Opinion
No. 79-6403-Civ-SMA.
November 4, 1982.
Bruce Rogow, Fort Lauderdale, Fla., Nan D. Hunter, American Civil Liberties Union Foundation, New York City, Joel Lumer, Miami, Fla., Mary Ellen Shoemaker, Fort Lauderdale, Fla., for plaintiff.
Jim Smith, Atty. Gen. by Gerald B. Curington, Asst. Atty. Gen., Dept. of Legal Affairs, Tallahassee, Fla., for defendants.
MEMORANDUM OPINION
The Court is required to address again the difficult and controversial issue of the permissible degree of governmental regulation of a woman's fundamental right of privacy in the abortion decision. In June 1979, the Florida Legislature enacted the Medical Practice Act, which contains a section regulating abortions in Florida. Subsection 4(a) of Fla.Stat.Ann. § 390.001 (1982) requires an unmarried minor to have either the written informed consent of a parent, custodian or legal guardian or an order from the Florida circuit court prior to obtaining an abortion. Subsection 4(b) requires a wife who is neither separated nor estranged to furnish her husband with notice of the proposed abortion and an opportunity to consult with her concerning the procedure. Any person who willfully performs, or participates in, a termination of pregnancy in violation of the statutory provisions is guilty of a felony.
Chapter 79-302, Fla.Stat.Ann. ch. 458, § 458.0015 et seq. (1979), amended by laws 1980, chapter 80-208.
Fla.Stat.Ann. § 390.001 (1982) (previously codified as Fla.Stat.Ann. § 458.505 (1979)).
Section 390.001(4)(b) reads as follows:
If the woman is married, the husband shall be given notice of the proposed termination of pregnancy and an opportunity to consult with the wife concerning the procedure. The physician may rely on a written statement of the wife that such notice and opportunity has been given, or he may rely on the written consent of the husband to the proposed termination of pregnancy. If the husband and wife are separated or estranged, the provisions of this paragraph for notice or consent shall not be required. The physician may rely upon a written statement from the wife that the husband is voluntarily living apart or estranged from her.
Fla.Stat.Ann. § 390.001(10).
On July 13, 1979, this Court preliminarily enjoined the enforcement of Section 458.505(4)(a), now Section 390.001(4)(a), regulating the performance of abortions on unmarried minors. Jones v. Smith, 474 F. Supp. 1160 (S.D.Fla. 1979). And in December 1979, the Court declared both the parental consent for unmarried minors and the spousal notification requirements unconstitutional in that they impermissibly invade a woman's fundamental right of privacy in the abortion decision. Scheinberg v. Smith, 482 F. Supp. 529 (S.D.Fla. 1979).
This case was originally styled Jones v. Smith but the caption was amended on Plaintiff's request to reflect the Plaintiff's true name.
On appeal, the Court of Appeals for the former Fifth Circuit affirmed the district court's determination that the provision regulating a minor's access to abortion is unconstitutional, but vacated this Court's determination that the spousal notification provision is constitutionally defective. Scheinberg v. Smith, 659 F.2d 476 (5th Cir. 1981). Regarding the spousal notice provision, the appellate court found a compelling state interest in maintaining and promoting the marital relationship and in protecting a husband's interest in the procreative potential of the marriage are compelling. Id. at 483-486. Therefore, the burden placed by the Florida Legislature on a woman's abortion decision is justified if the statute, as enacted, is drafted narrowly enough in furtherance of these interests. The district court found section 4(b) underinclusive "because it does not require that a married woman notify and consult with her husband about an impending hysterectomy or tubal ligation." Scheinberg, supra, 482 F. Supp. at 540. The Court of Appeals, however, disagreed, holding that a court may not demand that the legislature, particularly in the area of abortion, address each manner of furthering a specific state interest when it legislates on one discrete matter. Scheinberg, supra, 659 F.2d at 486. The trial court had also found the notice provision overinclusive because it "makes no exception for a married woman's carrying the child of someone other than her husband." Scheinberg, supra, 482 F. Supp. at 540. However, the appellate court held that if the abortion procedure poses a substantial enough risk of a decrease in fertility to affect detrimentally, in more than a de minimis fashion, the procreative potential of a marriage, it is irrelevant whether the husband is the father of the fetus and the legislature acted within the proper constitutional sphere despite its failure to limit the notice and consultation provisions to jointly conceived children. Id. at 486-487. On the other hand, if the abortion procedure poses a de minimis risk or less to the procreative potential of a marriage, then it is relevant whether the husband is the father of the fetus, and the statute as drawn is underinclusive and unconstitutional. Id.
The appellate court also rejected the district court's holding that the statute was drafted inadequately because compliance with its notice and consultation requirements might result in an unacceptable delay for a woman wishing to secure an abortion. It found that any such delay would be a volitional delay referable to the actions of the individuals rather than from a constitutionally impermissible, statutory mandated delay. Scheinberg, supra, 659 F.2d at 486, n. 6.
