Hispanic women are less likely to undergo hysterectomy than are white women, according to an analysis of the 1998-1999 National Health Interview Survey (NHIS).1 Only 12% of Hispanic women aged 25 or older have had the procedure, compared with 23% of white women. The gap between the two groups lessens with increasing acculturation and increased education among Hispanic women. For example, compared with white women who have no high school diploma, Hispanic women who are not high school graduates have about 60% lower odds of hysterectomy, but the odds are reduced by only about 30% among Hispanic women who have any postsecondary education.
The researchers analyzed the use of hysterectomy among Hispanic women because data on this subject have been lacking, largely as a result of the absence of accurate information on ethnicity in national surveys, and because other data point to underuse of medical services among Hispanic women. The NHIS is an annual study that collects health and health care utilization information on a representative sample of the American population. Since 1995, the NHIS has oversampled Hispanic and black households, providing more accurate estimates of health and behavior among these groups, and making possible this kind of analysis.
The researchers based their analyses on respondents' self-reported history of hysterectomy and identification with any of a variety of Hispanic subgroups. To determine women's level of acculturation, they examined whether the interview was conducted in Spanish, English or a combination of the two. They also classified women into three groups: born on foreign soil and living in the United States for fewer than 10 years, born on foreign soil and living in the United States for 10 years or more, and born in the United States.
Hispanic and white women differed significantly both in the proportions who had had a hysterectomy--12% vs. 23%--and in their background characteristics. Sixty percent of Hispanic women, compared with 44% of white women, were aged 25-44; 16% and 11%, respectively, were married. Among Hispanic women, 44% had no high school diploma, 23% had graduated from high school and 33% had some postsecondary education; by contrast, 14% of white women had not completed high school, 33% had graduated and 53% had had some further schooling. Higher proportions of Hispanic than of white women reported having a family income of less than twice the federal poverty level (55% vs. 24%), being in fair or poor health (17% vs. 12%) and having no usual source of medical care (16% vs. 7%).
In initial analyses, the interaction between ethnicity and education emerged as a key predictor of the odds of hysterectomy, a finding that was borne out by the results of logistic regression. Hispanic women without a high school diploma had roughly 60% lower odds of having had a hysterectomy than white women who were not high school graduates (odds ratio, 0.4). The gap closed, however, as education increased among Hispanic women: The odds were reduced by about 40% (odds ratio, 0.6) among those with a high school education and by 30% (odds ratio, 0.7) among those with some schooling beyond high school.
Because of the diverse backgrounds of Hispanic women, the researchers also evaluated the association between hysterectomy and education among women of different national origins. In these analyses, they found that for Cuban or Cuban American women and Puerto Rican women, level of education did not make a difference in the association--odds of hysterectomy were 40% lower among these groups than among white women regardless of whether they had schooling past high school. However, among women who identified themselves as Mexican, Mexican American or Chicano, the relative odds were lower for those who had less than a high school education (odds ratio, 0.4) than for those who had ever attended college (0.7). The pattern was similar for other Latin American women (odds ratios, 0.3 and 0.6, respectively).
Level of acculturation also had strong associations with the odds of hysterectomy. Hispanic women who had lived in the United States for fewer than 10 years were the least likely to have had a hysterectomy (odds ratios, 0.1 for those interviewed only in English and 0.3 when Spanish was used in the interview); those born in the United States had odds much more similar to their white counterparts' (0.7 and 0.5, respectively). As with particular ethnicity, education had a significant association with the likelihood of hysterectomy among Hispanic women when analyzed in conjunction with level of acculturation. Foreign-born Hispanic women with no college education had 60% lower odds of hysterectomy than white women; those with some college had a smaller reduction in odds (40%). No difference by education was found for U.S.-born Hispanic women.
One limitation of the data, the authors note, is that they were collected cross-sectionally, so it is not possible to determine when a woman had her procedure and how much her characteristics have changed since that time. The data set also does not allow for an analysis of parity or age at first birth, factors considered to be important predictors of hysterectomy.
The researchers point out that because Hispanic women had less education, lower incomes and poorer health status than white women--all factors with significant effects on hysterectomy rates--it is uncertain how much ethnicity alone accounts for the differences they found. The association between level of acculturation and hysterectomy, however, indicates that ethnicity may indeed have an independent effect. The researchers postulate that hysterectomy, which generally is used to treat conditions that are not life-threatening, may be overused among white women, rather than underused among Hispanic women. They suggest that the "next step in this research would be to look at the medical necessity of hysterectomy among Hispanic and non-Hispanic women and assess any differences." --D. Feivelson
1. Brett KM and Higgins JA, Hysterectomy prevalence by Hispanic ethnicity: evidence from a national survey, American Journal of Public Health, 2003, 93(2):307-312.