点击显示 收起
GROUP 1—DRS GUIDOTTI (CHAIRMAN), JACOBS, JOHNSTON, KAHN, LADIPO, RUSH, WINICHAGOON, AND YIP
Dr Johnston: Dr Guidotti introduced the topic by identifying the World Health Organization's (WHO's) concerns about what to do, what instructions to give, and what guidance to provide to frontline community workers about the proper entities of care. We then focused on what we can do to reduce maternal mortality in relation to severe anemia. Some of us agreed that there is a known risk of increased maternal mortality with severe anemia and that this question does not need to be researched. Others reiterated that it has not been demonstrated that reversing severe anemia lowers the associated risk of maternal mortality. We also recognized that there are several causes of severe anemia.
The first questions were, Can we identify severely anemic pregnant women? Can we use the diagnostic tools or do we need to identify individual cases? What interventions should be considered? Can we reverse severe anemia using these interventions? Two untested strategies were suggested—a package of interventions and improved service delivery. A third strategy might be a combination of both. Can these strategies be tested on a community basis or do they need to be targeted? Our fundamental question was, How can we prevent severe anemia?
Another question was, What proportion of pregnant women have severe anemia? Can we identify these women with severe anemia at the community level? If so, do we need a lab to do this or are there community-based methods that can be used or developed, such as one centrifuge for the community to test for hematocrit? We agreed that detection at a community level is not a research question, particularly by relatively reliable and specific clinical screening. We came back to the question about individual screening and targeting. Can we use an epidemiologic basis to intervene for whole communities? Epidemiologically, we can look at the nutrition problem in the community, the infection prevalence, and genetic causes. We can get that general picture of the community. Can we rely on that or do we need to do individual testing and screening? If we use epidemiologic data, do we have the appropriate tools to look at all causes of anemia and full baseline data or do we need a rapid tool for parasites, malaria, and nutrition? Those were questions that were raised but not answered: we kept coming back to our primary research question, which concerned the options for intervention to reduce the severe anemia in pregnancy.
The 2 main interventions were, first, targeted identification and screening women for severe anemia and, second, treating them individually based on the evidence of the community epidemiology. The women would receive a packet of anthelminthic, antimalarial, and nutritional advice depending on the epidemiology of these in the community. The comparison group would comprise all pregnant women in the area and would get the intervention. These women would receive whatever was the common primary health care practice and a packet of anthelminthic, antimalarial, and nutritional advice, again depending on the epidemiology of anemia in the area. The primary outcome measure would be a change in prevalence of severe anemia at the end of pregnancy. Secondary outcomes would be complications at delivery, maternal mortality if the sample was sufficiently large, and birth outcomes including birth weight if obtainable. There should also be a control area. There were ethical questions concerning identifying women in the control area with anemia who did not receive treatment. Any woman identified with anemia would require treatment for ethical reasons, but this control group would be needed.
Hemoglobin would be monitored periodically throughout the intervention. Height was of questionable use, but it would be important to identify women who thought they were going to have a large baby. Services or some pattern of care would need to be provided for women who were concerned about the size of the baby. We did not decide whether the research should be prospective or cross sectional. If it was prospective, you probably would not get very many severely anemic women because you would have treated them before they became severely anemic, which might favor a cross-sectional design.
The cost-effectiveness of the 2 interventions would need to be monitored. Outcomes have already been mentioned. The provision of obstetric services should be equivalent for all the groups and probably sufficient to deal with any serious problems.
Dr Martorell: Why do you need a control group if your target is the current practice?
Dr Johnston: We would be testing a new package that includes multiple micronutrients and antimalarial advice. The control areas would get the current practice.
Dr van den Broek: I am not clear about the groups. For the targeted intervention, are you going to screen out women with severe anemia, do a diagnostic protocol on them, and treat according to what is wrong with them?
Dr Johnston: There will be no diagnostic protocol, only treatment that is based on what we know about anemia in the area. If the incidence of malaria is high, we will give them malaria prophylaxis.
Dr van den Broek: Is that the same package you are giving in the untargeted second group but you will give it to everybody?
Dr Johnston: Yes.
Dr Rush: We would monitor hemoglobin concentrations in the targeted group. Those who did not respond would be referred to the next level of medical care for further evaluation and modification of treatment.
Dr van den Broek: For those with moderate anemia, you are offering whatever is currently being offered.
Dr Davidson: Would the research be done in diverse settings, different ecologic situations, where you knew there was parasitic infestation of different sorts and used the appropriate package?
Dr Johnston: Yes. It assumes that we know the causes of anemia: nutrition, infection, and genetics. It provides treatment for all of those known causes in a region plus a nutrition package, which could be multiple micronutrients, nutrition education, or both. It would be completely different in each place.