In the decision declaring the spousal notification provision unconstitutional, this Court found that the testimony and record "overwhelmingly supported Plaintiff's position on the spousal notice and consultation provision." Scheinberg, supra, 482 F. Supp. 533, n. 11. The unrebutted testimony of various medical, psychological and social work experts, all of whom agreed, indicated that in a host of various marital situations, compelling notification would "at the least, produce anxiety and stress for the woman and her marriage," or by virtue of delay or the seeking of an illegal abortion, cause "serious, perhaps irreversible consequences" to herself. Id. at 538. Consequently, it was this Court's view that the notification provision did not further marital harmony and created a potential danger to a woman's health, including her childbearing capacity. Because of that conclusion, this Court did not find it necessary to specifically address the "procreative potential" question, stating:
Moreover, the Court questions whether the notice and consultation requirement furthers a husband's interest in the procreative potential of his marriage. The parties presented conflicting expert testimony concerning whether a first trimester abortion creates a substantial risk of loss of fertility in women.
The Court of Appeals found that because the district court did not make a specific finding as to the degree of risk posed by abortion to the procreative potential of a marriage, its findings were inadequate to sustain invalidation on the alternative grounds that, although the state interest in the procreative potential of the marriage is compelling, the statute is overinclusive if it includes a requirement of notice to husbands who may not have sired the fetus. Scheinberg, supra, 659 F.2d at 486. Accordingly, the case was remanded:
. . . for a specific finding of whether the legislature could have concluded reasonably, from the evidence in the record, and from any further evidence the parties may present in light of our holding today, that the abortion procedure, as performed properly by licensed medical practitioners according to methodology approved by prevailing medical authority and generally in use in Florida, poses a greater than de minimis risk to a married woman's future ability to bear children.Id. at 487.
Plaintiff argued at the evidentiary hearing on remand that this Court should also address the issue of the Florida Legislature's intent and purpose in enacting § 390.001(4)(b). Plaintiff contends that the Court of Appeals admonished the litigants to assist the appellate court on any future review of this case by advancing evidence of state interests at the remand hearing. Scheinberg, supra, 659 F.2d 4756, n. 2. Although the Court would respectfully disagree with the appellate court's conclusion that the parties stipulated as to the legislature's intent, the Court agrees with the Defendants that the Court of Appeals' acceptance of the alleged stipulation, and attendant determination that the state interests are compelling, means that the legislature's intent is now the law of the case. See Carpa, Inc. v. Ward Foods, Inc., 567 F.2d 1316 (5th Cir. 1978); EEOC v. International Longshoremen's Assn., 623 F.2d 1054, 1058 (5th Cir. 1980). However, the Court permitted the Plaintiff to proffer evidence supporting his contention that the two state interests identified by the Court of Appeals were not actually considered by the legislature and that the real purpose of the statute was to limit a woman's right to an abortion. (Testimony of Kenneth Myers and Elaine Gordon; Plaintiff's Exhibits 23, 24, 25 and 26). The State proffered that the legislature wished to further two state interests through the enactment of the spousal notification provision: maintaining and promoting the marital relationship and protecting a husband's interest in the procreative potential of the marriage. (Testimony of Edgar Dunn).
In the event the Court has misconstrued the holding of the appellate court, or if on reconsideration that court determines that the intent of the legislature was not stipulated, and if therefore the matters were deemed properly before this Court on remand, the Court holds that Defendants have failed to show a legislative intent other than to limit a woman's right to an abortion. The testimony on proffer clearly shows that the two identified interests were not considered by the Florida Legislature prior to enactment of the Medical Practice Act. The Court does not find that the intent is even nominally related to any identifiable state interest other than to limit the availability of abortions in Florida.
At the onset, the Court notes that the statute in question impinges upon abortions during all trimesters of pregnancy. therefore, it is appropriate to test its constitutionality by focusing, as the parties did in the 1979 trial, on the first trimester when the protection is strongest against state interference with a woman's decision to terminate a pregnancy. See Planned Parenthood v. Danforth, 428 U.S. 52, 69, 96 S.Ct. 2831, 2841, 49 L.Ed.2d 788 (1969); Charles v. Carey, 627 F.2d 772, 782 (7th Cir. 1980). The Court also concludes that because first trimester abortions comprise at least 90% of the abortions performed in Florida, the methodology of those abortions is, for all practical purposes, the "methodology approved by prevailing medical authority and generally used in Florida."
If the law is unconstitutional as to abortions performed in the first trimester, it must, therefore, be invalidated. Plaintiff objected to any testimony by Defendants' witnesses, or the introduction of literature, addressing abortions performed in the second or third trimester, arguing that it is irrelevant since the focus must be on the effect of first trimester abortions on the procreative potential of the marriage. While the Court agrees with Plaintiffs that its evidentiary focus should be on the effects of first trimester abortions, and consequently finds that evidence as to the sequela of second and third trimester abortions is not necessarily relevant to the issue at hand, nevertheless, it has permitted the introduction of evidence also dealing with second and third trimester abortions. However, such evidence will be accorded less weight than will be studies focusing exclusively on first trimester abortions and its methodology in the Court's ultimate factual determination.
The witnesses agreed that abortions performed in the first trimester, because of the safer methodology used and the shorter gestational age of the fetus, are a safer medical procedure than abortions performed in later trimesters and, accordingly, are less likely to create any complications or produce any adverse effects upon the women.
The Court interprets the de minimis test stated by the Court of Appeals to mean that the Defendants must show by a preponderance of the evidence that an induced abortion as performed in Florida, has a definite, even if small, adverse effect on a woman's future childbearing ability. The term de minimis is used by the Court of Appeals to describe the extent of the harm; it does not diminish the underlying burden of proof upon the state nor does it permit this Court to be satisfied with evidence which, taken as a whole, amounts to speculation.
