Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Chapter 3 ABORTION AND THE RISK OF MEDICAL COMPLICATIONS This chapter describes the medical complications associated with legal abortion in the United States and examines the recent trend in reported complications from illegal abortion. The chapter first discusses the difficulties encountered in defining medical complications and then presents data on complication rates from a national study carried out by The Population Council in 1970-1971. A brief review of the reports on long-term complications of legal abortion also is provided, largely based on foreign studies. Defining Medical Complications Data on medical complications associated with legal abortion are more difficult to assess and compare than are those on maternal mortality. Although the medical community might agree in a general way on a definition of "medical complication," in any given situation physicians' subjective judgments may vary widely as to whether a complication does or does not exist. Some bleeding is expected with every surgical procedure, and certainly with every birth; whether the blood loss is defined as a complication is generally left up to each physician. Furthermore, even if a complication of induced abortion is defined as "when a certain number of sutures must be taken in the cervix," two physicians may have different thresholds for deciding when to suture and when to apply other treatment. Consequently, what may be termed a complication by one physician may not be so regarded by another, and what may be interpreted as a normal occurrence in one instance may be judged as a minor complication in another. The time at which the complication occurs must also be considered. Guidelines for dealing with this factor have been provided by Christopher Tietze: Complications can be divided into two categoriesâearly, ... within ... 30 days after the abortion, and late, ... occurring more than 30 days after the procedure. Early complications may be further subdivided into immediate complications, defined as those occurring (although not necessarily discovered) during the procedure or within three hours after the procedure; and delayed complications, developing later, but still within 30 days.17 47
48 Using these guidelines, immediate complications for suction and D&C would include an adverse reaction to anesthesia, perforation of the uterus (with or without injury to the other organs), hemorrhage, and laceration of the cervix. Early detrimental effects from saline abortion would include possible damage to the central nervous system if the saline solution entered the blood stream, and disturbances of the blood clotting mechanism. The use of prostaglandins to induce second-trimester abortions was approved in 1973 by the Food and Drug Administration. Some preliminary information on the types of medical complications resulting from prosta- glandin use is now available. In general, these tend to be less serious than those from saline abortion, and include such minor complications as vomiting and diarrhea. There appears to be little or no damage to the central nervous system if the prostaglandins enter the blood stream, although there may be a temporary change in blood pressure. Labor generally tends to be of a shorter duration than with saline abortion.^/ The most frequent delayed complications from any induced abortion include retention of fragments of the placenta in the uterus, resulting in subsequent bleeding, various degrees of infection in the pelvic area, and thrombosis, which may lead to an embolism. All of these complications are relatively rare with legal abortion except for the retention of placental fragments in the uterus. It is often difficult to determine the extent to which any complications are actually abortion-related. Since pre-existing conditions may also contribute to post-surgical complications, it is often not correct to ascribe all post-abortal complications to the abortion procedure itself. Similarly, when such surgery as sterilization accompanies an abortion, it is difficult to isolate the complications arising from the surgery from those attributable solely to the abortion procedure. Complication Rates Associated with Legal Abortion The lack of a consistent definition of "complication" makes it difficult to compare data on abortion-related complications collected by different investigators from different localities. Consequently, most of the findings reported in this chapter are drawn from the Joint Program for the Study of Abortion (JPSA) which provides a fairly consistent data base for assessing the incidence of immediate and delayed abortion-related complications in the United States. Joint Program for the Study of Abortion (JPSA). The Joint Program for the Study of Abortion was a coordinated effort sponsored by The Population Council in 66 institutions (60 teaching hospitals and six abortion clinics) from July 1, 1970 to June 30, 1971,
49 for the purpose of collecting information on early medical complications associated with abortion.â¢3/ The study surveyed almost 73,000 abortions, or about 1/7 of those performed legally in the United States during that period. Thirty-four of the reporting institutions were in New York state and the rest in 12 other states.* The JPSA study computed complication rates for total patients and for local patients with follow-up. Follow-up was defined as any contact with the woman 10 or more days after the abortion. The complication rate for total patients is a minimum estimate of the incidence of complications because follow-up was not available for all patients. Because patients with complications are more likely to return for care than those without complaints, and since the proportion of non-private patients (who had more complications) was higher among local patients than among non-residents, the complication rate for local patients with follow-up approaches a maximum estimate of the incidence of complications. Table 8 summarizes the findings of the JPSA study for all abortion patients and for local patients with follow-up. Complication incidence is itemized according to gestation, abortion procedure, pre-existing complications** and concurrent sterilization. The complication rates (number of women with one or more complications per 100 women obtaining abortions) are further distributed in terms of both total and major complications. "Total" complications include all complaints registered by the patient, as well as diagnoses made by the physician. Minor problems, such as a single day of fever or vomiting, would be included in total complications, while the need for major surgery, one or more blood transfusions, prolonged illness, and sustained fever would be classified as "major" complications. *The other institutions were situated as follows: California (9), Pennsylvania (6), District of Columbia (5), Maryland (3), Massachusetts (2), and one each in Arizona, Connecticut, Hawaii, Kansas, North Carolina, Oregon, and Washington. **Types of pre-existing complications identified in this study include diseases of the circulatory system (heart disease, hypertension), diseases of the reproductive system (uterine fibroid tumors), asthma, diseases of the urinary tract, anemia, diabetes, mental disorders, and epilepsy. Diseases of the circulatory system and the reproductive organs represented almost one-half of the conditions presented as pre-existing complications. As would be expected, the number of women with pre-existing complications increased with age and parity.