Dr Fleming: Where you do not know the causes of anemia, are you going to have etiologic studies before? How would you do it?
Dr Johnston: If we did not have that information, we would have to get it through a survey.
Dr Heywood: Is this all done through the public system?
Dr Johnston: Yes.
Dr Rush: The assumption is that these are areas in which village health workers are already in place. The effort would fail unless we ensure that there is a mechanism for referral to the next level of medical care.
Dr Martorell: Was there discussion of the ethical issues, because under current practice you would have to deny treatment to women with moderate levels of anemia?
Dr Rush: No, there is no need to deny routine supplementation, or any other intervention currently in place.
Dr Martorell: You are going to screen for severe anemia, but I do not know what screening method you are going to have. Presumably you are going to know the anemia level.
Dr Rush: Yes. I envision a clinical screen with either a hematocrit or hemoglobin concentration for both verification and monitoring. We would give this enhanced package to those who are severely anemic; moderately anemic women would receive current standard care.
Dr van den Broek: Did you discuss what the new package would be?
Dr Yip: We said that there are 3 major causes of anemia or contributors to anemia: nutrition, infection, and genetic factors. There is nothing we can do about the genetic factors. Infections include parasitic diseases such as malaria, hookworms, acute bacterial infections, and chronic infections. In many areas where severe anemia is relatively common, we would assume that the general intake of multiple nutrients is going to be deficient as a result of a very poor diet. The package would probably consist of iron and folate with vitamin A, riboflavin, vitamin B-12, and zinc as a multisupplement. We are willing to err on the side of too much nutrients rather than too little, as this would be the nutrition package. Ecologic information will be needed, for example, about what type of infection-based intervention should be added to this package.
Dr van den Broek: What about the group with chronic inflammation and chronic anemia?
Dr Rush: It would be very difficult to interpret the results of such a trial if rates of HIV infection in the test area were high. The vexing problem of HIV infection and anemia would have to be addressed separately.
Dr Yip: That could be an important research question. If there is severe anemia or significant anemia with evidence of infection and inflammation, but you have no easy way of identifying the sources of infection by conventional diagnostic microbiological means, you can give antibiotics to a subset of women in the trial and see what happens to the anemia, say 3 wk later, compared with a subset of women who did not get antibiotics. It is an operational question.
Dr Fleming: If sickle cell and hemoglobin sickle cell are identified, for example, they indicate a severe anemia in West Africa. Those individuals need to be helped in different ways. At what point do you identify them?
Dr Rush: After screening, the level on anemia would have to be verified independently, by some laboratory procedures. After a trial of therapy, nonresponders would be referred for further evaluation. At that stage, the etiology of the anemia could be specified.
Dr Belizan: In addition to division, you are thinking of a cluster randomization—clusters and randomized clusters.
GROUP 2— MR ALNWICK, DR BELIZAN, DR CAULFIELD, MS ELDER, DR FREIRE, DR HEYWOOD, DR KITANGE, DR LABBOK (CHAIRMAN), DR LOUDON, DR MAINE, AND DR MARTORELL
Dr Caulfield: We limited our discussion to iron, anemia, and maternal mortality. We identified what we thought were important questions and considered 2 topics specifically. The first was the severity of anemia and whether there is a relationship between mild, moderate, and severe anemia and maternal mortality and the need to focus explicitly on that within observational studies. We should disaggregate or examine hemoglobin as a continuum. The second was the issue of toxicity and whether one could generate harm through iron supplementation. The result of this was a consensus, at some level, that regardless of whether there might be a need to do research, there was no need to change the program or programmatic perspective at this time.
We recognized 3 types of research that might be conducted that would compress the array of questions we had generated. The one that received the greatest priority is an intervention study that could address different areas of research related to the need for iron and other nutrients in some type of supplement or another approach. It would be a package of interventions. It would be important not only to understand the effectiveness of these interventions in preventing or reducing severe anemia, but also the cost-effectiveness of this package of interventions. We discussed the need to understand these issues and the cost vis-a-vis the package and the individual components. There was some disagreement on the need, for policy purposes, to describe a package or describe individual components.
We thought about whether we would confine ourselves to the construct of the physiology of pregnancy or work also with the nonpregnant state. We would focus on looking at a continuum of hemoglobin or a continuum of anemia and not try to screen and treat. Also, perhaps some other issues, such as weekly versus daily regimes, could be incorporated into this type of intervention study.