The state agreed at the hearing on remand that it had the burden to prove a greater than de minimis harm. Although de minimis is not a term subject to precise definition, it has appeared in constitutional law decisions analyzing whether legislative reapportionment plans meet Fourteenth Amendment standards for protecting the one-person, one-vote principle. The Supreme Court has held that courts designing reapportionment plans may not allow more than a de minimis difference in population size between voter districts with the greatest and the least population. Chapman v. Meier, 420 U.S. 1, 26-27, 95 S.Ct. 751, 766, 42 L.Ed.2d 766 (1975). In redistricting plans devised by state legislatures for state elections, deviations under 10 percent have been "considered to be of prima facie constitutional validity." Connor v. Finch, 431 U.S. 407, 418, 97 S.Ct. 1828, 1835, 52 L.Ed.2d 465 (1977). In Wyche v. Madison Parish Police Jury, 635 F.2d 1131, 1158-1159 (5th Cir. 1981), the Court of Appeals held that, under the stricter standards applied to court-ordered reapportionment plans, an 8.2 percent deviation was too large, but a 4.11 percent deviation was sufficiently de minimis. Although this Court adopts neither 10 percent nor 4 percent as a definition of de minimis, the prior use of those figures in the context of determining constitutional violations supports the view that de minimis is meant to describe an amount which is small, but not infinitesimal.
"De minimis" is a legal term which is not used in the scientific profession. Therefore, the parties' medical experts, in giving their ultimate opinion on the issue posited by the Court of Appeals, have sought to interpret medical findings using a legal term. The medical profession uses terms like "significant", "substantial" and "association" to express the effect or risk of abortions on the incidence of certain medical complications. Effects or risks may be quantified by either an odds risk ( e.g., a woman has a 1 in 100,000 risk of death from an abortion (.001%)) or by expressing the woman's "relative risk" of suffering the occurrence. Relative risk is an expression of the risk ratio of an event with a risk ratio of 1.00 meaning no greater risk; .05 meaning a reduced risk; or 2.0 meaning an enhanced risk of an occurrence because of a given event. (Testimony of Dr. Daling). However, even if a finding shows a slightly higher risk, like 1.31, if the "confidence interval" of the finding includes numbers showing a decreased risk ( e.g., .09) then one can't say that the occurrence didn't happen by chance and, therefore, the finding is consistent with no association at all and the risk is not statistically significant. ( See, e.g., Plaintiff's Exhibits 2 and 21; testimony of Dr. Daling).
Accordingly, pursuant to the remand of the Court of Appeals and Rule 52, Federal Rules of Civil Procedure, the Court adopts and affirms the findings and conclusions of its December 13, 1979 Memorandum Opinion, as modified by the Fifth Circuit Court of Appeals, and as herein supplemented by the following Findings of Fact and Conclusions of Law based upon the record of its July 29 and 30, 1982 and August 6, 1982, evidentiary hearings on remand, and the September 14, 1979 trial and July 10, 1979 preliminary injunction proceedings. Based upon this review and for the reasons discussed below, the Court finds that the Florida Legislature could not have reasonably concluded from the evidence adduced in this case that the abortion procedure, as performed properly by licensed, medical practitioners according to methodology approved by prevailing medical authority and generally in use in Florida, poses a greater than de minimis risk to a married woman's future ability to bear children; that is, it does not pose such a risk. Therefore, Fla.Stat.Ann. § 390.001(4)(b) on spousal notice is constitutionally defective.
FINDINGS OF FACT
The Court initially finds that certain facts are uncontested by the parties. The method of abortion which is approved and generally in use in Florida is that of vacuum aspiration (VA), which is also referred to as suction curettage. Testimony from expert witnesses on both sides agreed that the methodology used in Florida was comparable to that described in national data compilations. Nationwide statistics reveal that more than 90% of all abortions are performed by vacuum aspiration or suction curettage. The experts also agreed that there is no reason to believe that the gestational age at which abortions are performed in Florida differs from the national average. Over 50% of all abortions are performed within the first eight weeks of pregnancy, and at least 90% are done within the first trimester. Evidence also revealed that induced abortion is one of the safest operations performed on women with fewer than 1 woman per 100 suffering a major complication. According to Plaintiff's Exhibit 16, which is a study issued by staff members of the Bureau of Epidemiology, Center for Disease Control, Atlanta, Georgia, at page 189, "the death-to-case rate of 3.7 deaths per 100,000 abortions is lower than that of tonsilectomy" in the United States. Defendants' chief witness, Dr. Denis Cavanaugh, agreed that there is no data which suggests that an abortion properly performed, without complications, poses a greater than de minimis risk to a woman's future childbearing ability. Therefore, the Court's focus is on whether the abortion procedure, because of the complications caused by some abortions, poses a greater than de minimis risk.
The Court is only concerned with the methods and risk associated with "induced" abortions, those initiated with the intention to terminate a pregnancy, and not with "spontaneous" abortions, generally defined as all other abortions. ( See Plaintiff's Exhibit 10, p. 1).
