Oregon Vital Statistics Annual Report 1995, Volume 1
Technical Notes — Methodology
INDUCED TERMINATIONS OF PREGNANCY
Except for incomplete reporting by providers, the data represent
abortions performed in Oregon during the current data year. That is, the data constitute events associated with the place of occurrence rather than the "residence data" used in estimating births. This is necessary because many abortions obtained out-of-state by Oregon residents are not reported to Oregon's Center for Health Statistics. It reflects the great variation in abortion reporting procedures among states (e.g., some states do not record the patient's residence) as well as the fact that a comprehensive data collection network among all states, similar to that used in reporting births, does not exist in regard to abortions.
In using "occurrence" data rather than "residence" data to estimate abortion rates for Oregon residents, an implicit assumption is made that the number of Oregon residents who leave the state to obtain an abortion equals the number of out-of-state residents who obtain an abortion in Oregon. In formulating generalizations which involve trends or long-term behavioral patterns, annual totals are treated as sample values generated by ongoing social, economic, or political processes and thus subject to "chance" variability. For most purposes, numbers offered in this report should be viewed only as careful approximations and interpreted only within the framework of statistical safeguards developed to take sampling variability into account.
Some rates in this section are based on relatively
few events and for most comparisons may be used only with extreme caution--due to the chance fluctuations associated with small numbers. A small percentage of abortion reports lack certain data items. This may greatly affect the estimation of rates. To minimize the potential bias inherent in such estimates, unknown events are assigned to the categories of analysis proportional to the distribution of known events. In this way, rates calculated for subsets (e.g., "abortions per thousand black females") are, on average, less affected by incomplete data.
ESTIMATION OF THE CUMULATIVE PROPORTION OF FEMALES WHO HAVE EXPERIENCED AN ABORTION
This figure is estimated by tracing the abortion experience of a specific cohort of females over an extended time period. In the table on the previous page, an approximation of the "cumulative total" of first-time abortions by one of the cohorts may be obtained by summing the figures in the boxed area.
To obtain this value, it is necessary to sum the number of first-time abortions for 15- to 19-year-olds from 1975 to 1979 and those of 20- to 24-year-olds from 1980 to 1984 with those of 25- to 29-year-olds from 1985 to 1989. This provides an estimate of the numerator in the following equation:
The denominator may be estimated by averaging the size of the cohort during 1975-1989. Table 10-1 lists the annual estimate of the number of females within each cohort. For example, in 1975 the number of 15- to 19-year-old females was estimated to be 110,334; in the next year it was 111,184. The average size of this age group from 1975 to 1979 was 112,047. Similarly, the number of 20- to 24-year-old women between 1980 and 1984 was 114,553 on average; the number of 25- to 29-year-olds averaged 111,724 between 1985 and 1989. Thus, between 1975 and 1989 the cohort of interest had an average population size of 112,775.
Substituting into the formula given above:
This figure approximates the proportion of females in the 25- to 29-year-old cohort who, by 1989, had
ever had an abortion. This method of estimation assumes that factors such as deaths and migration have not altered the composition of the female population in Oregon--that is, the women who have left the state display the same characteristics as those who have moved into Oregon. It also assumes that patients with a history of previous abortions do not report the current procedure as a first abortion.
Pregnancy estimates are based upon the estimated number of teen births and induced terminations among Oregon teens; they do not include the number of fetal deaths or miscarriages (spontaneous abortions) which occur. The estimation of teen births is considered to be relatively complete and includes births to resident teens even when they occur out-of-state. The estimation of teen abortions is based on all reported abortions to teen age residents of Oregon; however, because states often do not report abortions obtained within their borders to the state of residence as occurs with vital events such as birth and death, an unknown number of Oregon teens obtain abortion services out-of-state. As a consequence, estimates of teen abortions and teen pregnancies should be considered minimal in nature.
Furthermore, because estimates of abortion for teens are based on "residence data," figures given in Chapter 4 do not correspond exactly to those in Chapter 3, which are based on "occurrence data." (See Induced Terminations of Pregnancy methodology section above.)
The estimation of rates requires an estimate of the size of the appropriate population. Such estimates are now available for 15- to 17-year-olds and 18- to 19-year-olds for each county on an annual basis. Because estimated rates based on a small population may vary greatly due to chance factors, rates of teen pregnancy, birth, and abortion were calculated for these age groups only if there were 100 or more female residents of the appropriate age group in the county.
Similarly, rates for 15- to 19-year-olds were calculated whenever a county had 200 or more female residents in this age group.
Great caution must be taken in the use of pregnancy statistics associated with females under 15 years of age. This is due to the fact that relatively few events are recorded each year for this group. Also, rates are based on the estimated population cohort of 10-14 year old females—many of whom are physiologically not yet at risk of pregnancy. Thus, any
direct comparison of rates between this group and another age group—e.g., 15- to 17-year-olds—would be inappropriate.
The extent to which Oregon's demographic composition may affect its national ranking is indicated by comparisons shown in the sidebar. In 1990—the most recent year for which this comparison is possible—Oregon's birth rate for all teens (regardless of race or ethnic affiliation) was nine percent lower than that of the U.S. and, among all 50 states, it had the 24th lowest teen birth rate. Yet, if comparisons were made in terms of births to non-Hispanic white teens only, Oregon would have been 36th and the rate would have been 19 percent
than that of the U.S. This results from the fact that 87 percent of 15- to 19-year-old females in Oregon were non-Hispanic whites and only seven percent were either Hispanic or non-Hispanic African Americans. By comparison, 70 percent of the U.S. female population of that age were non-Hispanic whites and 26 percent were Hispanics or non-Hispanic African Americans.
ABORTION FACILITY DAMAGED BY FIRE
During April 1992, a fire damaged the facility of a major abortion provider in Jackson County. In the preceding five years the clinic had provided abortion services to residents of more than one-half of Oregon's counties. Still, it accounted for less than seven percent of all abortions performed in the state and could be considered a significant provider of services in only six counties (Jackson, Josephine, Curry, Klamath, Douglas and Coos).
Because about 80 percent of women seeking abortion services in Jackson or Josephine County had obtained them at the clinic which burned, the loss of abortion services might be expected to have major impact there. Because alternate abortion providers in Oregon were less accessible, residents of Curry County appear to have been more affected by the fire than those of Klamath, Douglas or Coos counties.
The number of abortions to residents of Jackson and Josephine Counties nearly dropped to one-half of the average number observed during the previous five years. Such a sharp decrease in the number of unwanted pregnancies during a single year seems unlikely. It is thought that while many women sought an abortion at an alternative provider within the state, others may have traveled to California or Nevada and, as a result, their abortions were not recorded. Comparison of 1992 birth and abortion patterns with those of previous years suggests that as many as 350 women from southwestern Oregon counties may have done this. Because of the level of confidentiality needed in regard to use of abortion services, no way has been found to satisfactorily confirm this.
Return to Table of Contents