Pap Test and HIV testing behaviors of South Carolina women 18–64 years old

Abstract

Background

Pap tests are used to detect abnormal cell growth in the cervix. Early detection of precancerous cells increases the likelihood of treatment success. In fact, the Pap test is one of only a few cancer screening procedures that can prevent cancer by virtue of identifying and intervening upon abnormal precancerous cells in the early stages. There has been a growing prevalence of cervical cancer among HIV-positive women. Early diagnosis is essential to improving survival outcomes of women living with HIV/AIDS. The purpose of this study was to describe Pap test behaviors among women in South Carolina and examine its relationship with HIV testing.

Methods

Behavioral Risk Factor Surveillance System (BRFSS) data from 2008 were analyzed. Chi-square tests and logistic regression analyses were performed to describe Pap test behaviors among our sample (n=3,404) of non-Hispanic White and non-Hispanic Black women (18–64 years old).

Results

The majority of participants (97%) reported ever having a Pap test. Participants who reported never having a Pap test were 50% less likely to have ever been tested for HIV.

Conclusions

The participants in our study met the Healthy People 2010 target of ever having a Pap test. However, less than half of participants (42%) had ever been tested for HIV. A larger proportion of women who have had a Pap test had also been tested for HIV.

Practice Implications

Our findings suggest that offering women an HIV test during routine Pap tests may present an opportunity to increase the number of women in South Carolina who know their HIV serostatus.

Background and Significance

Oncogenic genital human papillomavirus (HPV) infection has been identified as the main etiological cause of cervical cancer.^1,2 Studies have shown that women with human immunodeficiency virus (HIV) have a high prevalence, incidence, and persistence of HPV infection.^3–7 Because of the elevated risk of HPV-associated cervical cancer among women with HIV/AIDS,⁸ the early detection of both HIV and cervical abnormalities will play a vital role in reducing incidence and mortality rates for these diseases which disproportionately affects African American (AA) women.⁹

Papanicolaou (Pap) tests are used to detect cellular changes in the cervix. There is a growing debate in the scientific literature over the use of HPV DNA testing either alone or in conjunction with (i.e., contesting) the cytological exam.^1,10–13 Nonetheless, the Pap test remains a useful screening tool for the prevention of cervical cancer. Current screening guidelines recommends routine Pap tests for women three years after sexual debut, but no later than age 21.¹⁴ Behavioral Risk Factor Surveillance System (BRFSS) data suggest that among women 18 years and older residing in South Carolina (SC) in 2006, 96.1% reported that they had received a Pap test.¹⁵ These findings show that just past the mid-year review, SC was very close to meeting the Healthy People 2010 objective for 97% of women 18 years and older to have ever had a Pap test.¹⁶

On the other hand, with regard to HIV testing the news was not so good. Only 35.4% percent of women (18–64 years old) from SC who participated in the 2006 BRFSS survey reported that they have ever been tested for HIV.¹⁵ Even more alarming is the fact that unlike other AIDS-defining cancers such as Kaposi’s sarcoma and Hodgkin’s lymphoma, the incidence of invasive cervical cancer among women with HIV/AIDS has not decreased with the arrival of highly active antiretroviral therapy (HAART).^7,17 These data underscore the importance of early diagnosis of HIV as women diagnosed with both AIDS and cervical cancer do not do as well as those who have an intact immune system.⁸ Thus, it is imperative that women learn their HIV serostatus as soon as possible so that those who are infected with HIV can initiate treatment early to prevent (or at least delay) the progression to an AIDS diagnosis.^18–20 As the incidence and prevalence of HIV/AIDS continues to increase among females,²¹ it will become increasingly important to ensure that women are screened for both cervical cancer and HIV.

The Centers for Disease Control and Prevention (CDC) revised its recommendations for HIV testing to include all persons 13–64 years old regardless of risk. The goal is to increase the number of persons who are aware of their HIV status by offering HIV testing where other diagnostic tests and routine screenings are performed.¹⁹ Subsequently, the American College of Obstetrics and Gynecology (ACOG) recently released a formal committee opinion that supports the CDC revised HIV testing guidelines. The ACOG recommends that women between the ages of 19 and 64 years old be tested for HIV as a routine part of gynecological care.²² The purpose of this study was to describe Pap testing behaviors among women in SC and examine the behaviors in relationship to HIV testing. We hypothesized that women who had ever had a Pap test will be more likely to have also been tested for HIV.