50 TABLE 8 local and Major Postabortal Complications per 100 Women Obtaining Abortions by Gestation, Procedure, Pre-existing Complications, and Concurrent Steriliza- tion, Total Patients and Local Patients with Follow-up (FU) JPSA, July 1, 1970 - June 30, 1971 Type of Abortion All patients 12 weeks or less 13 weeks or more Total Complications Major Complications Total Local patients Total Local patients patients with FU patients with FU Complication Rates Per 100 Women Obtaining Abortions 5.2 22.2 7.8 26.1 0.6 2.2 1.1 3.0 Patients without pre-existing com- plications, by procedure Suction a/ D&C a/ Saline a/ Hysterotomy b/ Hysterectomy Patients without complications or sterilization 12 weeks or less 13 weeks or more Patients with pre-existing complications, without steriliza- tion 12 weeks or less 13 weeks or more Patients without pre-existing com- plications, with sterilization 12 weeks or less 4.2 6.0 23.4 33.4 49.9 4.2 20.6 12.7 29.9 25.9 35.8 6.1 8.2 27.2 32.9 50.9 6.2 26.0 17.1 35.1 28.0 35.4 0.4 0.5 1.7 6.7 14.3 0.4 1.6 1.4 4.6 .6.1 8.2 0.6 0.8 2.4 6.9 15.6 0.6 2.1 2.0 6.7 7.2 8.0 13 weeks or more Patients with pre-existing complications and sterilization 12 weeks or less 13 weeks or more 43.0 56.5 46.2 60.4 14.9 13.8 17.1 17.4 a/Without tuti.il sterilization b/With tubal sterilization Source: Christopher Tietze and Deborah Dawson. "Induced Abortion: A Factbook," Reports on Population/Family Planning 14, December 1973.
51 As shown in Table 8, abortions performed in the first twelve weeks had a lower complication rate than did those performed subsequently, regardless of whether or not the patient had pre-existing complications or whether she obtained sterilization at the same time as the abortion. Those women who were aborted with suction or D&C, both of which are first trimester methods, experienced the lowest total complication rates and the lowest rates for major complications. The risk of medical complications increases with the length of gestation, even within the first trimester. Women with pre-existing complications and women who underwent con- current sterilization had substantially higher complication rates than did women who were subject only to the complications of the abortion procedure itself. A woman with pre-existing complications obtaining a first trimester abortion was more than three times as likely to develop post-abortal compli- cations as was a healthy woman (a woman going to term with pre-existing complications is also at greater risk). A woman without pre-existing complications who obtained concurrent sterilization during a first-trimester abortion was more than six times as likely to develop complications as was a comparable woman who was not sterilized at the time of abortion. A woman with both pre-existing complications and concurrent sterilization at the time of first-trimester abortion was more than 10 times as likely to develop complications as was a woman without pre-existing complications or sterili- zation. In terms of major complications, the risk was more than 35 times as great for the woman with both pre-existing complications and con- current sterilization. The same patterns are seen in second-trimester abortions, although the differences generally are not as great. The high complication rates reported in the JPSA survey for women undergoing concurrent sterilization reflect the fact that many of these sterilizations required abdominal surgery with general anesthesia either to remove the uterus (hysterectomy) or to perform a tubal ligation. Since the JPSA study, more sterilizations are being performed by laparascopy, which does not require as extensive surgery and has significantly less morbidity than other sterilization methods.* In a recent studyAV complication rates were compared between a group of 108 patients obtaining a combined suction abortion and laparoscopic sterilization, and a group of 195 patients obtaining only a suction abortion. Anesthesia was either local or light general anesthesia in both groups and there were no deaths reported. The complication rates for the two groups were 9.2 and 7.2 respectively, a significantly smaller differential than in the JPSA study and much lower rates than the rates reported with other sterilization techniques. *A laparoscopy is a procedure in which a slender instrument with a lens at the end is inserted into the abdominal cavity through a very small incision, permitting visualization of the organs contained therein and the performance of various surgical procedures, including tubal ligation.
52 If complication rates in JPSA are analyzed by the type of facility in which the abortion was performed, total complication rates associated with the method of suction are substantially lower for clinic patients than for either outpatients or inpatients in hospitals (Table 9). Except for "fever only," where the complication rate was highest among clinic patients, and the category for "all other" complica- tions, where it was marginally higher for outpatients, clinic patients had the lowest incidence of each specific type of complication, and hospital inpatients had the highest. The high rates of "fever only" for clinic patients, combined with low rates of pelvic infection, reflect the dependence of clinics on reports directly from patients rather than from physicians who would have made more specific diagnoses. 5_/ To some extent, this difference in complication rates can be explained by the fact that more patients with pre-existing complications are likely to be handled in a hospital setting. Clinic physicians, who usually perform a higher number of abortions, may be more experienced than other physicians in performing abortions; and there may be less complete reporting of delayed complications for clinic patients than for those treated in hospitals. TABLE 9 Number and Complication Rates of Local Patients with Follow-up Who were Aborted by Suction, by Type of Facility Type of Complication Number of Patients a/ Perforation of uterus Other Injury Hemorrhage Pelvic infection Hemorrhage and Infection Fever only All other Hospital Inpatients 5,350 0.6 1.8 2.0 1.4 0.4 0.9 0.7 Type of Facility Hospital Outpatients 11,538 Rates per 100 women 0.3 1.2 1.2 1.1 0.3 0.4 1.0 Free-Standing Clinic Patients 6,968 0.2 0.2 0.6 0, 0. 2.0 0.4 a/Excludes patients with pre-existing complications and/or sterilization Source: Christopher Tietze and Sarah Lewit. "Joint Program for the Study of Abortion (JPSA): Early Medical Complications of Legal Abortion," Studies in Family Planning 13: 116, June 1972.