Vacuum aspiration or suction curettage is the procedure in which the uterus is evacuated with a flexible or inflexible suction cannula or tube under negative pressure (Defendants' Exhibit 3, p. 475).
Both Dr. Denis Cavanaugh, an expert for the Defendants, and Dr. Christopher Tietze, an expert for the Plaintiff, agreed on this fact. Florida itself compiles no statistics on how many abortions are done by a particular method or in which week of pregnancy.
Testimony of Dr. Tietze. According to Plaintiff Dr. Scheinberg, qualified as an expert witness, these approximate percentages reflect at least the Florida percentages in 1979, the time at which this statute was passed. In fact, Dr. Scheinberg opined that probably all first trimester abortions in Florida are by VA or suction curettage. Data from the Centers for Disease Control reveals that over 97% of all abortions performed in the United States in the first trimester are by suction curettage.
Testimony of Dr. Tietze or Dr. Cavanaugh.
Testimony of Dr. Tietze; Plaintiff's Exhibit No. 14 and No. 15 at 41, table 12.
"Abortion is a safe surgical procedure, as documented in numerous large studies." Plaintiff's Exhibit No. 16, p. 177; "Induced abortion is one of the most frequently performed operations on women — and one of the safest." Plaintiff's Exhibit 16, p. 189.
Arguably, a "properly performed" abortion is one without medical complications. However, neither the Court nor Plaintiff's witness, Dr. Tietze, have taken such a restrictive view; a properly performed abortion is not necessarily an error-free abortion. Instead, a "properly performed" abortion is done with a level of care, skill and treatment which is recognized by reasonably prudent similar abortionists as being acceptable. ( See generally, Fla.Stat.Ann. § 768.45(1)).
To the extent that statistics on the risks associated with abortions will inevitably contain complications and adverse effects caused by improperly performed abortions, all of the studies and literature presented to the Court are biased in an unquantifiable amount in favor of the Defendants and towards a finding of increased risk.
(a) The method used and (b) the gestational age of the pregnancy have a major impact on whether there are later adverse effects from an induced abortion. Both of the experts who testified for the state at the hearing on remand agreed that these two factors are the most important determinants of future complications. Medical testimony indicated that the widespread use of suction, or vacuum aspiration, is far safer to the pregnant woman and carries significantly less risk of injury from laceration, perforation or hemorrhage than does the now outmoded method of sharp curettage. The choice of method is related to gestational age, in that suction curettage is used primarily during the first trimester of pregnancy and, generally, earlier abortions are medically simpler procedures. The testimony of all of the expert witnesses leaves no doubt that the earlier in pregnancy an abortion is performed, the safer the procedure is and the more minimal the risk of any significant sequelae. Consequently, the fact that 9 out of every 10 abortions in Florida are performed in the first trimester, and the fact that over 97% of these first trimester abortions are performed by suction curettage, leads to the conclusion, borne out by the testimony and exhibits, that risk of harm to the woman patient for an induced abortion in Florida is extremely slight.
Testimony of Dr. Cavanaugh and Dr. Matthew J. Bulfin. See also, Plaintiff's Exhibit 16, p. 1.
Testimony of Dr. Scheinberg and Dr. Tietze.
The rate for all major complications from abortions by suction curettage, even those done in the second trimester, is only 0.4 percent. (Plaintiff's Exhibit 16, p. 178, table 2). The rate for all abortions, performed by all methods, including sharp curettage, installation of an abortifacient, hysterotomy and hysterectomy, is only 0.7 percent. (Plaintiff's Exhibit 16, p. 177). Also, not all major complications have an adverse effect on a woman's future childbearing potential.
The state attempted at the remand hearing to meet its burden of proof to show a greater than de minimis risk, by relying on the testimony of two physicians: Dr. Cavanaugh and Dr. Matthew Bulfin. Dr. Cavanaugh agreed that an induced abortion performed during the first eight weeks of pregnancy, when over 50% of all abortions are performed, does not pose a greater than de minimis risk to a woman's future childbearing capacity. However, Dr. Cavanaugh believes that abortions in the eight-to-twelve week period and in the second and third trimester do pose a risk which is significant. Dr. Cavanaugh's opinion that abortions performed after eight weeks pose a significant risk to future childbearing was based primarily on medical studies of the sequela of abortions performed in other countries, studies which rely upon methodology now outmoded and not generally in use in Florida, and those which include data ten (10) years old or older.
Testimony of Dr. Cavanaugh; Cavanaugh deposition pp. 44-45, 55: "I think it would be common medical sense to say that up to 8 weeks the death rate is minimal and the complication rate is minimal." Centers for Disease Control statistics reveal that 99 percent of abortions done at or before 8 weeks is by suction curettage.
Dr. Cavanaugh has never done any studies of his own on induced abortions and its effect on childbearing. Similarly, he has never published any articles analyzing the studies or results of others' research.