Methods

The BRFSS telephone-based survey was administered to a sample (n=414,509) of non-institutionalized adults 18 years and older in the United States (US), Guam, Puerto Rico, and the US Virgin Islands.²³ Participants were randomly selected from a sample of all telephone-equipped households, with post-stratification adjustments made for selection and response bias. Data were weighted annually according to state-specific demographics such as gender, race/ethnicity, age, and income.²⁴ This cross-sectional study used data from the 2008 BRFSS survey. The response rate for the SC 2008 BRFSS survey was 60.6%.²⁴

Non-Hispanic White and non-Hispanic Black females (18–64 years old) who answered either yes or no to having ever had a Pap test (n=3,404) were included in this study. The frequencies of Latinas and females falling in other racial groups (including multiracial) were too small to include in this analysis. Although we did consider combining these women into an “other” category, we did not for two reasons: 1) after collapsing these groups, numbers were inadequate to obtain reliable point estimates and 2) Latinas are culturally different thus, it is important to treat them as a separate group.

Although Pap tests are not recommended for women who do not have a cervix,¹⁴ we did not subgroup participants based on self-reported hysterectomy for three reasons. 1) Our outcome of interest was based on the Healthy People 2010 goal¹⁶ which measured the number of women 18 years and older that had ever had a Pap test. Therefore, it was appropriate for our sample to include participants for whom Pap tests were no longer recommended. 2) While the BRFSS does ask participants whether or not they have ever had a Pap test, the question regarding hysterectomy is not specific about whether or not the procedure was a complete or partial hysterectomy. As a result, we are unable to ascertain whether participants who reported that they had a hysterectomy also had a cervix. Thus, some participants (21%) may have been excluded unnecessarily. 3) Guidelines recommend that women who have a total hysterectomy to treat cervical cancer or a condition that might lead to cancer may still need to be screened.²⁵

Statistical Analysis

Chi-square tests and logistic regression analyses were performed using weighted 2008 BRFSS data. The main goal of the proposed research was to describe Pap test behaviors among participants (18–64 years old) in SC. Associations between Pap testing and HIV testing behaviors were also examined. SAS version 9.2 (SAS Institute Inc., Cary, NC, USA) was used to perform these analyses on the weighted data.

Having ever had a Pap test was the dependent variable or outcome. Having ever been tested for HIV was used as the main independent variable or exposure in the subsequent logistic regression analyses in which having ever had a Pap test was modeled. First, behavioral associations between the outcome and exposure were examined. Next, associations between the outcome and the covariates were assessed using simple logistic regression models. Similar associations were examined between the exposure and each of the covariates in which having been tested for HIV was modeled.

Covariates associated with both the outcome (ever had a Pap test) and the exposure (ever had an HIV test) were assessed for confounders in the subsequent multivariate logistic regression analyses. Variables were removed in a backward stepwise fashion. If the odds ratio of the exposure (having ever been tested for HIV) was changed by more than 10% then that variable was kept in the model as a potential confounder of the relationship between Pap and HIV testing behaviors. Otherwise, variables were removed one at a time until all remaining variables were either significant based on an alpha level of significance less than 0.05 or could not be removed because it was a potential confounder. The odds-ratio, 95% confidence interval, and p-value for these behavioral associations are reported.

Outcome Variables

Our primary outcome of interest was whether or not participants had ever had a Pap test. This was assessed by participants’ response to the following BRFSS question: Have you ever had a Pap test? Responses were dichotomized as “yes” or “no”. The secondary outcome of interest was whether participants had ever been tested for HIV. This was assessed by participants’ response to the following BRFSS question: Have you ever been tested for HIV? Participants were instructed to exclude HIV testing that was done as part of a blood donation and include rapid HIV testing that used fluid from their mouth. Response options were “yes” or “no” which were recoded as “tested” or “not tested”. Participants who responded don’t know/not sure or refused to answer these two items were excluded from the analysis.

Covariates

Sociodemographic characteristics (age, race/ethnicity, marital status, education, employment, poverty status, and urbanization) and healthcare access (insurance, personal doctor, affordable healthcare, and routine medical checkup) variables were used to describe sample characteristics as well as control for potential confounding between the outcome (having ever had a Pap test) and the exposure (having ever had an HIV test) in the multiple logistic regression models.