53 The Ljubljana Study Information on complication rates associated with first-trimester abortions also is available from a study undertaken in Ljubljana, Yugoslavia, during 1971 and 1972.6/ The study covered 4,700 women aborted for "social" reasons who were randomly assigned to suction or D&C. The duration of the pregnancy was seven to 12 weeks in 95 percent of the cases. Abortions were performed on an inpatient basis with general anesthesia; hospitalization was required for 48 hours following the procedure. Table 10 summarizes the complication rates experienced in this study. For every complication, women obtaining suction abortions ex- perienced lower rates than those obtaining D&C abortions. While low, these rates are consistently higher than those reported from the JPSA study. TABLE 10 Complications of First Trimester Abortions. Ljubljana. 1971-1972 Type of Complication Rates per 100 Women Obtaining Abortions Suction D&C Perforation of uterus 0.0 0.6 Laceration of cervix 0.7 0.7 Complications of anesthesia 2.9 3.7 Heavy bleeding requiring agents to tighten the uterine walls a/ 3.9 6.0 Blood loss exceeding 300 ml b_/ -at abortion 1.9 3.3 - total first 48 hours 5.4 8.9 Post abortion bleeding - requiring treatment 9.6 10.0 - requiring hospitalization 1.5 1.9 Retention of tissue 0.9 1.3 Pelvic Infection or fever c_/ during hospitalization 2.0 1.6 -requiring readmission a/ 0.7 1.3 a/Statistically significant difference (P-.01) b/These data based on a subsample of 530 cases. c/Ineludes endoraetritis, salpingitis, temperature rise to 37.6 degrees centigrade for more than two days. Source: L. Andolsek and M. Owen. "The Operation and the Operator"; L. Andol- sek. "Operative Events"; L. Andolsek and M. Owen. "Blood Loss"; and L. Andolsek. "Infection"; in L. Andolsek. The Ljubljana Abortion Study. 1971-1973. Bethesda: National Institutes of Health, 1974.
54 Saline Abortions As indicated by the JPSA data in Table 8, the complication rates associated with second-trimester abortions, usually performed by the saline method, are considerably higher than those per- formed by D&C or suction. Since second-trimester abortions will continue to be needed, methods are needed to reduce these complication rates.* In a recent paper, Berger and Kerenyi reported that hemorrhage and/or infection accounted for 97 percent of all complications associated with saline abortion.^/ Factors such as age, pre-existing complications, and a history of previous pregnanciesâall associated with an increased risk of hemorrhageâcannot be controlled by the physician. However, prolonged retention of the placenta after delivery of the fetus, which is frequently responsible for subsequent hemorrhaging, is subject to physician control. The study concluded that hemorrhage from saline abortion can be markedly reduced if the placenta is removed within one hour after the fetus is delivered, and infection and fever can be reduced if the highest standards of asepsis are maintained.8/ Long-Term Complications Much concern and controversy have arisen over the potential long-term effects of legal abortion on the reproductive capabilities of women obtain- ing such abortions. Cervical incompetence, scar tissue from infection, or other trauma to the reproductive organs following abortion have been suggested as long-term complications which might result in subsequent spontaneous abortion or prematurity, in subsequent ectopic pregnancies, or infertility. Various studies have been published dealing with these pro- blems, but in most cases methodological shortcomings seriously limit vali- dity of the findings and prevent definitive conclusions about the existence of these long-term complications. Among the problems encountered in these studies is the difficulty of correlating long-term complications with a particular abortion, because *For example, the procedure of amniocentesis, which is used to detect certain fetal defects, can rarely be performed prior to the 15th week of pregnancy. In this procedure, a portion of amniotic fluid is withdrawn from the uterus and the fluid or fetal cells obtained from it are cul- tured for two to four weeks and are subjected to tests for fetal defects. By the time the diagnostic procedures are completed, the woman is well in- to the second trimester of pregnancy. If fetal disorders are identified and the patient elects abortion, a second-trimester method, usually saline, must be used. Chapter 6 has further information on birth defects and the use of selective abortion.
55 intervening full-term pregnancies or miscarriages may also have contributed to the complications. In addition, without complete medical records, it has often proved difficult to conduct retrospective studies, i.e., to iden- tify previous abortions in the medical history of women who incur complica- tions that could potentially be related to prior abortions. Without complete medical records, full documentation on a woman's previous reproductive his- tory may not be obtained, which means that a researcher may have to rely on the woman's memory and willingness to admit previous abortions. There is some evidence that a woman will more readily admit to having had a previous abortion if she has subsequently experienced a premature delivery, still- birth, or neonatal death, than if she subsequently experienced a normal birth .j)/ Other problems include the lack of an appropriate control group to compare with women who have had prior abortions, and the possible in- validity of comparing long-term complications from abortions performed by different methods, in different types of health care delivery systems, and on women of different ages or socio-economic classes. Few studies have provided specific details on the extent to which they controlled for these factors, thereby limiting the applicability of their conclusions. And when significant differences of these types exist between countries, trans- national comparisons must be made with great caution. This seriously affects the ability to draw conclusions about long-term complications from induced abortion in the United States, since legal abortion has been avail- able for too short a time to enable the publication of American studies on this subject. All the data presented in this discussion are from foreign studies and their direct applicability to the United States is not known at this time. A World Health Organization task force has initiated prospective studies in European countries to examine the effects of induced abortion on the outcomes of subsequent pregnancies. The WHO Task Force will identify the reproductive history, including abortion history of pregnant women when they register for prenatal care. Subsequently, the pregnancy outcomes of women who have had a prior abortion and who are pregnant for the second time will be compared with those of women pregnant for the second time with one prior birth, and those of women pregnant for the first time. The Task Force Report should be available in 1977.10/ Premature Births A review of much of the currently available data on the relationship between legal abortion and subsequent premature births indicates that many of the studies did not control for smoking, age of mother, overall maternal health, or socioeconomic status. Many of the stu- dies also did not identify differences in the method of pregnancy termination or indicate whether prior abortions had been obtained illegally or in a medical setting. And in some cases, the existence of prior spontaneous as well as induced abortions was not taken into account. All of these factors are likely to be closely related to the causes of premature births, and a desire to isolate the impact of abortion from these other factors would seem to be a high priority for researchers in the field.
