The Complexity of Compiling Statistics

JACK C. SMITH, MS

SINCE LEGAL 'INDUCED ABORTION emerged as a the January 1973 Supreme Court decision concerning medical procedure, States have rapidly passed laws the Texas and Georgia (ALI) laws (2). That ruling by liberalizing abortion-the number of legal the Supreme Court invalidated most of the conditions has increased more than twentyfold in 5 years (1). In and restrictions written into the abortion laws of States the United States today, approximately one legal abor- that had reformed legislation. Furthermore, the Court's tion is reported for every five live births (1). ruling also invalidated most State abortion laws, As an adjunct to new State legislation on abortion, whether old or reform, on the grounds that the per- compilation of abortion statistics by the central health missible reasons for abortion-life, health, deformity, agency of the State is usually required. The Center for rape, and incest-were too restrictive. Although report- Disease Control (CDC), in its study of the ing of all abortions was required by Georgia's ALI-type epidemiology of legal induced abortion' in the United law, reporting was not argued as a constitutional issue States since 1969, has been communicating with State before the Supreme Court and therefore was not and local health agencies during their planning, in- responded to in the Court's decision. itiating, and refining of abortion reporting systems ac- The 1973 Supreme Court Decision precipitated a cording to each one's State laws and health regulations. new round of legislative and court actions. These ac- This communication has brought about a broad un- tions ranged from State laws passed in the spirit of the derstanding of the difficulties encountered in compiling Supreme Court's intent to laws passed in direct statewide abortion statistics. defiance of the Supreme Court decision (3). While the pre-1967 restrictive abortion laws were Legal Status similar from State to State, an appraisal of the current The legal status of abortion is the first major issue for situation indicates that legislative and judicial activities consideration. No other kind of information collected in since 1967 have resulted in much dissimilarity of laws. the arena of health statistics is more entangled than As a consequence of the shifting and unstable legal abortion in the legislative and judicial process. status of abortion in the United States, State health Until 1967, when Colorado enacted its reform abor- agencies, in general, continue to have difficulty in defin- tion law, all State laws were essentially the same, per- ing their roles and responsibilities for compiling mitting abortion only to save the life of the pregnant statistics on abortion. woman. From 1967 to 1972, more than a dozen States passed laws that liberalized the circumstances under Reporting Authority which abortions were permissible. For the most part, In States in which an abortion reporting system has these laws followed the American Law Institute (ALI) been implemented, authority for reporting has come in Model , which recommended that abor- one of the four following ways: tions be legally permitted for the following reasons: (a) Establishment of reporting authority as part of reform abortion if the continuance of a pregnancy might cause physical legislation. Reforming abortion legislation usually or mental impairment of the woman, (b) if the child means modifying or rewriting that portion of the might be deformed, or (c) if the pregnancy was the criminal code of the State that deals with circumstances result of rape or incest. Some States, however, passed under which performance of abortion is lawful. laws with no legal restriction on reasons for which an Authority to collect and compile statistics on abortion abortion could be obtained before viability of the fetus. is usually given to State health agencies in a general Reform abortion laws often explicitly stated other manner so that both the data items and the modus conditions which had to be met before an abortion was operandi must be determined by the health officials. performed, such as residency requirements or restric- Some State laws, however, in giving authority also tions on the type of health facility used. These con- prescribe explicitly who should report, what should be ditions and restrictions led to numerous court reported, and how it should be reported. The amount of challenges of various State laws which culminated in input that the central health agency had in drafting OMr. Smith is chief of the Statistical Services Branch, Family reform legislation usually determines whether the more Planning Evaluation Division, Bureau ofEpidemiology, Center explicit wording of the authority is more or less for Disease Control. This paper is based on one he presented at desirable than a general authority. the 102d annual meeting ofthe American Public Health Associa- Reporting authority based on State vital statistics statutes. All tion, New Orleans, October 1974. Tearsheet requests to Jack C. States have vital statistics laws that require reporting of Smith, Statistical Services Branch, Center for Disease Control, vital events, including fetal deaths. However, no unifor- Atlanta, Ga. 30333. mity exists regarding which fetal deaths are to be

