessary for us to focus on those issues. We cannot even begin to discuss vitamin A, zinc, protein energy, improving body mass index, improving weight gain of women, or any of those other issues because by definition we have not assembled the group to talk about those issues.
Dr Kitange: A lot of findings—both the Malawi study and our knowledge of what causes iron deficiency—support the idea of multiple supplementation for anemia. Iron supplementation might not solve the problem of anemia. We need to rethink the issue of multiple supplementation.
Dr Martorell: The comments that I have been hearing have more in common than not. I agree with Dr Caulfield that the meeting has been structured—and the presentations have focused—on the role of iron in maternal mortality, and to examine the question in its broadest fashion would require more information and more time.
I liked Dr Beard's questions and I came to some of the same conclusions as Dr Labbok. For example, we heard that if you only change iron in a setting where you have multiple nutrient deficiencies, you run into other limiting factors, and there is new evidence of nutrient interactions. I conclude that perhaps we need multiple micronutrients. We also heard about the great demands for iron during pregnancy. It is often said that pregnancy is not the time to correct anemia; by then it is much too difficult. Perhaps we should look at other approaches to help women with iron and folate. I have been hearing lots of agreement among us, more agreement than is apparent at the moment.
Dr Belizan: I would be pleased if we pose provocative questions to help researchers in the developing world. There are not many interventions proven to decrease maternal morbidity and mortality, which reflects how weak our knowledge is about our recommendations. After listening to comments coming from this meeting and seeing the limitations in the provocative paper by Dr Rush, I think that we are identifying what evidence really exists for making recommendations and what we need to know. The effort of many agencies and researchers is to identify what we need to know and how to do good research to answer what we do not know.
Mr Alnwick: Is there a consensus on Dr Belizan's statement that we need more evidence? If we agree, it means we need more research. Let me just caution to say that there are some things where the world is moving forward without evidence, for example, the benefits of brushing one's teeth in the morning. There has never been a controlled trial to show that brushing prevents dental caries or anything else. The world, give or take 60%, has iodized salt. There was never to my knowledge a randomized controlled trial of salt iodization. It was acknowledged that the benefit to be obtained from salt iodization was sufficiently clear, and the world could move forward without the need for further research or further randomized controlled trials. There are some things, rightly or wrongly, that have been done without further research. Are we of the opinion that in this area we need more research?
Dr Guidotti: WHO has the responsibility to provide guidelines that UNICEF and a lot of other agencies use. Obviously, this meeting is very important. I would like to relate how decisions were made when I joined WHO about 16 y ago. We would have a meeting like this. Someone, like Dr Rush, would present a paper. Other people would present other papers, and at the end of the meeting we would come to a consensus and write guidelines. That was the best we could do. Of course, those guidelines were predicated sometimes on personalities at the meeting. Later, Ian Chalmers came up with an alternative and he challenged us to an evidence-based method.
I assume you all know Dr Chalmer's work and what he does with meta-analysis and the Cochrane database [Cochrane Collaboration, Adelaide, South Australia, 1998]. One statement in the Cochrane database on routine iron supplementation during pregnancy, which was reviewed by Kassam Mohammed [Iron supplementation in pregnancy. Issue 2. Oxford: Update Software, 1999], is that there is no strong evidence to advise for or against a policy of routine iron supplementation in pregnancy. On the one hand, I have to sit with this challenge on my desk and, on the other hand, we have to realize that the world has been moving ahead without evidence on every single action made by humans. We may have to look for the evidence and if it is not there, we may have to do more research, but I would also like to try to get clarification. Dr Rush's review shows that there may be evidence against routine supplementation, which, I think, is the crucial part of this meeting. We have to decide after reviewing the observational studies whether there is strong evidence against iron supplementation and its effect on maternal mortality or any maternal health.
Mr Alnwick: So, if the FAO-WHO guidelines on iron intake say that people should be getting x mg of iron per day, and we do a reasonably accurate dietary survey in Bangladesh and find that women are getting 40% of requirements, we should not be concerned until we have evidence that this is causing some undesirable outcome? Or, if we observe that iron intakes in a large population are <50% of what the world has decided is a basic intake or a basic minimum, should we do something about it?
Dr Maine: We have not been prepared to deal with the issue of health. I propose that we talk about the role of iron in maternal mortality; the attributable portion; and the kinds of anemia that might be important based on Dr Rush's paper and other people's input, the research implications, and the program implications.