56 A positive relationship between prior induced abortions and the percent of premature infants, defined as weighing 2,500 grams or less, was reported in an Hungarian study of 20,400 births occurring in the second half of 1970.ll/ The proportion of premature infants born to second pregnancy women with a prior abortion was nearly 15 percent of total births to that group, and for second pregnancy women without any prior induced abortions, the proportion was only eight percent of the total births. The percentage of low weight births increased as the number of previous pregnancies increased, regardless of prior abortions, but the abortion group consistently had higher prematurity rates. All abortions can be assumed to be lega1, first-trimester abortions performed by D&C. Unfortunately, no data are presented on the age or socioeconomic charac- teristics of the women in the different groups or whether these factors might have affected the outcomes. Moreover, since D&C is used in only 13 percent of first-trimester American abortions and involves more surgical trauma to the cervix than suction, the relevance of these find- ings to the United States cannot be assessed. A Greek study by Pantelakis et^ al. 12/ provides data on more than 13,000 women registering for prenatal care at Alexandra Maternity Hospital between October 1, 1966 and September 30, 1968. The women were asked to supply information on previous pregnancies, including those terminated by induced abortion. The frequency of premature births to women in the study was then examined for women who had had no previous abortions, previous induced abortions, previous spontaneous abortions, or a combination of the latter two types. Premature birth rates for women with previous abortions were found to be twice those of women without any abortions (17 percent of the total and eight percent of the total, respectively).13/ Although Pantelakis' results are interesting, they are seriously limited by important methodological shortcomings. First, the women in the study were asked to supply the data on previous pregnancies themselves, making these data subject to incorrect recall. (The authors cited studies in which there had been difficulty in getting women to admit accurately a history of prior abortions.)14/ Second, although abortions were theoretically illegal in Greece during the period under study, the authors did not seek to determine whether prior induced abortions had been obtained in a medical setting or in another environment. Third, the study was not adequately controlled for either the number of prior births or the pregnancy order. And finally, the method of termination of previous induced abortions was not identified. These factors consider- ably weaken the conclusions of the study, and once again, the extent to which the medical experience of Greek women can be applied to the United States is unknown.
57 A different conclusion was made by RohC and Aoyama,l_5/ who obtained reproductive histories of nearly 2,800 Japanese women through a mail survey (2,170 respondents) and interviews (609 respondents) in April 1971. Pre- mature births (defined as live birth less than 2,500 grams) for the first post-abortion pregnancies in 587 women with prior abortions were compared with the premature births for women without prior abortions (both birth order and maternal age at completion of pregnancy were controlled for). The authors concluded "that women in their most active reproductive years, i.e., 20 to 39, do not seem to be at excess risk of prematurity due to a previous induced abortion."J.6/ However, this study also is flawed by the fact that the women participating were asked to provide not only their prior reproductive history, but also the birth weights of their children. The authors indicated that attempts to verify the data will continue. Several studies have concluded that there is no significant difference in the prematurity rates of women with prior induced abortions compared with other women. In a carefully controlled historical prospective study of 948 Yugoslavian women whose first pregnancies were terminated by induced abortion (222) or delivery (726) during 1968-1969, Hogue did not find any evidence of greater prematurity or other reproductive problems among women who had had abortions compared with other groups of women. There was no evidence of impaired fertility among aborters or of an increased incidence of first- or second-trimester spontaneous abortions. When second-pregnancy deliveries were compared, deliveries following induced abortion had a higher prematurity rate than did those following another delivery (approaching statistical significance), but there was no difference when first-pregnancy deliveries were compared with first deliveries subsequent to induced abortion. The pregnancy-order differential is thus due to the known higher risk of prematurity among first-order births. Apparently induced abortion when it precedes such pregnancies neither increases this risk nor protects against it.17/ Spontaneous abortions A few studies have addressed independently the relationship between induced abortion and subsequent spontaneous second-trimester abortion. The explanation for this concern is that cervical incompetence may result from forced dilatation of the cervix during vaginal abortion. The longer the gestation, the larger the cervix must be dilated, and the greater the risk of cervical laceration and subsequent inability of the cervix to bear the pressure of a full term pregnancy.
58 Wright et^ al^ 1^8/* studied 3,314 patients who completed a pregnancy at Queen Charlotte's Maternity Hospital in London during 1971. Patients were asked to provide detailed information on their obstetric history, including whether any therapeutic abortion was obtained, the length of gestation, method of termination and the hospital in which termination was performed. Ninety-one patients were identified with a therapeutic abortion immediately prior to their current pregnancy (Group A) and were matched by age with a control group of patients who had had one spontaneous abortion before the current pregnancy (Group B). Pregnancy outcomes on 3,223 other women were also examined (Group C). More than 90 percent of the women in the abortion group had had no prior pregnancies than the one originally terminated. Those pregnancies ending in second-trimester spontaneous abortion are summarized below. Spontaneous Abortions Total Pregnancies Number Percent Group A (Prior 91 89 Induced abortion) Group B (Prior Spon- taneous abortion) 91 11 Group C (Other Deliveries and abortions) 3,223 30 1 Based on these data the authors concluded that temporary or permanent cervical incompetence is induced by the procedure of dilatation of the cervix during termination. This is further suggested by the fact that a control group had signifi- cantly fewer second-trimester abortions. The only important dif- ference between the two groups was that the previous-termination patients had had forcible dilatation of the cervix.19/ *Spontaneous abortion in the second trimester is often dealt with as part of the prematurity problem. Data on first-trimester spontaneous abortion do not seem to be available, perhaps because so many of these abortions occur without the woman knowing she was pregnant.