The Court finds that the studies relied upon by Dr. Cavanaugh to support his conclusion suffer from major flaws which severely weaken the reliability and credibility of the studies and the concomitant credibility of the doctor's conclusions. Defendants' exhibits numbered 12, 14 and 16 are based on studies of illegal abortions in foreign countries over ten years ago and are of little, if any, value towards determination of the issue on remand; exhibits 3, 7 and 10 also are based on foreign data. There was no analysis of the abortion method used or of gestational age, the predominant factors influencing the morbidity of abortions, in Defendants' exhibits 7, 10, 14 and 16, and no analysis of the method of abortion in exhibit 17. The data in exhibits 3 and 5 was gathered prior to 1973, when the methodology in use was often times dissimilar from that currently employed and before the Supreme Court decision in Roe v. Wade, 410 U.S. 113, 93 S.Ct. 705, 35 L.Ed.2d 147 (1973), declared restrictive abortion laws unconstitutional; the relevant dates are unknown for the data in exhibits 7 and 10. Defendants' exhibits 2, 3, 4, 5 and 6 state generalized complication rates with no data breakdown or separate discussion as to individual complications, nor do they state conclusions as to whether the complications would affect childbearing. Exhibits 7, 18 and 19 are completely anecdotal in nature, with no use of a control group or cross tabulation for the effects of other variables. Dr. Bulfin primarily testified about an anecdotal study he conducted based on patients seen in his private practice. Such anecdotal evidence, as compared to well controlled scientific studies based upon control groups and cross tabulation, appears to be an unreliable means of determining the issue before the Court. See Weinberger v. Hynson, Westcott and Dunning, Inc., 412 U.S. 609, 93 S.Ct. 2469, 37 L.Ed.2d 207 (1973). The Defendants' third witness on this issue, Dr. Leslie Iffy, who testified at the 1979 trial, explicitly based his testimony on data from Hungary, which reflected a methodology and health care system very different from that in use in Florida at the time the statute in question was enacted. The Court concludes that the studies relied upon by the Defendants, therefore, are either not reliable or credible or offer little assistance in determining the effect on childbearing of an abortion as "performed properly by licensed medical practitioners according to methodology approved by prevailing medical authority and generally in use in Florida."
Dr. Cavanaugh failed to include the most important American studies and instead sought to highlight those foreign studies with adverse outcomes or those American studies, the result of which have not been substantiated or duplicated by other studies. As a result, the Court found Dr. Cavanaugh's overview of the sequela of abortions to be unscientific and biased and selective and, therefore, not a fair, accurate or reliable review of the literature.
Defendants' Exhibit 13, which disclosed a 0.5% increased risk of placenta previa from an induced abortion was based upon a small sample and failed to control for whether the women studied were smokers, a characteristic previously found to be associated with an increased risk of placenta previa. More significantly, the patients studied were only asked about a history of prior abortions after the onset of a complication. Dr. Cavanaugh acknowledged that selective recall in such a situation would bias the result. In fact, several of the studies introduced by Defendants could be cited in support of the argument that no greater than a de minimis risk exists. Exhibit 11 was a World Health Organization study which found no enhanced risks from abortions in cities in which vacuum aspiration was the primary method used. Exhibit 9 reflects no increased risk of spontaneous abortion from one prior abortion but a greater risk from multiple prior abortions. Its authors, however, specifically state that their results "do not indicate that these women ultimately are unable to achieve a successful pregnancy or that they are unable to achieve their desired family size." Exhibit 9 at 2499. Exhibit 17 finds that induced abortion does not lead to an increased risk of placenta previa, abruptio placenta, incompetent cervix, ectopic pregnancy, spontaneous abortion, congenital anomalies, or low birth weight infants. Exhibit 17 at 519. Although finding a statistically significant relationship between prior abortion and subsequent poor pregnancy outcome, the authors note that "the difference in outcome disappeared" when they controlled for marital and economic status, contraceptive use, and smoking. Exhibit 17 at 519 and Table III.
Dr. Bulfin's data on 802 of his patients between 1972 and 1978, some of whom had abortions outside of Florida, did not compare the complications, which he diagnosed or which were related to him, to any control group of women who had not received abortions. Nor were any of the conditions he found cross-tabulated with such variables as sexually transmitted disease, method of birth control, abortion method, or other factors which could have significantly altered the findings. The Court finds that this study was not conducted in a way which indicates scientific reliability. Lastly, the Court notes that even Dr. Bulfin's data indicated only a 1.7 percent incidence of infertility, which itself can be considered de minimis in light of the natural incidence of infertility in the population. Testimony of Dr. Tietze and Dr. Cavanaugh concurring.
The literature primarily relied upon by Dr. Iffy was a Hungarian study, which collected data from 1953 onward. [Transcript of September, 1979 trial, p. 201.] Dr. Iffy related that Hungarian abortions were performed with "ovum forcepts" [T. 221-222], and that in Hungary, "sharp curettage is being used with any abortion" [T. 221]. The vacuum or suction curettage method was only beginning to be recognized in Hungary in 1979, according to Dr. Iffy. [T. 233].
At the remand hearing, the Plaintiff presented testimony from two experts in the field of reproductive epidemiology. Dr. Christopher Tatze, an internationally recognized expert in the field and the author of approximately 100 articles specifically about abortion, testified on the basis of his own research and his published analysis of the medical literature. Dr. Janet Daling described her published research findings on the issue of whether abortion affects fertility or other subsequent pregnancy outcome. Both concluded that induced abortion does not have a greater than de minimis impact on future childbearing capacity. The Plaintiff, Dr. Scheinberg, a Florida physician who has performed several thousand abortions, testified at the remand hearing that he had encountered or observed no significant complications from first trimester abortions performed by himself or other physicians.