Dr Loudon: I am interested in the disparity you pointed to. Either we have to decide to stick to iron and maternal mortality or we have to say, having discussed it, that we feel it is unrealistic to narrow the topic to iron deficiency, and we ought to look at anemia and maternal mortality. That is the very first thing we have to decide, because what we are going to discuss is going to be totally different based on this decision.
A second point: from time to time speakers talk about maternal morbidity. If anybody is going to speak about maternal morbidity, I would like them to tell me how they measure it. From a historical point of view, I was repeatedly asked why I was not talking about maternal morbidity and, after many attempts to do so, I decided it is an unmeasurable quantity.
Dr Harris: I am very cognizant of Dr Maine's frustration with how we keep moving around. One of the treatments of anemia that is offered is iron supplementation. One reason that iron supplementation is offered as a treatment for anemia—at least we say it is one reason—is that it is going to reduce maternal mortality. The question is, Do we know that? Some say yes and some say no. What are the issues that we can narrow the discussion down to? How do we decide whether the question is still open or has not been fully looked at? How do we decide what we need to do? We are not set up to deal with the broader nutrition issues.
Dr Maine: As far as we can tell from the very poor studies done, not all anemia is fatal, only severe anemia. That has a lot of program implications. What do we need to know that we do not know to design better programs? What are the program implications of what we know?
Dr Caulfield: The critical point of discussion to begin with is precisely what was just articulated. Is it only severe anemia that may kill women or do mild and moderate anemia also kill women, and is it during childbirth or around the period of childbirth, etc?
Dr van den Broek: I am keen also to look at feasible things we can measure. I was impressed with Dr West's [West KP Jr, Katz J, Khatry SK, et al. Double blind, cluster randomised trial of low dose supplementation with vitamin A or beta carotene on mortality related to pregnancy in Nepal. The NNIPS-2 Study Group. BMJ 1999;318:570–5] studies with a vast number of women that still do not provide conclusive evidence. Can we, going back to the toothbrush issue, not assume that anemia is something that contributes to maternal death? It is very difficult to measure the "how," and I am not sure we need to put more effort into more programs and research to measure that directly. Can we not assume that anemia—if your population is on the left side of the curve—is going to be related to high maternal mortality figures, so that you need to shift the whole curve to the right?
Dr Maine: I think the data that we have on maternal mortality do not support a major role for nutrition. I can give you 2 examples. One is all of Western history, which Dr Loudon talked about. Anemia and other things that related to the standard of living and people's well-being decreased and maternal mortality did not. The other is from Bangladesh [Maine D, Akalin MZ, Chakraborty J, de Francisco A, Strong M. Why did maternal mortality decline in Matlab? Stud Fam Plann 1996;27: 179–87], which is where we have the best data on the poorest population, where maternal mortality has been slashed in a fairly short period, say, under 10 y. We are not—even those of us who have spent a lot of time there looking at it—sure how this occurred, but nobody is saying that the nutritional state of the population improved. It has something to do with medical services.
Dr Caulfield: The iron status of those women has improved.
Dr Maine: What has not improved is the body mass index of those women. Yes, there are data to suggest that the iron status has improved significantly.
Dr Caulfield: Maternal mortality fell in very particular areas: 2 parts of the Matlab area, one where there was an intensive maternal mortality intervention involving transportation, comprehensive obstetric care, village midwives, and cesarean sections. The other was a place where nothing was done. It was part of the old control area but it was very close to the hospital that was functioning fairly well, and mortality fell just as much. This finding is a mystery. One thing we can say is that certainly in the control area—where there was no International Centre for Diarrhoeal Disease Research program, only the old government service—nutritional improvement or iron improvement was probably not an explanation. On the basis of limited data, nutritional iron deficiency anemia contributes a fairly small part of maternal deaths. However, we are concerned about all maternal deaths and how we could prevent those. Sometimes people say 30–50% of maternal deaths are due to anemia, but I do not think there is any support for such a statement.