59 Considerable caution must be taken in accepting these conclusions since the methodology used in this study has definite limitations. First of al1, as stated earlier, use of retrospective recall by the patient usually results in an incomplete and sometimes incorrect obstetrical history. Wright et al. admit that even their own experience has shown this to be true. Second, the differential effect of repeated terminations (or pregnancies) is not mentioned, although the authors indicate that three of the patients experiencing spontaneous second-trimester abortions had had two or more previous pregnancies. No data are provided in this area on Group A women going to term. Third, the timing of termination is not provided for all patients in Group A. Only one of the 13 patients with apparent cervical incompetence was terminated prior to 10 weeks gestation in her earlier pregnancy, and one was terminated vaginally as late as 18 weeks gestation. Though this is mentioned as a possible contributing factor to spontaneous abortion, data are not provided on the lengths of gestation of the earlier pregnancies of the other women in Group A. Fourth, as Liu et^ al. have pointed out,2J3/ the authors neglected to take into account the normal rate of cervical incompetence (1-2 per 1,000 deliveries). Finally, Liu questions the validity of the results given that the spontaneous first-trimester abortion rate was only one percent in the Wright sample, and in a normal population this rate is usually between 10-20 percent. This suggests that Wright's sample may not be a representative group of women for obstetrical comparisons. Roht and Aoyama also examined the spontaneous abortion rates following induced abortion in their 1971 sample. Although their methodology also has important weaknesses, some of which are similar to those mentioned about the Wright study, their conclusions are exactly the opposite. Roht and Aoyama found no difference in the spontaneous abortion rates for "no abortion" and "post-abortion" women of the same age group. Their data demonstrated only that there was an increasing risk of spontaneous abortion with older maternal age.2I/ Ectopic Pregnancies Existing data on the relationship between legal abortion and subsequent ectopic pregnancies are also limited but seem to suggest that no negative effects are likely to occur. This conclusion follows from a review of Eastern European literature by Emily Moore-Cavar22/ and from a controlled study conducted in Ljubljana, Yugoslavia.23/ In a study conducted in 1971 in Novi Sad, Yugoslavia, Beric and Kupresanin found that from 1960 to 1970, there was an increase in deliveries, an increase in legal abortions, a decrease in hospital admis- sions for "other" abortions (spontaneous and illegal), and a slight overall decrease in ectopic pregnancies. In another Yugoslavian study conducted in 1972, Masic noted a "relation between ectopic pregnancy and provoked criminal abortion, but not between ectopic pregnancy and legal artificial abortion."24/
60 A carefully controlled study undertaken in Ljubljana, Yugoslavia between January 1971 and July 1973 developed data from a sample of 200 women with ectopic pregnancies, and a pooled sample of 600 women having full-term intrauterine pregnancies and 240 women who sought induced abortions during the time of the study. The risk of having an ectopic pregnancy for women with a positive history of induced abortion was then compared with the risk of ectopic pregnancies among the women in the pooled sample. The data indicated that the risk of ectopic pregnancy for women with one previous induced abortion was exactly the same as for women with no history of induced abortion. For those women with two or more induced abortions, the risk of ectopic pregnancy was only 80 percent as great as it was for women without prior abortions.25J These data led the authors to conclude the following: We believe that our data do not support the hypothesis that induced abortion leads to an increased risk of subsequent ectopic pregnancy. In this study, women with several induced abortions are in fact at slightly reduced risk of subsequent ectopic pregnancies. This is not surprising, since having had several induced abortions indicates a normally functioning reproductive system.26/ Infertility Data concerning infertility following induced abortion are sparse, extremely poor, and generally inconclusive. In a review of the existing literature, Moore-Cavar concluded:27/ Methodological problems abound in any effort to establish statistical probabilities of infecundity attributable to abortion. "Abortion makes you sterile" is generally not accompanied by an explanation of the type of abortion (lega1, illega1, or spontaneous, single or repeated) which is presumed to have this effect, nor is there usually any mention of the proportion of comparable non-aborting control women who are found to be infecund because of aging or because of earlier pregnancies carried full term. To date, it can be said that there is no conclusive evidence of an increased risk of reduced fecundity from terminating a pregnancy by abortion. Even if data were developed with reasonable accuracy, they may not be relevant to the United States since the data base may be old, and may reflect medical complications arising from abortion methods such as hysterotomy, which are rarely used in the United States today.28/ Teenage Abortions and Long-Term Complications Teenage women con- stitute a special case, in that any pregnancy, be it terminated by abortion or by a full-term birth, seems to increase the risk of prematurity in subse- quent pregnancies.23/ Russell reported on the first pregnancy outcome and subsequent reproductive history of 62 pregnant women under 16 whom he treated between January 1960 and December 1971. Fifty of these pregnancies were terminated by therapeutic abortion, 11 continued to term, and a single one ended in spontaneous abortion at 10 weeks gestation. The majority (46)
61 of the induced abortions were performed by suction (30) or D&C (16); two each were performed by hysterotomy and prostaglandins. Only 38 of the pregnancies were terminated at less than twelve weeks gestation; nine pregnancies were terminated between 12 and 14 weeks.â¢30_/ Length of gestation was not matched with method of termination in the article cited so that it was not possible to determine whether first-trimester methods were used to terminate second-trimester pregnancies. Significant immediate complications were reported from the 50 abor- tions, and cervical laceration requiring sutures was reported in five cases. Russell explained that "the cervix of the young teenager, pregnant for the first time, is invariably small and tightly closed and especially liable to damage on dilatation."31/ Of 53 subsequent pregnancies experienced by this cohort of women, a substantial number ended in either fetal, neonata1, or childhood death. The outcomes of these pregnancies are shown below.32/ Outcome of Pregnancy Number Percent of Total Therapeutic abortion 6 11 Spontaneous abortion 19 36 Stillborn (30 weeks gesta- t ion) 1 2 Premature Delivery 7 13 Term delivery JZO ^8 Total pregnancies 53 100 Two of the premature babies died, one at three days and one at three months. Four of the term birth deliveries died, one during the neonatal period and three within the subsequent 30 months. Of the 11 teenagers whose initial pregnancies had been carried to term, nine subsequent pregnancies were reported, all of which went to term and all of whose babies survived. The data provided in this study are not adequate grounds on which to draw conclusions about the relationship between induced abortion in teenagers and subsequent reproductive impairment. There is no matched control group for the women undergoing therapeutic abortions. The age at first and subsequent pregnancies is not specified, nor is the interval of time between pregnancies. The pregnancy order of women having the 53 subsequent pregnancies is not examined, although Russell provides anecdotal evidence that two women have had three and four spontaneous abortions, respectively, without being able to have a successful pregnancy.33/