We called the second study a laboratory study—laboratory in its broadest sense. The idea was that it would be an observational or association study, which would allow us to look at the associations between hemoglobin concentrations over a continuum and maternal mortality within communities or communities over time. The more proximate causes of maternal mortality would be looked at to understand how the relations exist and how they change. We used the example of the maternal mortality registry, etc, from Tanzania, which would provide data that would allow us to look more carefully at the relationship between various hemoglobin concentrations and maternal mortality within an association framework.
There is a need for further discussion on how other aspects of maternal nutrition, beyond anemia, may be related to maternal mortality. The other issue is to go beyond maternal mortality. We need to focus more on research and policy efforts to understand quality-of-life issues for women and how women may benefit from nutritional and other interventions.
Dr Labbok: I think we still have to continue to look for proxies for maternal mortality.
Dr Martorell: Although we did not flesh it out as a research issue, we did talk about the added value of having a surveillance component and focusing additional interventions and attention to severely anemic women within programmatic settings.
GROUP 3—DR CSETE (CHAIRMAN), DR BEARD, DR BOTHWELL, DR FLEMING, MS AOMARI, DR SCHOLL, AND DR VAN DEN BROEK
Dr Csete: We decided that besides existing supplementation programs, which, as I will say more explicitly later, should be continued and improved, it would be possible to give more of a push and more strength to doing more about severe anemia. That implicitly means we think, if resources are being diverted away from dealing with severe anemia, this is a problem that could be corrected without abolishing existing supplementation programs. There would be several interventions, all of which have already been mentioned, which would be part of any programmatic effort to deal more systematically and better with prevention of severe anemia. We all agreed that controlling severe anemia is an essential part of maternal mortality reduction.
We talked a lot about nutrients other than iron and folic acid, about malaria in endemic areas being a concern very directly related to maternal mortality reduction. Among the program issues we noted was the obvious need to give attention to toxicity or safety issues if we are going to start building nutrients onto iron-folate supplementation. In that regard, we noted that WHO recently made a pronouncement on the appropriate levels for vitamin A in pregnancy and lactation.
For treatment of severe anemia, we talked about several program and research issues. Among them was the operational matter of which diagnostic tool might be best. I was delighted to hear about the experience from Malawi of using the new color chart that we heard so much about. This is something that really needs to be worked out for us to strengthen programming in severe anemia. This group felt strongly that there should be studies, probably one-time studies, with a rigorous protocol for assessing the etiology of severe anemia in a given location. We thought that not only should that be done, but that it would probably need some advocacy work to convince policy makers, program planners, and others that this is an exercise worth doing.
At the first contact with the health services there should be generalized identification of severe anemia in pregnant or nonpregnant women and not only of those who seem at risk. Appropriate protocols for treatment and referral should be developed. It would be important to emphasize this point, especially as we try to make bigger program pushes in the early stages in this area. It would be important, also, to be sure that follow-up and evaluation of these activities are built into whatever we do. We started talking a little bit about including cost-effectiveness in protocols, but it may not be feasible or cost-effective in every location. There may need to be some effort at adaptation to conditions focusing on the essential feasible actions.
Among the research questions we discussed was the concern that where malaria and HIV co-occur significantly, perhaps the protocols for dealing with severe anemia or malaria prophylaxis might not be the same. What can we say about the importance of infections, other than malaria and HIV, to severe anemia? What could we expect if vitamin A and zinc were built into a supplementation program? Would that make a contribution to the treatment or particularly the prevention of severe anemia? There was, also, the interesting question of, Why is there so much anemia with HIV but people do not show the symptoms of full-blown acquired immune deficiency syndrome; what is the mechanism there? I do not know if that is as directly programmatically relevant as some other questions.
On the subject of supplementation, we went through the things that have been said before in this meeting. The efficaciousness of supplementation for addressing iron deficiency anemia is obviously affected by all these factors at the field level. We want to make the point very explicitly that we think the supplementation programs are important to continue and improve for reasons other than the relationship to maternal mortality reduction. The point was made, importantly, that well-supervised and targeted supplementation can help promote the growth of young women who are pregnant but who have not completed their physical growth. We also thought that supplementation should not be something that is automatically just associated with pregnancy, that the accent needs to be put on improving the nutrition of women and adolescent girls before, after, and during pregnancy. The results for growth of pregnant girls seen in the Zaria studies [Harrison KA. Child-bearing, health and social priorities: a survey of 22774 consecutive hospital births in Zaria, Northern Nigeria. Br J Obstet Gynaecol 1985;92(suppl):1–119] that Dr Fleming mentioned seemed to be very interesting. We thought that this was work that should be looked at and replicated, if possible.
