Is a history of trauma associated with a reduced likelihood of cervical cancer screening?
- OBJECTIVE: We tested the hypothesis that a history of trauma (especially sexual trauma) was associated with a reduced likelihood of having had medically appropriate cervical cancer screening.
- STUDY DESIGN: A case-control study using mailed self-report questionnaires.
- POPULATION: The questionnaires were completed by an age-stratified random sample of adult women members of a large health maintenance organization. The sample included 364 women who had received medically appropriate cervical cancer screening and 372 who had not.
- OUTCOMES MEASURED: We defined cases as women who, according to their medical record, had not had cervical cancer screening within 2 years before the study. Controls were defined as women who had been screened. We evaluated exposures to trauma that we hypothesized to be associated with the case/control state.
- RESULTS: Women who had been sexually abused in childhood were less likely to have had a Pap smear within the past 2 years (36.0% vs. 50.4%, P = .050). Other traumatic events were associated with Pap testing in bivariate analyses but not when demographic characteristics and clinic location were controlled. Childhood sexual abuse remained associated with reduced odds of Pap screening in logistic regression analyses that controlled for clinic location, demographics, attitudes about Pap screening, and posttraumatic stress disorder symptoms (adjusted OR = 0.56, 95% CI 0.34 to 0.91).
- CONCLUSIONS: These findings suggest that childhood sexual abuse may lead to decreased probability of screening for cervical cancer, potentially contributing to the poorer health seen in other studies of women who have been sexually abused.
TABLE 1
Demographic characteristics of women with and without Pap screening
| No Pap (%) n = 372a | Pap (%) n = 364a | P | |
|---|---|---|---|
| Ethnicity | .001 | ||
| African American | 10.1 | 11.6 | |
| Asian American | 20.1 | 8.0 | |
| European American | 60.6 | 71.8 | |
| Other | 9.2 | 8.6 | |
| Age | .076 | ||
| Mean (standard deviation) | 43.8 (12.8) | 45.5 (12.4) | |
| Education | .187 | ||
| Elementary school | 2.7 | 1.4 | |
| High school | 39.6 | 34.8 | |
| College | 41.5 | 43.2 | |
| Post-college | 16.9 | 20.6 | |
| Family income | .002 | ||
| $20,000/year or less | 12.4 | 5.8 | |
| $20,001–$50,000 | 41.3 | 38.2 | |
| More than $50,000 | 46.2 | 56.0 | |
| Marital status | .012 | ||
| Never married | 34.3 | 24.8 | |
| Married | 47.0 | 59.1 | |
| Separated | 1.6 | 0.6 | |
| Divorced | 13.8 | 13.1 | |
| Widowed | 3.2 | 2.5 | |
| aSample sizes vary slightly because of missing data on individual demographic items. | |||
Prevalence of trauma
Commonly reported events during childhood included natural disaster (reported by 13% of the women), sexual assault other than rape (11%), and news of a death or injury (10%). Childhood sexual abuse or sexual assault was reported by 18.4% of the respondents. The most common traumas in adulthood were receiving news of a death or serious injury (46%), natural disasters (33%), actual or attempted robbery (27%), and serious accidents (14%). Of the respondents, 8.3% reported sexual abuse or sexual assault in adulthood. Their overall rate of childhood and adult sexual assault was 26.7%.
Associations of trauma history with pap screening
We investigated the association of trauma with screening using chi square analyses. Women who had been raped before age 18 (36% vs. 50%, n = 713, P = .050) and women who had been subjected to other sexual assaults before age 18 (35% vs. 51%, n = 694, P = .009) were less likely to have been screened. Nonsexual childhood abuse and neglect were not related to screening. Women who experienced a natural disaster during childhood (36% vs. 52%, n = 571, P = .009) and those who experienced terrorist acts during adulthood (20% vs. 49%, n = 715, P = .024) were less likely to have been screened. (Although the association with a terrorist act was significant, exposures were reported by only 3% of unscreened women and 0.9% of screened women.) Women who reported a household break-in during adulthood were slightly more likely to have been screened (53% vs. 47%, n = 656, P = .032).
In a hierarchical logistic regression model (Table 2), childhood sexual abuse, but not other traumatic events, was associated with lower odds of screening when clinic location, demographic characteristics, attitudes, and PTSD were controlled. The logistic regression model was repeated using CTQ subscales to assess trauma, with similar results. Unmarried women were less likely than currently married women to have been screened, and Latina, Native American, Asian/Pacific, or multicultural women were less likely than European American women to have been screened. Women who endorsed the statement, “I have no symptoms so I do not need a Pap test” and those who anticipated embarrassment during screening were less likely than others to have been screened; women who believed that testing would ease their mind were more likely to have been screened.
TABLE 2
Hierarchical logistic regression model of sexual trauma and attitudes as predictors of pap screening
| Predictor | Adjusted odds ratio (95% CI) |
|---|---|
| Traumatic events | |
| Break-in (adult) | 1.14 (0.77, 1.70) |
| Natural disaster (child) | 0.78 (0.45, 1.38) |
| Terrorist act (adult) | 0.28 (0.07, 1.07) |
| Childhood sexual trauma | 0.56 (0.34, 0.91) * |
| Site | |
| Santa Rosa | 0.68 (0.44, 1.04) |
| San Francisco | 1.0 (referent) |
| Oakland | 1.27 (0.80, 2.02) |
| Education | |
| Less than college | 1.09 (0.71, 1.69) |
| College | 1.0 (referent) |
| More than college | 1.01 (0.65, 1.57) |
| Ethnicity | |
| European-American | 1.0 (referent) |
| African American | 0.59 (0.33, 1.06) |
| Other than African | |
| American or European | |
| American | 0.46 (0.29, 0.71) ** |
| Unmarried (compared with married) | 0.67 (0.48, 0.94) * |
| Attitudes toward Pap screening | |
| “I have no symptoms so I do not need a Pap test” | 0.66 (0.51, 0.85) ** |
| “I’ve had negative experiences with my health care provider” | 0.90 (0.73, 1.10) |
| “Getting a Pap test would ease my mind” | 1.54 (1.25, 1.89) *** |
| “There is danger of infection from a Pap test” | 1.09 (0.83, 1.43) |
| “I do not trust the health care system” | 1.06 (0.81, 1.39) |
| “I would be embarrassed to have a Pap test” | 0.67 (0.52, 0.84) *** |
| “Women who have many sexual | |
| partners are more likely to have cervical cancer” | 0.88 (0.73, 1.06) |
| “Pap would cause sexual assault flashbacks, or health care provider looks at me in a sexual way” | 1.05 (0.77, 1.45) |
| PTSD diagnosis | 1.62 (0.91, 2.90) |
| Missing data | 0.96 (0.80, 1.13) |
| *P | |
Discussion
Childhood sexual abuse is reliably associated with a decreased likelihood of cervical cancer screening. This association persisted despite controlling for demographic characteristics, attitudes about Pap screening, and PTSD symptoms. These findings are strengthened by the consistency with which childhood sexual abuse is associated with low rates of Pap screening using 2 measures of trauma in 3 clinics. Although cost has been a major barrier to access in previous studies of cervical cancer screening, it is not a barrier for women who are members of a pre-paid health plan. It was therefore possible for us to investigate known and suspected barriers to cervical cancer screening with fewer confounding co-variables.
This study clarifies the role of childhood sexual assault in Pap screening. Sexual assault, but not other traumatic events or other types of childhood abuse, is associated with lower rates of cervical cancer screening. Furthermore, sexual assault during childhood, but not during adulthood, is strongly associated with decreased Pap screening.