Epidemiology is the study of disease in human populations. Epidemiologists analyze the occurrence of disease or other poor health conditions in representative population samples. Such studies provide the basis for scientific and statistically verifiable conclusions about the risk of future disease or injury. In determining the extent of the risk of injury resulting from a given medical procedure, the methods of data analysis utilized by epidemiologists provide the most reliable guidance to the finder of fact.
Concerning Defendants' witnesses, Dr. Cavanaugh testified that he had performed a total of approximately 10 medically necessary abortions in his career; Dr. Bulfin has never performed an induced abortion. Dr. Iffy is not a Florida physician.
The Court, in weighing the testimony of Defendants' experts, is not questioning the expertise of the Defendants' witnesses in their respective fields. However, the lack of experience and training of the Defendants' witnesses in both the area of abortion sequela and the performance of abortions in Florida, and the reliance of Defendants' witnesses on poorly controlled studies, dated studies and foreign studies, make their testimony and opinions less credible than those offered by Plaintiff's experts. Therefore, in the test of credibility, the Court finds the Plaintiff's experts, with both experience and training in abortion sequela, to be more credible.
At the 1979 trial, Plaintiff adduced testimony from two other practicing Florida physicians. Dr. Samuel Barr, the director of a Winter Park, Florida, clinic in which abortions were performed, stated his view that "99 percent of the time, absolutely no adverse effect upon the woman or the childbearing capacity" occurred. Raphael Good, one of four physicians in the United States board certified as both an obstetrician-gynecologist and a psychiatrist and then a faculty member at the University of Miami School of Medicine, had also performed induced abortions in Florida. He stated his opinion that if a woman has a termination procedure done by a licensed physician in an approved facility the chances of harm to her future fertility were "very slim."
Transcript, September 14, 1979, p. 41.
Id., p. 71. Good noted that there is no clear data that an early abortion affects future fertility. He explained that while there are some statistics which tend to show some effect, there are other uncontrolled complicating factors in the statistics which derogate any results suggesting such an adverse effect.
The Court finds, from the testimony of experts and its review of the exhibits, that there are four ways in which an induced abortion could possibly cause a woman to be unable or significantly less able to bear a child. The first is obviously if she were to die from complications of an abortion. The risk of death from all abortions, however, is only 1 in 100,000, and the risk from first trimester abortions is approximately .5 per 100,000. The second possible way would be sterility resulting from an unintended hysterectomy or loss of uterus. The best available data indicates that this risk is from 1 in 5000 to 1 in 10,000. Thirdly, if major infection resulted from an abortion, it could lead to the condition of pelvic inflammatory disease (PID), which in turn might produce occlusion of both fallopian tubes. Available data as to this possibility indicates that the risk is approximately 1 in 1200 to 1 in 1400. Lastly, if Rh blood-type testing were not performed on abortion patients and Rhogam (an applicable drug) not administered, Rh iso-immunization might result, endangering the life of future children. However, such testing and treatment is, as a practical matter, universal among those providing abortion in Florida. Thus, the risk of Rh iso-immunization is for all practical purposes nil.
Testimony of Dr. Tietze. The Centers for Disease Control found a death rate of only .6 per 100,000 for all abortions in 1978. (Plaintiff's Exhibit 15). "The death-to-case rate for legal abortions is remarkably low, lower than any other surgical procedure." (Plaintiff's Exhibit 16, p. 189). It is of some interest to note that, by comparison, the risk of death from carrying a pregnancy to term is 10 per 100,000. (Testimony of Dr. Tietze).
Testimony of Dr. Tietze. Dr. Tietze also noted that the risk of unintended hysterectomy from a ruptured uterus during childbirth is approximately 1 in 1500. Defendants relied upon a finding by Stewart and Goldstein of a .7% risk of uterine perforation to support their contention of a greater than de minimis risk. However, that study reflects data from 1968-1970 and, more importantly, other subsequent studies more accurately reflecting current abortion methodology and morbidity, have not borne out the Stewart and Goldstein results. ( See, for example, Defendants' Exhibit 4 and Plaintiff's Exhibit 8). Dr. Scheinberg testified that he had never seen an incident of perforation with a first trimester abortion. The Court accepts as more credible and more reliable the expert opinion of Dr. Tietze.
Testimony of Dr. Tietze. Although there are some American studies on the incidence of pelvic inflammatory disease from an induced abortion, unfortunately, no satisfactory studies have been done in the United States on the linkage between PID and subsequent tubal occlusion. The best data is from a Swedish study which found that only about 15% of the women who contract PID, with or without an abortion, become infertile. Therefore, any rate for pelvic inflammatory disease must be reduced to 15% in order to determine the influence on bilateral tubal occlusion. Testimony also revealed that bilateral tubal occlusion may on occasion be treated successfully and the woman restored to her childbearing capacity. Therefore, the risk to future childbearing from PID is actually even less than 1 in 1200 to 1400, since only 15% of those affected with PID suffer a loss of fertility and since such loss of fertility may be treated successfully.
Further, Dr. Scheinberg testified that although there may be some mild infection easily treatable with antibiotics from a first trimester abortion, he has never seen a significant infection such as PID.
Testimony of Dr. Tietze and Dr. Scheinberg.