Determining the etiology of anemia seemed to us to be something that we should be thinking about in all of our research and efficacy efforts. We took seriously the question of the safety of supplementation. The group considered that in public health terms this is not as much of an issue, but there are aspects of the question that could be looked at through reanalysis of existing data from studies about which we have heard. It might also be looked at more systematically by following the outcomes of women with high hematocrit or high hemoglobin in existing clinical settings or programs at various times. We can look at maternal mortality rates indirectly. One might look at such indicators as pregnancy-induced hypertension and premature labor and so on.
Dr Beard thinks there probably is a group of people who are iron deficient who respond to iron, or folate, or vitamin B-12, or whatever it might be, in a certain way. He also thought the resulting shift to the right in the distribution curve, which we were considering earlier, would not be very large.
Dr Heywood: I am unclear as to whether you are describing a general health service or a research activity.
Dr Csete: We did not conceive of this as a research trial. We conceived of it more generally and programmatically.
Dr Heywood: What is the cost likely to be of screening every woman on first contact with health services?
Dr Fleming: For hemoglobin, using the WHO hemoglobin color comparator method, the cost is negligible.
Dr Csete: We did not talk about universal screening. We talked about identifying cases of severe anemia, which may not be the same thing.
Dr Fleming: What we would like to see is more or less universal use of this color comparator. The cost in material terms is just a small piece of blotting paper.
Dr Heywood: I agree in material terms, but is it not going to end potentially in a situation similar to health center microscopy, which universally is done badly and is misleading?
Dr van den Broek: The field testing of this scale in antenatal rural clinics by local midwives who are not particularly highly trained suggests it is more accurate than looking at the conjunctiva that was totally inaccurate in our setting. Health care workers like the new scale. I think there are problems, but it is encouraging. Apart from the cost of the scale, I think it is also a way to raise awareness of the problem of anemia. Health workers like having something to measure and to write down. I think time will tell which way it is going to go. It has potential to work.
Participant: Are they testing the color scale?
Dr Fleming: The field testing is going on in Zambia, Malawi, and South Africa and a study in South Africa was recently completed.
Dr Guidotti: The multicenter trial has been completed and includes not only African countries but Asian countries. The results have been analyzed, and last week a group sat down to discuss them. It looks like it is a good test for screening blood donors [Maternal Health and Safe Motherhood Programme, Division of Family Health, Maternal and Child Health Division, WHO, unpublished report, 1999]. For blood donors the cutoff point is 120 g/L, and that seems to give us quite good results. We are still not sure about screening in antenatal care—whether you can screen below 80 g/L or below 70 g/L. We are going to reanalyze the data, looking at different possibilities, and hopefully will have the results.
Dr Rush: Can you just tell us something about the screening test?
Dr Guidotti: Are you all familiar with the old Talqvist method, which is a blotting paper technique, where you compare the shade of the blood on the blotting paper with a scale? We looked at that and asked, Can we improve this? A WHO researcher decided to take it upon himself to make this a better test. He took blood samples and diluted them so that he had a range including 120, 100, 80, 60, and 40 g/L. He fiddled until he got the blood scale comparable to the type of blotting paper against which he would compare them.
The cost is in printing the booklet using computers and getting the colors exact. He found the best blotting paper was Whatman, which is chromatography paper produced everywhere. You cannot use normal paper—you have to use this special blotting paper because it ensures that the blood is evenly disbursed and it dries very quickly. That seems to be very important in making this comparison. We do not like to use the word Talqvist because that method has a very bad reputation throughout the world. Most of the people I ask as to whether they use the Talqvist in their program say that it is not allowed in their region and feel very strongly about not using it. So, although we do not like to call it a Talqvist method, it is an improved Talqvist technique and for blood bank purposes below and above 120 g/L, it seems to have very good sensitivity and specificity. For use in antenatal care in general clinics we are not quite sure as yet.
Dr Rush: Our need for specificity is not as great as our need for sensitivity. It could be tragic to miss very many cases.
Dr van den Broek: I think the sensitivity for data for Malawi is 70–80% compared with that for conjunctivas, which is 30–40%.
Dr Labbok: I noticed that 2 of the groups concentrated on severe anemia and that our group specifically chose not to because we felt that there is a continuum. Often we use cutoff points, and there does seem to be an association, but obviously not as strong in moderate anemia. We need to look at the entire context of the women. Clearly, where there is a high level of severe anemia, you need to address severe anemia, but when there is more of a bell-shaped curve, what is the real contribution of this to the picture of maternal mortality? We have to look at the efforts on severe anemia only in the context of the whole maternal mortality issue, but I do not think that is necessarily the point of this meeting.
Dr Rush: That is sensible. Clearly, however, the weight of evidence is that death is associated with severe anemia, but no evidence of much, if any, impact from mild or moderate anemia. To consider all levels of anemia as important seems to mean an obligatory diversion of resources away from a subgroup of women who are at enormously high risk.