Dr Rush: I concur with Dr Guidotti that the major danger of current iron supplementation programs is that they may divert us from other problems that need to be addressed. There also remains the question of the possible toxicity of supplemental iron, which cannot be dismissed without further research. If, without fresh data, we were asked whether current programs are of much benefit in solving the problem of severe anemia, I believe we would have to conclude that the odds are very slim indeed that these programs could have much effect. I do not think we should or can address in this forum whether they could induce a small shift upward in the population distribution of hemoglobin concentration, nor what the value of such a shift might be. We can concur that severe anemia is almost certainly a problem that is serious and that needs urgent attention. We need to better define both the extent of the problem and how best to respond. We need to know more about the relationship between severe anemia and maternal outcomes, particularly death, and what to do to try to improve those outcomes.
Mr Alnwick: You have raised a very interesting issue that we should not forget and should come back to. Are current iron supplementation programs diverting resources away from something that could be more useful? This is a very important question that I have been asked repeatedly. For example, is the thrust toward measles control through immunization diverting resources from primary health care activities, etc?
Dr Heywood: I agree with Dr Maine, who has outlined what we should be addressing. We have to address whether the current practices are relevant to our aims. Groups like this have met for the past 30 y in a context of expanding economies and increasing funding. What they did was just add to what was already being done without removing anything; they did not have to consider choosing among various things. This meeting may also be avoiding that issue. You cannot do that any more because, for example, Indonesia has a negative growth rate of 25% this year. Next year it will be probably 5%, and it will not start to expand. It will probably not even get back to its current size for another decade, which means that there will be less money. The international donors will be gone very soon because there will be another crisis somewhere else. Thailand is not as markedly affected as Indonesia but it is nevertheless affected, and it will be several years before its economy starts to grow again.
Indonesia cut its health budget this year in real terms. Where should its health money be spent? Should some of it be allocated to preventing maternal mortality? It is very important for us to make sure that we evaluate the current dogma and what is being done at the moment to guide governments and policy. There are going to be many national economies that are not going to grow very much in the next 5 y.
Dr Fleming: Regarding the question of the contribution of severe (and moderate) anemia to maternal mortality, one has to go back to a time before blood transfusions. At that time, only 20% of pregnant women globally had any contact with the formal health care sector. We do not know how much anemia and mortality is prevalent today, but my guess would be that the prevalence is not very different from what it used to be.
The maternal mortality we see today in anemic women who come for care falls into 3 groups. In the first group are women from the pool of severe anemia whom we do not see and who die before transfusion can be started, which possibly represents 2% of all the severe anemia cases seen in the hospital. The second group comprises women who are anemic and have some overwhelming infection. The third group comprises those who are anemic and hemorrhaging. The Harrison figures, which analyzed 93 maternal deaths in women with hematocrit <27, found hemorrhage contributed to 33% of the latter deaths [Harrison KA, Rossiter CE. Maternal mortality. Br J Obstet Gynaecol 1985;92(suppl):100–15]. It is a figure that I know that Dr Rush will not accept. There is one intervention that prevents severe life-threatening anemia, and that is malaria prevention.
Dr Bothwell: High hemoglobin cannot be sidelined. It is a major point in Dr Rush's paper and he quoted, I think, figures from Pritchard and Hunt [Pritchard JA, Hunt CF. Surg Gynecol Obstet 1958;106:516–8], who showed iron supplementation leading to hemoglobin concentrations of 140 g/L. He inferred that these people may be at risk of high mortality. I think that is a very provocative statement and must be approached. The inference is that in trying to prevent the anemia you may be raising infant and maternal mortality.
Mr Alnwick: You have an important point.
Dr Johnston: This is an issue that needs to be resolved because there will continue to be iron supplementation trials. It should be fairly easy to answer this question.
Dr Bothwell: The point is to determine which questions are the most easily answered from the existing data.
Mr Alnwick: There is a difference here. Dr Johnston is saying that we ought to try to answer that question from the trials. Dr Bothwell is saying that it is easier to answer with existing data. Do we agree in fact that it is a provocative suggestion from Dr Rush that needs to be answered? If we do, there are 2 ways of answering it: one is from a new set of studies and the other is from existing data. Are there people who know the data well enough to agree with Dr Bothwell and say that ought to be done?
Dr Yip: I think this is an important point. Also, in response to the concern about severe anemia, we cannot have it both ways. If we can shift the distribution curve significantly upward by giving iron, more women would be in the so-called high-hemoglobin risk range. It also means fewer women would be in the low range, which is the very severe anemia range that is the dangerous one.