In addition to the data on each particular manner by which future fertility can be impaired, there have been published studies on the general question of the relationship of induced abortion to infertility. Of these, only one published study has focused on American women. Using data gathered from 1976 to 1978, it found no statistically significant relationship. A second, just-completed study of American women, prepared by physicians at Harvard University and awaiting publication, reaches the same conclusion that induced abortion does not lead to reduced fertility. Even among studies of women in foreign countries, all but one have found no relationship.
Infertility is the inability to become pregnant after twelve months of unprotected intercourse. (Testimony of Dr. Daling).
Plaintiff's Exhibit 2; testimony of Dr. Daling and Dr. Tietze. Statistical significance is a term used to describe whether a particular result can be attributed to the hypothesized causal factor being studied, in this case abortion, or whether the outcome is equally attributable to chance. See n. 10, supra. If the confidence interval for relative risk extends both above and below 1.0, the results are inconclusive as to whether women with prior abortions are at greater risk than women who did not have an abortion. See, e.g., Plaintiff's Exhibit 21.
Plaintiff's Exhibit 3. This study, with a much larger control group than in the study in Plaintiff's Exhibit 2, also concluded that the results of Daling were correct in concluding that secondary infertility is not increased by previous induced abortion.
The lone contrary study (Defendants' Exhibit 8), from Greece, included illegal abortions and reflected data which may have been computed incorrectly. (Testimony of Dr. Daling).
Since there is a natural incidence of infertility in the population, determined primarily by aging, any woman, including one with no history of abortions, has a chance of becoming infertile as she grows older, with the risk of infertility increasing with age until it reaches 100 percent at the time of menopause. No credible evidence was presented to the Court, however, which demonstrates that the risk of infertility is increased by a legal induced abortion as currently performed in Florida.
Testimony of Dr. Tietze.
The State based its case largely upon theory that abortions increase the risk of certain conditions related to reproduction which, although not causing infertility, may impact on whether a future pregnancy results in a "healthy" child. The State evidence sought to show that a legal induced abortion may increase the risk of spontaneous abortion, premature birth, low birth weight infants and placenta previa. Plaintiff responded with other studies which have found no increased risk of these conditions and with evidence showing that reputable medical studies have been unable to conclude that any relationship exists between a legal induced abortion and any of these conditions. While one study which was presented to the Court found a relationship between induced abortion and later spontaneous abortions, other studies have not. In fact, most studies have shown a decreased risk of spontaneous abortions, although none of the findings were statistically significant. Therefore, although the Court cannot conclude that there is a reduced risk, no studies permit the Court, or the legislature, to reasonably conclude that there is a greater than de minimis risk of spontaneous abortion for women who have had an induced abortion.
The Fifth Circuit opinion is unclear as to whether the issue of "healthy children" is a proper inquiry for this Court. The opinion speaks of "risk of a decrease in fertility," "impact on the marriage's child-bearing capacity," "affect adversely a woman's future procreative ability" and "risk to the procreative potential of a marriage," but stated that the particular issue on remand is the "risk to a married woman's future ability to bear children." Scheinberg, supra, 659 F.2d at 486-487. However, the Court agrees with the Plaintiff and has addressed the issue of "healthy" children.
Plaintiff's Exhibit 21; Plaintiff's Exhibit 13, Table 4.
Id.; testimony of Dr. Tietze.
No reliable data was presented which shows any increased risk of premature birth or short gestation. Similarly, the evidence introduced at trial does not indicate an increased risk of a low birth weight child. Therefore, the Court concludes that shortened gestation and low birth weight in pregnancies after an induced abortion has an incidence similar to that in first pregnancies and, hence, there is insufficient credible evidence to reasonably conclude that abortion creates an increased risk of greater than de minimis proportions. And, finally, the evidence does not permit the Court to find a greater than de minimis risk of placenta previa from a prior induced abortion. With the exception of one study, of questionable reliability, showing a 0.5% increase, the evidence does not support a finding of an increased risk of placenta previa.
Of seven studies, four found no association, two American studies found a decreased risk and one Israeli study found an increased risk. (Testimony of Dr. Tietze; Plaintiff's Exhibit 21). The Israeli study is not reliable, however, since it was done in 1966-1968 when abortions were illegal and does not state the method of abortion nor the gestational age at which the abortion occurred.
Of four American studies on low birth weight, one showed no association, one showed a slightly increased risk and two showed a slightly decreased risk. (Plaintiff's Exhibit 21). The findings do not establish any relationship between an antecedent abortion, particularly by vacuum aspiration, and a subsequent low birth weight child.
Plaintiff's Exhibit 5; Plaintiff's Exhibit 13, p. 23.
Placenta previa occurs when the placenta is found in the lower rather than the upper part of the uterus, resulting in bleeding and premature delivery with concomitant increases in still birth and increased parinatal and maternal mortality. (Testimony of Dr. Cavanaugh).
Defendants' Exhibit 13. The credibility of this single finding is harmed by its failure to consider the important confounding variables of smoking, neonatal age and prior pregnancies. The study also does not show natal deaths nor any effect on future pregnancies.
Three other controlled studies introduced into evidence did not show any significant difference in the incidence of placenta previa: Defendants' Exhibit 14, Table 4; Defendants' Exhibit 17 at 519; Plaintiff's Exhibit 5. This view was also supported by the testimony of Dr. Tietze and Dr. Scheinberg.