Dr Labbok: I was trying to say that we should eventually look at this issue as opposed to other issues in the causation of maternal mortality. From the data presented for the cutoff for severity, how we would measure severity and what it would take to measure severity in populations have certain costs. I raise this issue because 2 groups concentrated solely on severe anemia and we have to look at this in the context of all the causes of anemia.
Dr Rush: We contended with this in the first group through the design of our projected study. In one area, we would use screening and would give intense attention to severely anemic women. In the other area, we would distribute an extended (modified) package to all pregnant women. In the evaluation we would assess response across all levels of anemia. The paper we heard from Malawi yesterday [van den Broek NR, Letsky EA. Etiology of anemia in pregnancy in south Malawi. Am J Clin Nutr 2000;72(suppl):247S–56S] was extremely important. It demonstrated that the causes of severe anemia are more complex and different from those of mild and moderate anemia.
Dr Yip: If you are willing to consider pricking a woman's finger to get a drop of blood to put on the paper or a device, there is an alternative technology that will work in very basic settings, although most people do not like it. It is the copper sulfate test. You can set it at any level of hemoglobin you want. If you want to screen anybody who has hemoglobin <70 g/L, you can set the concentration at 70 g/L. If the blood does not sink, then that person has a hemoglobin <70 g/L. It is an all or none test. It is almost as little trouble as the filter paper comparison. You need to change the copper solution once per month.
Dr Nestel: I corresponded with people from South Africa and, in their experience, the difficulty is getting the copper sulfate solution at the right concentration. In fact, they could only get it from their central laboratory because unless it is made up in the right specific density you do not get the correct result.
Dr Yip: Somebody has to distribute it. It is not something you go to a local drugstore to buy. Somebody has to distribute the copper sulfate as a premix.
Dr Loudon: I do not think I heard anybody talk here about obstructed labor. The discussion on obstructed labor that we heard yesterday seemed to me to be awful. I think it probably deserves another meeting, but it did come up, and there was total confusion between cephalopelvic disproportion and prolonged labor for other reasons that will go down in the book as obstructed labor. I was really not aware of how obstructed labor was measured, whether cesarean section was used as a proxy for it. I do not know how important it is, but if it comes down to attributable causes of death in obstructed labor, I feel that we want very much better data than I heard yesterday. I would just like to hear other people's opinion on that. Maybe the data are okay.
Dr Johnston: It would be good to know how to measure obstructed labor reliably because if we are going to have interventions that bring infants from low birth weight to normal birth weight, and if some women are small, we need to reassure ourselves that a problem does not exist. We need to know how to measure it objectively.
Dr Ladipo: Obstructed labor is an important topic vis-a-vis the issue of maternal mortality. I quite agree that sometimes people may misinterpret prolonged labor as obstructed labor, but I think in a good setting that error must be very minimal because there are clear features of obstructed labor that can be defined. Whether we can do anything about it now, I do not know. One issue was the role of nutrition and obstructed labor. We know that women of short stature— <1.5 m in height—tend to have a contracted pelvis. The shape of a pelvis may be gynecoid, but there is reduction in the diameters and often there is a mechanical problem in labor for such women. Whether nutritional intervention will result ultimately—maybe in 2 or 3 generations to come—in increased pelvic diameter is a big question. The stunted pelvic architecture has in the past been attributed to things such as rickets and osteomalacia. Perhaps studies in the future need to see what proportion of cases of obstructed labor can be attributed to nutritional deficiency in childhood or adolescence.
Dr Loudon: It was not quite clear to me in the work done on obstructed labor how obstructed labor is diagnosed.
Dr Caulfield: The Merchant work [Merchant K, unpublished final report submitted to Safe Motherhood Research Programme, May 1991] looks at cesarean delivery in a hospital setting of primigravidae, where other known causes of cesarean section have been ruled out, and there is a presumption that what is left is cephalopelvic disproportion. That is the population-based literature that we have on that issue, and it is a deficiency, absolutely.
Dr Labbok: I hope we all recognize that we are dealing with data collected with certain things in mind. The whole area of causality, not the proximate but the earlier causes of what we call causes of maternal mortality, is simply not studied. We have no idea, for example, whether certain vitamins might have an effect on the stages of labor. It simply has not been studied, yet it is feasible because all of this is modified through prostaglandins, coenzymes, and such where the vitamins would come into play. Your question is huge, Dr Loudon. It is not small because we have not even begun to scratch the surface of how nutritional elements might have an effect not as the primary cause but as a contributory cause of maternal mortality.
