| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
EDITORIALS |
1 From the Center for the Study of Sensory Impairment, Aging, and Metabolism (CeSSIAM), Guatemala City (NWS), and the Technical University, Munich, Germany (KS).
2 Reprints not available. Address correspondence to NW Solomons, Center for the Study of Sensory Impairment, Aging, and Metabolism (CeSSIAM), PO Box 02-5339, Section 3163/Guatemala, Miami, FL 33102-5339.
See corresponding article on page 116.
INTRODUCTION
In this issue of the Journal, Sharma et al (1) present the results of a partially randomized efficacy field trial comparing 100 daily oral doses of iron with 3 intramuscular iron dextran injections given at 1-mo intervals for the alleviation of pregnancy anemia in a public hospital setting in New Delhi.
ETHICS OF RESEARCH IN UNDERPRIVILEGED AND ILLITERATE SUBJECTS
Issues of biomedical ethics are major considerations in any investigation involving human subjects (2), even in the publication of research results (3). Biomedical ethics derives from a series of principles, among which the most important are those of beneficence, justice, nonmaleficence, and autonomy. An investigator from the prestigious Indian Medical Research Council participated in the research done by Sharma et al, and the study was reviewed (and approved) by the institutional review board of the host medical school. In the study by Sharma et al, the principle of beneficence was adhered to, because the subjects were all iron-deficient anemic women who had a need for additional iron and who received an iron-containing treatment. In terms of justice, there was no evidence in the study of discrimination or arbitrary exclusions of persons in need.
However, serious problems arise regarding the principle of nonmaleficence, which derives from the Hippocratic principle of primum non nocere (first do no harm). As can be substantiated by the literature (4-6) and illustrated in the list of side effects given in Table 5 of the study, parenteral exposure to iron dextran imposes a series of serious risks to the health of the recipient. Among a sample of only 100 women in the intramuscular iron group, the documented frequency of local pain, skin staining, fever, and systemic symptoms—not to mention adverse reactions requiring hospitalization—is unacceptable.
As cited by Sharma et al, 3 (0.14%) life-threatening, immediate reactions were observed in a previous study in which 2009 subjects received intravenous iron dextran injections (4). The same study also included 8 (0.38%) instances of severe delayed reactions including arthritis, myalgia, fever, and thrombophlebitis with pulmonary embolus that required observation and treatment for several days (4). There is no corresponding large-scale study on adverse and near-fatal events with intramuscular administration of iron dextran. Several case reports, however, showed that anaphylaxis may be encountered in such cases and may be fatal (eg, see reference 5). In the era before the availability of disposable syringes, a considerable percentage of intramuscular iron injections in Indian hospitals were associated with abscesses (6). Hence, there seems to be no comparability of risk between the 2 treatments in the study by Sharma et al; one treatment (parenteral administration of iron) is clearly more dangerous and life threatening than is the other (oral administration of iron).
This brings us to the principle of autonomy, which dictates respect for individual free will based on informed choice. This takes on special dimensions when there is an inherent inequality of power between the investigators and the candidates for enrollment in a study (7). Given the impoverished and illiterate nature of the study's target population from the slums of New Delhi, fully and completely "informing" these subjects of the nature, purpose, benefits, and risks of the proposed intramuscular exposure to iron dextran was virtually impossible.
The investigators and their ethical committee obviously recognized ethical dilemmas because an initial proposal to enroll only women with severe anemia was considered and rejected. This proposal was disapproved because it was considered dangerous to leave such women at the mercy of oral iron treatment only; the gravity of the most severe anemia justifies the risk of parenteral iron administration. Exposing women with only moderate anemia to intramuscular iron treatment, however, was to err in the other direction.
PROGRAMMATIC WISDOM OF RECOMMENDING INTRAMUSCULAR ADMINISTRATION OF IRON
Parenteral administration of iron dextran is an uncomfortable, inconvenient, and dangerous procedure (4-6, 8). These disadvantages are equally present in public health settings, clinical settings, and applied intervention research. In fact, "routine" usage is likely to be more dangerous because the quality-control features of efficacy research are necessarily relaxed in routine usage, with lower standards of training and fewer financial resources. An example of this can be found in the study itself, in which Sharma et al point out that the 36 women who were rejected for entry into the study because of intolerance to oral iron and who returned to the routine prenatal care clinic of the hospital "were given a prescription for orally administered iron according to the hospital's protocol."
The differential economics of the 2 intervention options, however, should represent further impediments and inhibitions to adopting parenteral iron treatment. By the authors' own reckoning, the base costs of the inputs for oral iron administration are one-eighth of those needed for intramuscular administration. Although the costs of both treatments are very low by the standards of developed countries, in a poor nation of 1 billion inhabitants, of whom one-half are female, an 8-fold difference in price has important fiscal implications, even before excess hospitalization and lost productivity are taken into account. Moreover, the corrective measures necessary for intolerance to orally administered iron require only stopping the administration of the tablets, whereas complications of intramuscularly administered iron may require hospitalization and medications.