Both Plaintiff's and Defendants' witnesses agreed that the conditions associated with spontaneous abortions, premature birth, low birth weight infants and placenta previa, if they occur at all as a result of an induced abortion, may not have any lasting negative effect on either the mother or her child. Although the immediate outcome of a pregnancy is adverse, this does not compel the conclusion that the child is unhealthy. Premature babies and those with low birth weight often go on to become healthy children. Studies have not found an elevated risk of infant morbidity, or of prenatal, neonatal or postnatal morbidity from an induced abortion. Furthermore, a woman who has suffered from a spontaneous abortion, placenta previa, low birth weight or short gestation may be treated and, once treated, should be able to have future childbearing success. All of the studies discussed by witnesses for either side related only to one pregnancy next following an induced abortion; no study has indicated any effect on a second or third pregnancy after abortion. Therefore, studies on adverse pregnancy outcomes do not indicate that those women ultimately are unable to achieve their desired family size. Because the State introduced no evidence which could quantify the risk of an abortion causing permanent injuries to either mother or child, if the abortion were performed properly according to prevailing medical practices in Florida, the studies on adverse pregnancy outcome are helpful, but are not dispositive of the issue of the adverse impact of abortion on the procreative potential of a couple. The Court concludes that the evidence does not support a reasonable conclusion that risks to the birth of healthy children exist from undergoing an induced abortion. The risk of problem pregnancies resulting from abortions are highly speculative at most, both as to whether they exist at all and, if they do exist, whether they have more than a negligible impact on future childbearing or future healthy children.
Testimony of Dr. Scheinberg.
Plaintiff's Exhibit 13, p. 20; Plaintiff's Exhibit 5.
Testimony of Dr. Scheinberg.
Id.; Defendants' Exhibit 9, p. 2499.
The Mandate to this Court from the Court of Appeals was to determine whether the legislature could have concluded reasonably, based on the evidence in the record of this case, that properly performed abortions in Florida create a greater than de minimis risk to a married woman's future ability to bear children. The Court finds, based on the evidence before it, and giving due deference to the legislature's ability to consider the same evidence presented to this Court, that the Florida Legislature could not have reasonably arrived at that conclusion. The evidence in the record clearly indicates that any such risk is a de minimis risk.
The Court notes that both Dr. Tietze and Dr. Daling testified that in their professional opinions, no reasonable person could conclude that legal induced abortions pose a greater than de minimis risk to a woman's future ability to bear children.
CONCLUSIONS OF LAW
When a state regulation implicates directly a woman's abortion decision, the Supreme Court has held that "regulation limiting these rights may be justified only by a `compelling state interest' . . . . and that legislative enactment must be narrowly drawn to express only the legitimate state interests at stake." (citations omitted). Roe v. Wade, 410 U.S. 113, 155, 93 S.Ct. 705, 728, 35 L.Ed.2d 137 (1973). The Court of Appeals determined that the state interest in protecting a husband's interest in the procreative potential of the marriage is sufficiently weighty to justify the burden imposed by the spousal notification requirement on a woman's abortion decision. Scheinberg, supra, 659 F.2d 476. However, in addition to furthering a recognized state interest, the spousal notice provision must be drawn narrowly enough in furtherance of that issue to avoid constitutionally offensive overbreadth. Scheinberg, supra, 659 F.2d at 486; see also Shelton v. Tucker, 364 U.S. 479, 488, 81 S.Ct. 247, 252, 5 L.Ed.2d 231 (1960); Griswold v. Connecticut, 381 U.S. 479, 504, 85 S.Ct. 1678, 1692, 14 L.Ed.2d 510 (1965); Poe v. Gerstein, 517 F.2d 787, 793 (5th Cir. 1975). Hence, Section 390.001(4)(h) must be narrowly drawn to express only the legitimate state interest at stake.Since the Court of Appeals has held that the notification provision could be constitutional if a greater than de minimis risk to childbearing capacity resulted from the abortion procedure, and that if no such greater than de minimis risk exists the statute is unconstitutional, the role of this Court is limited. Because the Court has found that the Florida Legislature could not have reasonably concluded, from the evidence introduced into the Court's record, that the abortion procedure, as performed properly by licensed medical practitioners according to methodology approved by prevailing medical authority generally in use in Florida, poses a greater than de minimis risk to a married woman's future ability to bear children, the Court must make the correlative finding that Florida Statute § 390.001(4)(b) is unconstitutional under the rationale and authority of Scheinberg v. Smith, 659 F.2d 476 (5th Cir. 1981).
In addition to a violation of a woman's right of privacy, including the right to choose abortions, the Plaintiff has alleged violations of First Amendment and marital privacy rights. Plaintiff contends that the state, by compelling married women to speak, has infringed both the First Amendment liberty to refrain from speaking and the Fourteenth Amendment right to familial privacy, which restrains state interference in private marital matters. These constitutional claims were not reached either in the original opinion of this Court, 482 F. Supp. at 540, n. 60, or by the Court of Appeals.
Accordingly, a Final Judgment shall be entered herein under even date herewith in accordance with this Memorandum Opinion and the Findings of Fact and Conclusions of Law reached. The Court retains jurisdiction for the entry of an Order determining Plaintiff's attorneys' fees pursuant to 42 U.S.C. § 1988.