502 Public Health Reports reported. Ten States require reporting of all fetal procedure on a pregnant woman. The health communi- deaths, which means any pregnancy termination before ty is faced with the responsibilities that logically relate delivery of a live-born infant, while the remaining to both the fetus and the pregnant woman. The point of States set criteria on reporting, which conflict is seen when a physician performs a legal usually excludes reporting of fetal deaths of less than 20 medical procedure during the first trimester of weeks' gestation. pregnancy and then is required to report that Most States that require reporting of all fetal deaths procedure by certifying a fetal death. There appears to use their vital statistics authority to include reporting of be a growing consensus that abortion performed before legal induced abortions. However, a major problem viability should be viewed as a medical procedure and with use of the vital statistics statute as a basis for ob- that the responsibility of the health agency is to the taining abortion data is certification. Laws governing woman; therefore, reporting a procedure rather than a vital statistics, for the most part, require that legal cer- fetal death is more appropriate. tification of an event be filed permanently with the cen- Another issue related to the question of what con- tral health agency. Cooperation and compliance with stitutes an abortion concerns the procedure called reporting provisions of the law have usually proved "menstrual extraction. " Since the risk of morbidity and more difficult to gain in data collection systems requir- mortality associated with-abortion increases with dura- ing a certificate as opposed to a non-certified statistical tion of pregnancy, there is a continuing effort to per- reporting form. Furthermore, the recommended stand- form abortions as early in gestation as possible. ard fetal death certificate was developed before the ad- Menstrual extractions are performed before pregnancy vent of legal induced abortion and does not lend itself to can be confirmed by means of standard pregnancy the collection of pertinent statistics on induced abor- tests. Thus, only after the menstrual extraction has tion. Some States have dealt with both issues by been done and the contents of the examined modifying their vital statistics statutes to maximize pathologically can pregnancy be verified. Technically, authority and optimize reporting. menstrual extraction cannot produce an abortion in the absence of a pregnancy; therefore, it is not correct to Reporting authority given by health regulations. In some definitionally equate an abortion with the performance States the health agency gathers abortion information of an abortion procedure. Consequently, just as the under an existing or newly established regulation. health statisticians have always had to be aware of Authority by regulation usually occurs only in States definitional problems, they must be alert to definitional that have neither of the two legislative authorizations problems that arise in relation to abortion statistics. mentioned previously. Two problems have been observed in States that rely on the regulatory power of the Use of Abortion Data health agency for obtaining abortion statistics. First, Finally, a word regarding abortion data after they have the health agency is often reluctant to put teeth into a been collected by the central health agency. Most regulation to collect abortion data when the legality of States have made little effort to fully use abortion abortion is not legislatively clear. Second, the coverage statistics. Abortions can be used as an indicator of un- of reporting under regulations may not be broad wanted pregnancies; thus, family planning personnel enough to include the free-standing abortion clinics, could use abortion statistics in planning and directing which are established outside the usual health care their programs to those population groups which have delivery system, and abortions performed in the the highest rate of abortions. Since abortions are direct- physicians' offices. ly related to , abortion statistics could be incor- Voluntary reporting. The States that have an operative porated into analyses of general fertility trends and voluntary reporting system usually have key persons in population growth estimates. Maternal morbidity and the central health agency who invest much time in mortality associated with childbearing should be in- soliciting the cooperation of various medical groups fluenced by the practice of abortion as should perinatal throughout the State. Certainly the voluntary reporting and infant mortality. approach-albeit better than none-is fraught with dif- Also, perhaps because of the sensitive nature of abor- ficulties, and the health agency usually seeks legislative tion, States often do not publish a periodic statistical or regulatory authority for reporting as soon as it is report on abortion or include abortion statistics in other feasible. established statistical reports. It is hoped that there will soon be more widespread dissemination and use of abortion data in each State. What is an Abortion? The question, "What is an abortion?" leads to another References necessitates of old es- complex area that rethinking and 1. Center for Disease Control: Abortion surveillance; 1973. Atlanta, tablished definitions. A fundamental issue is whether Ga., May 1975. abortion should be thought of in terms of the fetus or 2. Roe v. Wade, 410 U.S. 113 (1973); and Doe v. Bolton, 410 U.S. 179 the pregnant woman. Vital statistics laws relate to the (1973). fetus as a product of conception. Abortion laws relate, 3. Seven States adopt new abortion laws following Supreme Court for the most part, to the performance of a medical decision. Fain Plan Pop Rep 2:47, June 1973.

November-December 1975, Vol. 90, No. 6 503