Mr Alnwick: Dr Loudon accepted there was a previous period in history when mortality decreased. Are we quite sure that maternal heights—adult heights—in the United Kingdom and most of those other countries were short in those days? Were women much shorter? Is it not reasonable to assume that some of the higher mortality back then in the mid 19th and 18th centuries may have been partly attributed to shorter stature?
Dr Loudon: I think not, because the commonest cause of cephalopelvic disproportion was a rickety (or rachitic) flat pelvis. Rickets was much more common after the Industrial Revolution in the late 18th century. Also, there should have been a higher prevalence in the 19th century than before. There is no doubt that rickets became more common particularly in the northern area in the late 18th, early 19th century. Against that you have smaller stature. There was smaller stature in the 17th and 18th centuries, but it was very much smaller than in the early 19th century. So, I would be doubtful about that. I think this is a can of worms, if ever there was one, about the size. I would put my money on the worst period for cephalopelvic disproportion as the mid to late 19th century for Europe.
Dr Rush: The rate of all cesarean sections is not a perfect surrogate for the rate of obstructed labor because it almost surely understates rather than overstates the problem. The other conditions that lead to cesarean section, such as toxemia and obstetric emergencies, tend to be associated with undergrown babies. There is a terrible paucity of data on this issue. The rate of intrapartum cesarean sections is a very good index, and that is what Merchant used in her urban Guatemala study.
Dr Belizan: I used to work in a hospital and now I am working in a very poor community, where I learned a term I had never heard before: abandoned transverse lie. I want to talk about the question of obstructed labor in poor rural areas. A woman starts contractions, the midwife comes and stays, they measure the time, let us say, by the moon or the sun. They say, "Okay, it's taking more time than expected," it is a 2-moon thing or whatever, and then they go to the hospital. They look for transportation to go to the hospital and that takes 2 more hours, then it takes 6 h to get the hospital. When they arrive at the hospital, the woman has been in labor for 2 d and may die. That is normal obstructed labor. The reason was a big fetus, transverse lie, breech, and malposition of the fetus. You may have malposition of the fetus more than a big fetus. This is considered obstructed labor in a hospital, where we can follow progress with vaginal examination or the position of the fetus. We can follow labor. That is my experience with obstructed labor.
Dr Labbok: That implies hemorrhage as the cause of death by the time they are done, by the time she has ruptured and bled.
Dr Loudon: My guess would be it is a transverse lie, which occurs with grand multiparity, a woman who had many babies. It would be rare to find a transverse lie in a first baby without some other pathologic reason. It may occur in association with large families. Would you agree there?
Dr Ladipo: I think you are quite correct, without other pathologies such as placenta previa or uterine deformity.
Mr Alnwick: I felt we left the calcium discussion a little loose yesterday. There are several fairly good meta-analyses on calcium [Carroli G, Duley L, Belizan JM, Villar J. Calcium supplementation during pregnancy: a systematic review of randomized controlled trials. Br J Obstet Gynaecol 1994;101:753–8; Bucher HC, Guyatt GH, Cook RJ, et al. Effect of calcium supplementation on pregnancy-induced hypertension and preeclampsia: a meta-analysis of randomized controlled trials. JAMA 1996;275:1113–7; Hamet P. The evaluation of the scientific evidence for a relationship between calcium and hypertension. J Nutr 1995;125(suppl):311S–400S]. It is a little worrying to have several studies that suggest that increased calcium intake, if it could be obtained, could reduce preeclampsia by a fairly substantial proportion, for example, the studies from Argentina and other Latin American countries. A very well done multicenter trial in the United States [Levine RJ, Hauth JC, Curet LB, et al. Trial of calcium to prevent preeclampsia. N Engl J Med 1997; 337:69–76], however, shows absolutely no difference. It seems a little unfortunate to abandon all interest in calcium because of a study in the United States in women who seem to have quite adequate calcium intakes. If preeclampsia or hypertensive disorders are such an important cause of maternal mortality, it sounds like this ought to be an open question. I had the opportunity to discuss this with the principal investigator of the multicenter trial in the United States, who thought it was still a researchable question in developing countries. His interpretation of his own data was that they did not answer the question. My recollection for Bangladesh, for example, is that calcium intakes are typically perhaps one-quarter of the intake levels of the low calcium intake group in the United States.
Dr Guidotti: Jose Villar did another meta-analysis, which was not published, that looked at populations with a low-calcium intake. He has reassured me that calcium makes a difference against eclampsia. My question programmatically is, Where are these populations that have low calcium? Is it all developing countries? I tried to get this information from the Food and Agriculture Organization of the United Nations, as it has a database on nutrition in all populations. Unfortunately, it does not have any information on calcium in different populations.