Results

Pap Testing Behaviors

A majority (97%) of participants between the ages of 18 and 64 who completed the 2008 SC BRFSS survey reported ever having a Pap test. Participants between 25 and 34 years old had similar proportions for those reporting they had ever had a Pap test (28%) and those who had never had a Pap test (30%). However, almost half (47%) of participants who had never had a Pap test were between 18 to 24 years old. Compared to the proportion of participants 35 to 49 years old (9%) and 50–64 years old (14%) who had never had a Pap test, a larger proportion of participants in these age groups reported that they had ever had a Pap test, 35% and 31% respectively. These differences were statistically significant (p<.0001) as reported in Table 1.

Table 1.

Characteristics of SC women (18–64 years) by Pap test (n=3,404), BRFSS 2008

Characteristic		Pap test		No Pap test
		n	%	n	%	p-value
Totala		3321	97	83	3
Age Category (yrs)						**
	18–24	94	6	20	47
	25–34	443	28	14	30
	35–49	1170	35	14	9
	50–64	1614	31	35	14
Race/ethnicity						NS
	NH white	2151	70	44	68
	NH black	1170	30	39	32
Marital Status						**
	Married/partnered	1917	68	24	15
	Divorced/separated/widowed	903	16	16	12
	Never married	494	16	43	73
	Missing	7
Educational Level						NS
	Did not graduate from high school	253	6	17	17
	High school graduate	1046	29	36	36
	Some technical school/college	974	33	16	27
	Technical school/college graduate	1047	32	14	20
	Missing	1
Employment Status						NS
	Employeda	2380	70	46	59
	Unemployedb	937	30	36	41
	Missing	4		1
Federal Poverty Guidelines (%)
	Poverty Statusc					NS
	≤100	523	17	30	19
	125–200	757	22	21	27
	>200	2040	61	32	54
	Missing	1
Urbanization						NS
	MSA (urban/suburban)	1971	75	54	82
	non-MSA (rural)	1350	25	29	18
Health Care Accessd
	Had health insurance (REF)	2729	82	46	54	*
	Had a personal doctor (REF)	2978	89	61	65	**
	Was able to afford needed healthcare (REF)	2596	78	55	68	NS
	Had a routine medical checkup (REF)	2343	69	42	43	*
HIV Testinge						*
	Tested	1271	43	27	20
	Not tested	2050	57	56	80

^*P<.05,

^**P<.0001,

^NSnot significant, P>.05

^aincludes self-employed and retired

^bincludes homemakers and students

^cbased on the USDHHS 2008 Federal Poverty Guidelines

^dwithin the past year

^eexcludes HIV testing done as part of blood donation; includes oral (rapid) HIV test

Sociodemographics

Among participants who reported that they had ever had a Pap test, the majority was married or had a partner (68%). Whereas the majority of participants who reported they had never had a Pap test had never been married (73%). These data are reported in Table 1 and are also statistically significant (p<.0001). For other sociodemographic characteristics such as race/ethnicity, educational attainment, employment status, poverty status, and urbanization there were no statistically significant differences between the proportions of participants who had ever had a Pap test compared to those who had never had a Pap test. These data are reported in Table 1.

Healthcare Access

While there were no statistically significant differences between the proportion of participants who reported that they had ever had a Pap test and those who had never had a Pap test based on their ability to afford needed healthcare (p=0.1536) within the past year, there were statistically significant differences for other healthcare access issues in the past year such as having health insurance (p=0.0001), a personal doctor (p<.0001), and a routine medical checkup (p=0.0019). Most of the participants who had a Pap test had health insurance (79%), at least one personal doctor (89%), and a routine medical checkup (69%) within the past year. These data are reported in Table 1.

HIV Testing Behaviors

There were statistically significant differences in HIV testing behaviors between the proportions of participants who had ever had a Pap test and those who had never had a Pap test (p=0.0010). Although the majority of participants who had ever had a Pap test had never been tested for HIV, 43% who reported that they had ever had a Pap test had also been tested for HIV. Among the participants who had never had a Pap test, the vast majority of them (80%) had never been tested for HIV. These data are reported in Table 1. After adjusting for age, race/ethnicity, marital status, educational attainment, percent federal poverty guidelines, urbanization, and having health insurance within the past year participants who had never been tested for HIV were less likely to report that they had ever had a Pap test (OR 0.50, 95% CI 0.28, 0.89) (See Table 2).

Table 2.