62 Finally, neither the method of abortion nor the existence of immediate post-abortal complications is compared with future reproductive history. It is impossible to know, for example, whether the women suffering cervical lacerations were a significant proportion of those having subsequent spontaneous abortions or premature deliveries. Perhaps recognizing these difficulties, Russell does not conclude a causal relationship between therapeutic abortion and subsequent reproduc- tive difficulties. He cites the higher complication rates for teenagers undergoing full-term pregnancies (including subsequent prematurity, fetal deaths, and perinatal mortality), and the serious socia1, familial and psychological difficulties faced by teenage mothers. He concludes that teenage "pregnancy...clearly carries medical hazards for the girl and her baby. But the full significance of these early pregnancies will only be established by carefully planned prospective studies involving several disciplinesâobstetrics, pediatrics, psychiatry and sociology."34/ An example of the complications arising from teenage pregnancies is provided by a study during 1967-1969 of 180 first pregnancies of school age women in New Haven who terminated their first pregnancies with a birth. A significantly higher risk of prematurity and perinatal death was found in subsesquent pregnancies. Only 12 percent of the first pregnancies ended in premature birth (defined as less than 2,500 gram), but 27 percent of the 103 subsequent pregnancies resulted in premature births. Increasing birth order was associated with increased risk.35/ Although Roht and Aoyama focused primarily on women 20-39 years of age, some data were reported on teenagers in their study. The sample size is too small to represent conclusive findings, but the data show that teenage women had higher rates of prematurity than other women, whether or not the birth occurred post-abortion. The number of completed pregnancies in the sample is too small to assess whether induced abortion increases this risk.3j6/ In conclusion, the fact that teenage pregnancies have a greater risk of death and medical complications than pregnancies of older women has been documented extensively in the literature and has been repeated here. What is not yet clear, however, is under what conditions and to what extent induced abortion aggravates that risk in subsequent pregnancies. Since nearly one-third of the legal abortions in the United States are obtained by teenage women, it would seem that further research in this area would merit high priority.
63 Rh Immunization One other potential long-term complication of induced abortion is the sensitization of women with Rh-negative blood by Rh-positive red blood cells from the fetus during abortion (or during miscarriage, full-term birth, or even amniocentesis). This process will cause the formation of antibodies, which could be harmful to the health of fetuses in future pregnancies. However, it is possible to prevent the formation of the antibodies in Rh-negative women by injecting them with Rh immune globulin (commercially distributed as RhoGAM), which essentially neutralizes that part of the Rh-positive blood of the fetus that would otherwise have stimulated the production of maternal antibodies. If Rh-negative women undergoing abortions are injected with immune globulin, there should not be any complications in a subsequent pregnancy. To summarize, no definitive conclusions can be drawn from the diverse data obtained from the studies cited above about the long- term complications of legal abortion in the United States. Not only is there contradictory evidence, but many of the studies have serious problems of method. Reliance on retrospective methodology may lead to biased recall; poorly designed hypotheses and poor sampling, with or without controls, lead to inaccurate cause and effect conclusions. Finally, it is extremely difficult to isolate the many factors that can have a long-term effect on a person's health status. Not only is the method and trimester of abortion important, but also the woman's physical health, her socioeconomic status, and condition of intervening pregnancies may be as important or more important than the abortion itself in affecting a woman's reproductive capabilities in the long term. Illegal Abortion and Medical Complications In countries with laws that prohibit or greatly restrict legal abor- tion, induced abortions that are performed by physicians outside the strict limitations of the law, or those performed by non-medical personnel or by the woman herself are illegal. In the discussion below, illegal abortion is generally meant to refer only to those cases where abortion is performed by non-medical personnel or by the pregnant woman in a setting which is outside the formal health care system.
64 Among the non-medical procedures used for inducing abortion are eating or drinking quinine or other drugs, introduction of chemicals into the vagina, and mechanical methods such as inserting blunt or sharp instru- ments into the uterus through the vagina. The drugs quite often lead to poisoning, or vomiting so intense that it results in dehydration and eventual death unless fluid replacement compensates the loss. Intense vomiting is sought because it might cause contractions and thus expel the fetus. Inserting chemicals or implements in the vagina or uterus can lead to (1) infection; (2) injury to the membranes of the vagina; (3) perforation of the uterus with the possibility of injuring other organs in the abdominal area; (4) bleeding due to retained fetal or pla- cental tissue and (5) air embolism. These medical complications have re- sulted in numerous deaths, and serious illnesses requiring lengthy hospita- lization. They represent, therefore, a serious public health problem, although one that has declined in importance as legal abortion has become more accessible. A septic abortion is an infection of the uterine cavity or of the retained products of conception following an abortion. If not treated promptly, the infection may extend to the area surrounding the uterus (peritonitis) or the blood stream (septicemia) and may result in death. Septic abortion can occur following spontaneous and legal abortion, but it most commonly results from illegal abortion. An incomplete abortion occurs when the placenta is not completely expelled from the uterus. This may cause severe bleeding until the remaining tissue has been removed, and also predisposes the uterine cavity to infection. The frequency of medical complications due to illegal abortions is difficult to determine. One measure that has been used is the number of hospital admissions due to incomplete or septic abortions.* Although admissions due to incomplete/septic abortions are often recorded as mis- carriages (as some actually are), or of an "unknown" origin, it is be- lieved that the majority of abortions leading to major complications, and especially those with fatal outcomes, are illegally induced.** *Not all illegally induced abortion complications are admitted to the hospital. Many injured women are treated on an outpatient basis or by a private physician. Additionally, some women probably resort to self- treatment with varying degrees of success. The number of hospital admissions due to incomplete/septic abortion can only be used as a partial indicator to measure the incidence of illegal abortion. **See for example, David N. Danforth, ed. Textbook of Obstetrics and Gynecology. 2nd ed., New York: Harper & Row, 1971, p. 345; and J. P. Greenhill and Emanuel A. Friedman. Biological Principles and Modern Practice of Obstetrics, Philadelphia: W. B. Saunders Co., 1974, p. 380.