Dr Ladipo: That is a very important topic that we should consider. Eclampsia is still a major cause of maternal mortality in developing countries and virtually every country in the southern hemisphere will give you tales of bad cases coming in. Last week, I was told that the incidence at a South African college was 1 in 200, and that is about the same pattern in most other developing countries. Eclampsia is even seasonal. In Nigeria, for example, it occurs more in the rainy season than in the dry season [Wacker J, Schulz M, Fruhauf J, Chinwora FM, Solomayer E, Bastert G. Acta Obstet Gynecol Scand 1998;77:712–6; Jamelle RN. Eclampsia: is there a seasonal variation in incidence? J Obstet Gynaecol Res 1998;24:121–8; Bergstrom S, Povey G, Songane F, Ching C. Seasonal incidence of eclampsia and its relationship to meterological data in Mozambique. J Perinat Med 1992;20:153–8]. Is this related to any nutritional factor? Somebody mentioned briefly yesterday the role of antioxidants. I think one paper from South Africa showed vitamin C was the only antioxidant to be recorded as very low in most of the eclampsia patients [Bowen RS, Mars M, Chuturgoon AA, Dutton MF, Moodley J. The response of the dietary antioxidants vitamin E and vitamin C to oxidative stress in pre-eclampsia. J Obstet Gynecol 1998;18:9–13]. There are no data relating eclampsia to nutrition. A paper has been published showing that the use of 1.2 g prophylactic calcium prenatally reduces the risk of pregnancy-induced hypertension [Bassaw BS, Roopnarinesingh A, Roopnarinesingh H. Homer prevention of hypertensive disorders of pregnancy. J Obstet Gynaecol 1998;18: 123–6]. We need to follow this up and perhaps look at the other determinants, the antioxidant versus the oxidant determinants.
Dr Winichagoon: Most of the Asian countries (eg, China, Thailand, and Malaysia) probably have a lower calcium intake.
Dr Kahn: This results from high rice eating. Thailand has a high incidence of bladder stones that is due mainly to high phosphate and lower calcium.
Dr Winichagoon: That is a quite different story.
Dr Kahn: But it is the low calcium that was a contributing factor.
Dr Rush: Many, many nutrients have been implicated as causes of toxemia, but the basic pathophysiology remains obscure. I believe that before institutionalizing calcium supplementation in calcium-deprived parts of the world, another trial is going to be needed. The meta-analyses did include studies from the industrialized as well as developing worlds [Bucher HC, Guyatt GH, Cook RJ, et al. Effect of calcium supplementation on pregnancy-induced hypertension and preeclampsia: a meta-analysis of randomized controlled trials. JAMA 1996;275:1113–7]; the situation is very confused and confusing.
Mr Alnwick: I would not disagree with Dr Rush. We did not have the time to go into these other issues. It sounds to me that we have a 50-50 chance of doing something good with a relatively low-cost intervention such as calcium. I say that advisedly, because my understanding is that the cost of giving 0.5 or 1 g Ca daily is much, much greater than the cost of other micronutrients. So, it does not work out to be a particularly inexpensive intervention, unfortunately, but it is a relatively inexpensive intervention if it is a candidate for reducing a major cause of maternal mortality. We should be more enthusiastic about looking at it, and I would agree with Dr Rush that it is an opportunity to do the right kind of trial in the right kind of environment. Surely we should pick an environment where calcium intakes are very low. Bangladesh would seem to be a beautiful example. People get their calcium from fish bones.
Dr Loudon: If you put calcium in with everything else, and you get a reduction in deaths from eclampsia and toxemia, how do you know it is going to be the calcium and not something else?
Mr Alnwick: I was not necessarily saying that you had to put in everything. Your question is very interesting. If you put everything together, how do you know what causes the reduction in deaths? I am not sure we would need to know. If we had a supplement that reduced mortality by a decent chunk, why would we need to worry about what caused it?
Dr Heywood: Even if the cost is relatively low per person, that is not the figure that we need to know. It is the cost per large setting that we need to know and compare that with the cost per life saved from other interventions. The issue is not the marginal cost of the nutrient. The issue is the cost of that intervention per life saved, and I think you may well get a different answer if you do it that way. To give calcium to every pregnant woman throughout pregnancy, the cost might well be very high per life saved. This is because the relative rates for maternal mortality are low compared with some other mortality rates and preeclampsia does not contribute to most of those maternal deaths anyway.
Dr Martorell: We are not talking about effectiveness. We are trying to answer the efficacy question (ie, the scientific question).
Dr Heywood: That is not what I have heard others say. You were making that point.
Dr Martorell: Before we get into effectiveness we are addressing research needs, and it seems to me it is an efficacy question. We need to answer the scientific question first and then think about programs.