BIOLOGICAL IMPLICATIONS OF COMPARATIVE FINDINGS ON HEMATOLOGIC STATUS
The biological implications of the findings of a comparative assessment of 750 mg Fe injected into the gluteal muscles of women and 10 g orally ingested Fe are potentially significant. Although Sharma et al assumed 10% fractional absorption of iron from 100-mg tablets, we estimate that 7.5% of orally administered iron is absorbed on average. No difference in the rates of anemia were observed between the treatment groups, the mean increase in hemoglobin concentration in each group was 7.3%, and the initial anemia was corrected in slightly less than one-third of the subjects in each group. More than 70% of the subjects receiving either treatment remained anemic. This "modest" hematologic response suggests impaired utilization of the administered iron perhaps because of the general nutritional status of the subjects. Semiquantitative modeling of total body and compartmental iron exchanges in Indian pregnant women might help to illustrate the possibilities and limitations of even 750 mg internalized Fe for preventing or correcting gestational anemia. A typical woman of reproductive age from New Delhi might weigh 50 kg before conception and have a circulating blood volume of 3.5 L (7% of body weight). At the time of delivery, when blood volume would have expanded by 35% to 4.7 L, a nonanemic (hemoglobin concentration of 110 g/L) woman would require 1760 mg Fe within the circulating red blood cells (3.4 mg Fe/g hemoglobin in a total of 
520 g hemoglobin). During pregnancy, however, iron is needed not only for new red blood cells but also for the fetus and placenta (360 mg) (9), and an additional 230 mg is needed for the 0.8 mg daily endogenous Fe lost over 280 d of gestation; thus, a total of 590 mg is needed. If a woman were to depend on only the 750 mg Fe taken in from the orally or intramuscularly administered supplements, then only 160 mg of this iron would be left over for erythropoiesis, and she would inevitably end her pregnancy with a hemoglobin concentration that was lower than that with which she started her pregnancy. A total of 1100 mg Fe is needed to account for the basal losses, products of conception, and new red blood cells in the expanded blood volume of an initially nonanemic woman who remains nonanemic through term. If the woman were to enter pregnancy with a hemoglobin concentration at the lower limit of the entry criterion for the present study, ie, 80 g/L, her cumulative needs to support herself, her pregnancy, and her fetus plus the total demand for erythrocytes would be 1400 mg. Thus, the 750 mg Fe from either supplementation format would supply only about one-half of the demand for new iron. In the study by Sharma et al, somewhat less than one-third of the women recovered from the anemic state during the course of pregnancy. In addition to receiving the 750 mg from the supplements, those women would have had to absorb 
1.8 mg Fe from dietary sources on each day of pregnancy.
With respect to the effect of iron administration on iron reserves, the interpretation by Sharma et al of ferritin concentrations as biomarkers for iron storage is problematic. In theory (10), the body uses newly received iron to replenish red blood cell mass before committing the nutrient to storage reserves. Because anemia was not corrected in most of the women who participated in the study, it would be premature to look for a buildup of iron reserves. Moreover, these low-income women are susceptible to inflammatory stress, which could increase ferritin concentrations in a manner unrelated to any increase in storage iron (11). Women of the underclasses of New Delhi often live in squalid conditions in which immunostimulation from microbes in the unsanitary environment (12) is a continuous condition. In addition, as described by Sharma et al, tetanus toxoid boosters are part of prenatal care in India, and inoculations produce a systemic immune response. The intramuscular injection of iron itself could have been a nidus for local irritation and inflammation in the women in the intramuscular iron group. Conversely, daily consumption of 100 mg elemental Fe by the oral iron group may have produced irritation in the alimentary tract and portal zone with attendant systemic inflammation (13). Hence, without the benefit of monitoring the acute phase response status of the women with a circulating biomarker such as C-reactive protein or 
1-glycoprotein, clear delineation of iron reserves from ferritin responses is impossible to achieve.
Well over a gram of new iron would probably be needed to satisfy the needs in the various expanding compartments in an already anemic pregnant woman to bring her to full sufficiency and out of anemia. Although both oral and intramuscular routes appear to be equally effective, indigent, pregnant Indian women need more iron than can reasonably be forced through gut or muscle to restore them to the nonanemic state. The challenge of treating iron-depleted and anemic women is unlikely to be met safely or comfortably with either parenteral or enteral iron treatment if treatment begins after the conception of pregnancy. The real essence of public health interpretation of the findings of Sharma et al is to shift the focus from the gestational period to the prepregnancy or interpregnancy period and from the maternity clinic to the community (14). Building women's iron reserves to 1 g Fe would need to be achieved through some concerted prophylactic programs of targeted iron fortification or periodic supplementation of the entire population of women of reproductive age.
LESSONS LEARNED
Because of ethical considerations for human subjects in research and the inherent lack of safety of intramuscular iron administration for wide-scale application in public health systems, the research hypothesis of Sharma et al probably never should have been pursued. Furthermore, although this finding is counterintuitive and contrary to most prognostications, parenteral administration of iron did not alleviate moderate gestational anemia any better than did enteral administration. The Indian Health Ministry should be disabused of any interest in using iron dextran as an alternative to oral iron treatment, given the former's higher cost, greater inconvenience and discomfort, and, above all, more severe toxic risk. Moreover, we can glean lessons from this experience about the pitfalls of designing investigative procedures to be conducted in underpriviledged and unempowered populations.
REFERENCES
Related articles in AJCN:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