If we assume that iron therapy can shift the curve significantly upwards, we can establish unequivocally that in the very severe range, with hemoglobin at 30, 40, or 50 g/L, you get sick, lose more blood, and die. We do not need a trial to show that a hemoglobin concentration of 30 g/L kills. I could show you data for the percentage of potential maternal mortality that can be attributed to severe anemia. The point is that we have not been able to show effectiveness. The hemoglobin distribution curve can be shifted to the right to reduce severe anemia, but we have to decide whether we want to shift the curve. If it can be shifted, there will be a potential strategy for reducing severe anemia. If it cannot be shifted, this preventive strategy cannot work and there will be no need for concern about the high hemoglobin end. Either we have an intervention that works for both ends or for neither end. That is what I am trying to argue.
Dr Ladipo: From the clinical point of view, the main concern is severe anemia. The presentation from Africa and other comments show that anemia is multifactorial. Programmatically, to address anemia as a cause of mortality, we have to remember the major causes of anemia. The severe anemia itself could be chronic or acute and may be related to hemoglobinopathy. Whether an individual dies depends on how the individual has adjusted hemodynamically or what immediate services are available to assess the person, not transfusion per se. Some Jehovah's Witnesses still refuse blood transfusion and some have a progressive decline in hemoglobin in pregnancies to 50 g/L. They refuse blood and yet they survive pregnancy, labor, and the postpartum period, but that is not necessarily the ideal situation.
From the point of view of prevention of severe anemia we have no clues. Besides giving iron supplements, we have to ensure that the pregnant woman has treatment for major causes of anemia such as malarial infection and helminthic infestation that are quite common—it was about one-third for Malawi. In Nigeria, 40% of people have worms of various kinds. In clinical practice I have not seen high hemoglobin values in women who have been given prenatal iron tablets. There must be other reasons for high hemoglobin values. Iron per se on its own during pregnancy, either throughout the 9-mo period or in the last trimester, would not lead to high hemoglobin levels. I do not regard it as a problem.
Mr Alnwick: I am not sure that the earlier question from Dr van den Broek was answered. It was partly related to competition for resources and something Dr Heywood also referred to. I would have thought that a very reasonable consensus would be that one of the most effective things to be done to reduce maternal mortality is to strengthen essential obstetric care and district referral systems Wherever that is possible, it is being done. It is certainly a priority of WHO, UNICEF, and now the United Nations Population Fund. I am not sure that anybody in those agencies—and perhaps other agencies as well—sees waiting for district services to be strengthened as an alternative. I think the real issues that people are struggling with are in situations such as found in much of Tanzania, where a reasonably good health infrastructure has collapsed. Even with the best will in the world, people think it will take years to reestablish and retrain health workers. A second example can be found in countries such as Mozambique, where there never has been any health infrastructure and people are trying to develop one. Here, too, I believe it will be a significant number of years before we have a system of training and referral for the basic essential elements of obstetric care.
What people are asking themselves is whether in the interim there are large-scale public programs that could be put in place that would have a chance of contributing some element of prevention. Could those programs deliver something to women in the interim? Maybe they would need to continue, if they were effective, while the other programs are being put in place. I would not see it as an either-or situation.
To come back to the point that Dr Labbok raised, we are certainly not thinking only of iron and folate supplements. We might be thinking of multimicronutrient supplements because of what we heard about vitamin A and the interesting data from Dr Fawzi [Fawzi WW, Msamanga GI, Spiegelman D, et al. Randomised trial of effects of vitamin supplements on pregnancy outcomes and T cell counts in HIV-1-infected women in Tanzania. Lancet 1998;351:1477–82] that was mentioned yesterday. There are a growing number of population-based studies suggesting something is going on with micronutrient supplements and outcomes, possibly including maternal mortality. In circumstances where it will take a while to strengthen district facilities, I suggest there is also a need to consider whether there can be alternative mechanisms for getting other effective ways to prevent death from hemorrhage. For example, an alternative would be the use of oxytocins by people who normally would not be considered licensed or able to use such drugs if those are lifesaving measures.
Dr Labbok: Programmatically, we are talking realistically about what we can do. I think the concept of shifting the curve is much more feasible. The people who are already in good status probably will not change much, except those few people with a high hemoglobin problem. What we would get, if we treated more broadly, is a skewing and not the creation of problems. I suggest that the curve would be shifted to the right. I do not think we have to worry about the upper-tail problem being as broad as the savings in the reduction of anemia. However, I do not think the curve would definitely move if we were just dealing with iron deficiency anemia, because the bottom end of the tail is multicausal.