Dr Heywood: I agree.
Dr Martorell: But it is not the issue of, "Okay, if it is a 50-50 thing, let us do it." That is a different issue.
Dr Caulfield: In thinking about the calcium preeclampsia or maternal mortality issue, I do not understand the implication that there is a not a body of evidence that suggests that calcium supplementation could reduce pregnancy-induced hypertension from populations such as the 4 countries studied in Latin America [Carroli G, Duley L, Belizan JM, Villar J. Calcium supplementation during pregnancy: a systematic review of randomized controlled trials. Br J Obstet Gynaecol 1994;101:753–8; Bucher HC, Guyatt GH, Cook RJ, et al. Effect of calcium supplementation on pregnancy-induced hypertension and preeclampsia: a meta-analysis of randomized controlled trials. JAMA 1996;275: 1113–7]. There was also a large trial in a not particularly calcium-deficient population that showed no effect [Levine RH, Hauth JC, Curet LB, et al. Trial of calcium to prevent preeclampsia. N Engl J Med 1997;337:69–76]. Why is it that a study conducted in the United States that finds no effect can in fact require the need for further research that has already demonstrated efficacy? Why is there a need for further research in low-calcium-intake populations to put that back on the drawing board? Are there sufficient problems in the studies done in the developing countries, for example, that caused us to question the benefits demonstrated? I do not think so, but I have not reviewed that rigorously. When one is looking at the potential or the demonstrated efficacy of interventions, nothing has received greater scrutiny than this particular one. When you look at the results in the Cochrane database [Cochrane Collaboration, Adelaide, South Australia, 1998], or in some of the meta-analyses, the conclusion has been that there is efficacy demonstrated in low-calcium-intake populations, regardless of this one study. It really is not clear to me why there is a need for another large study to drive home that point. I do not see why the US study [Levine RH, Hauth JC, Curet LB, et al. Trial of calcium to prevent preeclampsia. N Engl J Med 1997;337:69–76] negates or reduces what we know from the previous work.
Dr Rush: Possibly because prior trials, on which the meta-analysis has been based, showed positive results in both industrialized and developing countries. There is a contradiction between the results of the new trial and some of the data on which the meta-analyses are based. There is no simple dichotomy between calcium-deprived Third World countries and the United States. Past studies in Western Europe and the United States had also shown positive results.
Dr Belizan: The main interpretation of the calcium meta-analysis is publication bias. Many big trials have tended not to be published and lots of small trials have been published. Whenever we had a study with a good sample size, they did not follow the pattern. What is the reason for this? Is it that we do not have negative results for the small sample studies? Is this research dealing with calcium deficiency? There are small trials with positive results, but no small trial with negative results has been published. All the bigger trials are negative trials that do not show anything. We can assume confidently that the biggest trial showing no effect is a trial on a population with high calcium intakes, and the smallest trials are on populations with low calcium intakes. I started 20 y ago observing the high calcium intakes of the indigenous populations in Guatemala. I have invested greatly in this hypothesis, but I would like to be sure be sure that there is an effect before any recommendations are made. We are going to do the trial. We have one population in Africa and we have a site in India, and WHO is going to fund these 2 trials. We need another site in Latin America with low calcium intakes. We know the intakes of calcium in Latin America are very, very low, between 200 and 400 mg/d in pregnant women.
Dr Johnston: Do we have the women's perception of taking the multiple micronutrient supplements, any qualitative understanding and their perception of anemia?
Dr Martorell: Yes, from Mexico where the public sector was offered iron and the private sector, through pharmacies, the multivitamin and mineral supplements. There is increasing pressure on the Mexican government to say, Why do you not offer the multivitamin and mineral supplements to the public sector?, but this information is anecdotal.
Dr Fleming: In Nigeria there is the tradition of taking traditional medicines in preparation for an event, whether it is an examination or something in the future. Providing antenatal supplementation in Nigeria was accepted as it was preparation for an event.
Participant: The research is coming together—being analyzed and written up—and in the different sites you get interesting parallels. For instance, in Malawi there is tremendous enthusiasm on the mothers' part at our study site to take iron, and they see it as their right. They want it. They demand it. They demand it postpartum and antenatally. It really depends on the setting. In that situation in Malawi, the issue has always been one of supply. We also have some really interesting material coming out of India. Our researcher in India, at one site, pointed out the issue of eating down in pregnancy that is related to our earlier discussions. Although it is very much a cultural tradition and a heritage as a belief among the women, they are quite ready to set that aside if they are convinced. There is room for behavior change if they are convinced that it is to their and their child's benefit. We are getting a lot of rich material coming out of the past 5 y of work that will be coming together shortly.