65 With the introduction of non-restrictive abortion laws, the number of hospital admissions for incomplete and septic abortion has declined sharply. For example, municipal hospitals in New York City have reported a steady decline in the number of admissions due to incomplete abortions (Table 11). In 1969, the year before the non-restrictive legislation was implemented, 6,524 incomplete abortions were reported; in 1971, the first year after the new legislation went into effect, this number declined by 31 percent. In 1973, 3,253 admissions were reported, or a rate of 133 per 1,000 deliveries. TABLE 11 Admissions to Municipal Hospitals in New York City for Incomplete Abortions, 1969-1973 Incomplete Abortion Admissions per Years Number of Admissions Number of Births 1,000 Births 1969 6524 27,842 234 1970 5293 31,308 169 1971 3643 27,998 130 1972 3538 24,989 142 1973 3253 24,502 133 Source: Personal communication to Emily Moore-Cavar from Jean Pakter, Director, Bureau of Maternity Services and Family Planning, Department of Health, New York City and published in International Inventory of Information on Induced Abortion, New York: Inter- national Institute for the Study of Human Reproduction, Columbia University, 1974, p. 642. Data from Brooklyn, Atlanta, Los Angeles and San Francisco also show that a decline in the number of hospital admissions for incomplete abortion is related to the growing availability of legal abortion, with its lower frequency of medical complications. A study done at Kings County-State University Hospitals in Brooklyn during 1967-1971 reported that the number of "spontaneous" abortions fell to 20 percent during this period. The authors attributed this decline in part, to "a decline from the previous number of abortions initiated illegally outside a hospital and then brought for completion to the hospita1, where for the lack of better information the abortion was classed as spontaneous.-^37/ Other factors mentioned were elective abortion of pregnancies which might have subsequently resulted in a spontaneous abortion, and a general decline in the number of pregnancies.38/
66 A study of the frequency of admissions due to complications resulting from illegal abortion was undertaken at Grady Hospital in Atlanta from 1969 to 1973. Only those women who actually admitted having had an illegal abortion were included in the study. Its results, shown in Figure 7, indicate a decline in the number of admissions for complications due to illegal abortionâfrom a high of 33 in the last quarter of 1970, to five in the first quarter of 1973. However, a dramatic drop in this seriesâ from 24 to nine admissions in one quarterâdid not occur until the second quarter of 1972. This correlates with the changes in Georgians abortion laws during this period. Georgia adopted a moderately restrictive abortion law in 1968, but procedural requirements did not become less cumbersome until 1970, and access to abortion on request did not become readily available until 1973. The number of legal induced abortions at Grady Hospital during this period increased from 30 in all of 1969 to 498 in the first quarter of 1973.^9/ These data suggest that there may be a time lag in women's awareness of changed laws, and that both abortion legislation and practices must be non-restrictive before some women turn from illegal abortions to legal abortions.40/ FIGURE 7 to 00 V) Q < O z Number of Hospital Admissions for Complications due to Illegal Abortions, Grady Hospita1, Atlanta, 1969-1973 35 iâ 30 25 20 15 10 0 I I I J I J I I I I I I III IV I III IV I II III IV I II III IV I QUARTER 1969 1970 1971 1972 1973 Source: Ronald S. Kahan, Lawrence D. Baker and Malcolm G. Free;aan. -The Effect of Legalized Abortion on Morbidity Resulting from Criminal Abortion," American Journal of Obstetrics and Gynecology 121:115 January 1, 1975.
67 Summary Medical complications associated with legal abortion may occur at the time of the abortion (immediate), within thirty days following the procedure (delayed), or at some later time (late). The frequency and severity of complications vary as a function of the trimester in which the abortion is obtained, the method used, and whether or not pre-existing complications or sterilization are also present. As presented by the JPSA study, major complications in women undergoing first-trimester abortions by suction or D&C are rare. More frequent and serious complications occur in women undergoing second-trimester abortions and/or concurrent sterilization, and in those having pre-existing complications. The impact of legal abortion on long-term complications is more difficult to evaluate, particularly in the United States, where the history of non-restrictive abortion practice is too short to provide longitudinal data. Although there is some evidence from Hungary and Greece associating a history of repeated abortions with subsequent premature births, different studies from Japan and Yugoslavia conclude that prior induced abortions cannot be statistically linked to prematurity. Similar contra- dictory evidence is found on infertility. The data on ectopic pregnancies, particularly from the Ljubljana study, are somewhat more reliable and lead to the tentative conclusion that induced legal abortion does not lead to a greater risk of ectopic pregnancy post-abortion. There is evidence that spontaneous second-trimester abortions may be related to previous induced abortions, particularly for teenagers. But it also appears that a teenager is at risk if she carries a pregnancy to term. The length of gestation, method of termination, and other charac- teristics relating to the pregnant woman are not adequately sorted out at this time. Thus, the inconsistent findings of these diverse studies do not permit definitive conclusions to be drawn on the long-term complications of legal induced abortion in the United States, and particularly if that abortion is an early, first-trimester abortion performed by suction. Although it is impossible to estimate the frequency of medical complica- tions due to illegal abortions, one indicator that has been used for this purpose is the number of hospital admissions due to incomplete or septic abortions. Existing trend data on hospital admissions and discharges indicate that abortion-related complications have declined over the past several years; in several hospitals the numbers and rates of women admitted for treatment of incomplete septic abortions fell sharply after the states in which the hospitals were located had introduced nonrestrictive abortion legislation or practices. These data lead to the conclusion that many women who previously relied on illegal abortions are now obtaining their abortions legally with a lesser risk of severe medical complications.