Dr Kahn: The hemoglobin curve is not a double-tailed Gaussian curve. It is really a single tail if you look at data of healthy populations: you come up to a stone wall and you may have some tailing. The front of the curve does not move much when it gets up to that wall.
Dr Caulfield: One of the good things about the new INACG guidelines is that they focus on interventions and prophylaxis for malaria with a specific focus on what to do in severe anemia and emphasize what we think we can do and combining approaches. People forget and focus on what we thought previously rather than on what is currently out there, although this is not the case for programs. The new recommendations include specific recommendations on what we should be doing and include more than iron supplementation.
Dr Rush previously stated that iron supplements could, in fact, cause women to have pathologic hemoglobin, meaning that they could have higher hemoglobin concentrations than they were perhaps intended to have. I think that most people believe that is not true. There is, however, the other issue—that women who have high hemoglobin concentrations seem to have poor pregnancy outcomes. I do not know that they are more likely to die during pregnancy. Certainly in observational studies they seem to be more likely to have poor pregnancy outcomes from the perspective of the fetus in terms of preterm delivery, perinatal death, and low birth weight. There is a difference between what we think about the mother's outcomes and the baby's outcomes.
Some women are going to have hemoglobin levels of 130 and 140 g/L. The issue is whether their biological hemoglobin, by itself, predisposes them to have poor outcomes such that one might want to prevent them from having their biological hemoglobin concentration. There is an argument in obstetrics that entering pregnancy with high hemoglobin levels—meaning 130 g/L—and being rich in oxygen may lessen plasma volume expansion. The question—which I think it is ludicrous—might be about whether those women should give blood to lower their hemoglobin concentration to below their physiologic value to have better pregnancy outcomes. You have to consider the whole topic extremely broadly and not only think about whether iron supplementation can cause a pathologically high hemoglobin, higher than somebody is intended to have. We also need to think about what it means to enter pregnancy with a physiologically appropriate hemoglobin concentration.
Dr van den Broek: We are making a number of mistakes and making lots of assumptions about whether the curve will or will not shift. On the whole, in developing countries fewer women have the problem of high hemoglobin than have the problem of low hemoglobin, although we should look at it. Can we not think about a multinutrient supplement to try to shift the curve and then carefully observe what happens at both ends in the shift of this curve?
I also wanted to comment that there is a lack of papers on obstetric outcomes, not perhaps on maternal mortality but on premature labor and low birth weight, complicated by issues of plasma volume, which also needs to be studied. However, such studies would be of individual patients in a hospital setting rather than population supplementation.
Finally, can we look for something feasible to do? For example, if I were to design another randomized controlled trial for an intervention in pregnancy I would look at the entire hemoglobin curve and try to observe both ends so that we could study what happens to high hemoglobin levels. We do not have evidence to suggest that we should only be supplementing women who are severely or moderately anemic. Supplement everyone and observe what happens closely with whatever outcome measures you want, and study the individual woman's physiology because that is the question. Why are women physiologically anemic? Maybe it is not such a good idea to have 130 g/L if you started with 120 g/L. Maybe if you started with 130 g/L you should go to 140 g/L, but that needs to be studied in individual patients.
Dr Maine: When Dr Bothwell said that the question of toxicity could be answered with existing data, I was not clear about whether this meant re-analysis of existing data sets or that Dr Bothwell thought that published studies exist that answer this question sufficiently. There is a difference, and we need to be careful that we do not conflate deaths from hemorrhage with those from anemia. Most of the women who die of hemorrhage may not be anemic. If I had a postpartum hemorrhage and no treatment, I would die. It does not mean I am anemic now.
Another thing relates to what happened with traditional birth attendants—which went on for decades and drained away all resources and attention from getting obstetric services functioning. I think there are cases when we have to admit that right now we probably cannot do much about maternal mortality for, for example, a whole country. We know we can do something in an area where we can get the district hospital functioning, but for the whole country, let us do what we can do. We may be able to do something for kids. We may be able to do something for onchocerciasis or some other terrible scourge of humanity, but I think to mislead people into thinking there is something you can do in a major way for maternal mortality will only delay true progress. There are lots of important things to do in the world.