68 In San Francisco, the number of septic abortions per 1,000 deliveries dropped by 68 percent between 1967 and 1969.41/ The Los Angeles County/ University of Southern California Medical Center reported a comparable decline of 78 percent from 1967 to the first nine months of 1971.42/ Although other factors, such as the increased use of effective contraception and a declining number of unwanted pregnancies might have contributed to the decline in high risk abortions, it is likely that the introduction of less restrictive legislation in 1967 was a major factor in reducing total illegal abortion-related complications in California.
69 REFERENCES 1. Christopher Tietze and Deborah Dawson. "Induced Abortion: A Factbook," Reports on Population/Family Planning 14: 28, December 1973. 2. "Symposium: A Report on Prostaglandins for Abortion," Contemporary OB/GYN. 2: 83-108. 3. Christopher Tietze and Sarah Lewit. "Joint Program for the Study of Abortion (JPSA): Early Medical Complications of Legal Abortion," Studies in Family Planning 3: 97-122, June 1972. 4. John I. Fishburne, David A. Edelman, Jaroslav F. Hulka, and Jack P. Mercer. "Morbidity of Combined Outpatient Laparoscopic Sterilization and Therapeutic Abortion vs. Abortion Alone: A Comparative Study," Obstetrics and Gynecology (in press). 5. Tietze and Lewit, p. 116. 6. L. Andolsek, ed. The Ljubljana Abortion Study. 1971-1973, Bethesda: National Institutes of Health, 1974. 7. Gary S. Berger and Thomas D. Kerenyi. "Control of Morbidity Associated with Saline Abortion," A paper presented at the Twelfth Annual Scientific Meeting, Association of Planned Parenthood Physicians, April 16-17, 1974. 8. Ibid. 9. Carol J. Hogue. "Low Birth Weight Subsequent to Induced Abortion: An Historical Prospective Study of 948 Women in Skopje, Yugoslavia," American Journal of Obstetrics and Gynecology (forthcoming, 1975). 10. Personal communication from Christopher Tietze. 11. Hungary: Central Statistical Office. Perinatalis Halalozas. Budapest, 1972, cited in Tietze and Dawson, p. 42. 12. Stefanos N. Pantelakis, George C. Papadimitriou, and Spyros A. Doxiadis. "Influence of Induced and Spontaneous Abortions on the Outcome of Subsequent Pregnancies," American Journal of Obstetrics and Gynecoloey 116: 799-805, July 15, 1973. 13. Ibid., p. 802. 14. Ibid., p. 805.
70 15. Lewis H. Roht and HIdeyasu Aoyama. "Induced Abortion and its Sequelae: Prematurity and Spontaneous Abortion," American Journal of Obstetrics and Gynecology 120: 868-874, December 1, 1974. 16. Ibid., p. 872. 17. Carol J. Hogue. "Prematurity Subsequent to Induced Abortion in Kospje, Yugoslavia: An Historical Prospective Study," unpublished dissertation, University of North Carolina, Chapel Hil1, North Carolina, 1973. 18. Charles S. W. Wright, Stuart Campbell and John Beazley, "Second Trimester Abortion after Vaginal Termination of Pregnancy," The Lancet 1: 1278-1279, June 10, 1972. 19. Ibid., p. 1279. 20. David T. Y. Liu, H. A. H. Melville, and Terry Martin. "Subsequent Gestational Morbidity after Various Types of Abortion," The Lancet 2: 431, August 26, 1972. 21. Roht and Aoyama, p. 873. 22. Emily Moore-Cavar. International Inventory of Information on Induced Abortion. New York: International Institute for the Study of Human Reproduction, Columbia University, 1974, pp. 453-454. 23. M. Hren, T. Tomazevie, and D. Seigel. "Ectopic Pregnancy," The Ljubljana Abortion Study. 1971-1973. pp. 34-38. 24. Moore-Cavar, pp. 453-454. 25. Hren et^ al., pp. 34 and 37. 26. Ibid., p. 38. 27. Moore-Cavar, p. 455. 28. See Jan Lindahl. Somatic Complications following Legal Abortion. Stockholm: Svenska Bokforlaget, 1959. p. 146. The author cites a 4.6 percent infertility rate following legally induced abortion in Sweden. 29. J. K. Russell. "Sexual Activity and its Consequences in the Teenager," Clinics in Obstetrics and Gynecology 1: 683-698, December 1974; and James F. Jekel et al. "A Comparison of the Health of Index and Subsequent Babies Born to School Age Mothers," American Journal of Public Health 65: 370-374, April 1975. 30. Russel1, pp. 692-693.
71 31. Russell, p. 693. 32. Ibid., p. 694. 33. Ibid. 34. Ibid. 35. Jekele^al., pp. 371-372. 36. Roht and Aoyama, p. 871. 37. Jonathan T. Lanman, et^ al^. "Changes in Pregnancy Outcome after Liberalization of the New York State Abortion Law," American Journal of Obstetrics and Gynecology 118: 490, February 15, 1974. 38. Ibid. 39. Ronald S. Kahan, Lawrence D. Baker, and Malcolm G. Freeman. "The Effect of Legalized Abortion on Morbidity resulting from Criminal Abortion," American Journal of Obstetrics and Gynecology 121: 115, January 1, 1975. 40. Ibid. 41. Gary K. Stewart and P. J. Goldstein. "Therapeutic Abortion in California: Effects on Septic Abortion and Maternal Mortality" Journal of Obstetrics and Gynecology 37: 510-514, April 1971. 42. Paul Seward, etâ¢ al. "The Effect of Legal Abortion on the Rate of Septic Abortion at a Large County Hospita1," The American Journal of Obstetrics and Gynecology 115: 335-338, February 1973.