Odds of SC women (18–64 years) having an HIV test by Pap test (n=3,390)^a, BRFSS 2008

Characteristic	Pap test (n=3,308)		No Pap test (n=82)		Odds-Ratio (OR)		Adjusted Odds-Ratio (AOR)b
	n	%	n	%	OR (95% CI)		AOR (95% CI)
Ever tested for HIVb
Tested	1267	43	27	20	1.00	Ref.	1.00	Ref.
Not tested	2041	57	55	80	0.51	(0.28, 0.92)	0.50	(0.28, 0.89)

^*P<.05

^an=3,390 after excluding observations with missing values for either the response or explanatory variables

^bafter adjusting for age, race/ethnicity, marital status, education, poverty status, urbanization, and health insurance

In addition to observed differences in HIV testing behaviors by participation in Pap testing we also observed statistically significant differences in HIV testing by age groups (p<.0001). Of the 42% of the participants (n=1,298) that reported that they had been tested for HIV only 8.1% were between the ages of 18–24 years old. The majority of the participants that had been tested for HIV were between 25 and 49 years old, with 38% of those 25–34 years old and 38% of those 35–49 years old reporting that they had been tested for HIV. However, only 16% of women aged 50 years and older reported that they have ever been tested for HIV.

Discussion

The participants in this met the Healthy People 2010 for 97% of all women 18 years and older to have ever had a Pap test.¹⁶ These data are consistent with previous years since women in SC have been progressing towards the Healthy People 2010 goal and near the mid-year review, had nearly met the 97% target.^15,26 Although Oster et al.²⁷ found Pap testing to be negatively associated with increasing age, this was not the case in our study. There were similar proportions of women 50 years and over that reported having ever had a Pap test compared to women in the 25–34 and 35–49 age groups. Nonetheless, when examining Pap testing behaviors especially among populations that are disproportionately burdened by cervical cancer, it is important to keep in mind that although higher Pap testing rates may be reported among minority women, this group of women continues to have disproportionately higher incidence and death rates of cervical cancer. As a case in point, data reported by CDC on the top ten cancers in 2005 for SC only included Black women for both cervical cancer incidence and death for which they ranked 10^th and 7^th respectively with age-adjusted incidence (9.3 per 100,000) and mortality (5.5 per 100,000).²⁸

In contrast to these Pap test findings, only 42% of the women in this study have ever been tested for HIV. These data are somewhat consistent with previous years (2001–2006) where HIV testing rates among women in SC ranged from 35.4% to 53.3%¹⁵, albeit still a long way from reaching the CDC’s goal of all persons knowing their HIV serostatus.¹⁹ Although providers have not fully implemented the CDC’s revised recommendations¹⁹ to date, the ACOG’s support of these revised HIV testing guidelines will hopefully increase the adoption of routine HIV testing practices during regular gynecological care visits.²² Subsequently, our finding that women who have ever had a Pap test were more likely to have also been tested for HIV suggest further research is needed to explore how Pap tests may be used to facilitate HIV prevention efforts to reduce the incidence and mortality rates of both cervical cancer and HIV/AIDS. These data are especially relevant given recent findings of Oster et al.²⁷ that routine cervical cancer screening (i.e., annual Pap test) is suboptimal among women infected with HIV. This evidence is particularly concerning given the increased risk of cervical cancer among women with HIV/AIDS^3–8 which has not been shown to be decreasing in the era of HAART.^7,17

Strengths and Weaknesses

A major strength of this study is the fact that both Pap and HIV testing behaviors were assessed from the same individuals at the same time. The BRFSS which is large nationally representative surveys allowed for the examination of the relationship between HIV testing and routine gynecological screening procedures (i.e., Pap test) post the CDC’s 2006 revised HIV testing recommendations. To date there are no published studies that have examined the relationship between these two preventive health behaviors among both HIV-negative and HIV-positive persons. Our findings for women 18–24 years old may be limited as the study’s participants were sampled from households with land-line telephones and this age group is increasingly using only cell phones. Thus, we suspect that HIV testing rates for this age group are underreported. However, the fact that this age group had the lowest Pap testing rates is not surprising given current Pap test screening recommendations of being screened by age 21 and previous reports of low Pap test screening rates in this population.¹⁴ We are further limited by the study’s design in that both of the outcomes of interest were self-reported. However, Ferrante et al.²⁹ recently reported 61% congruence between self-reported cervical cancer screening and medical records data. Therefore, self-reported Pap testing behaviors may not be a limitation of this study. In addition, HIV-status was not reported which would have facilitated being able to assess cervical cancer screening behaviors for women infected with HIV.

Implications for Practice and Public Health

The CDC estimates that about 21% of all persons living in the US are HIV-positive but unaware of their serostatus.¹⁹ In response to this public health threat, the CDC revised its HIV testing guidelines and currently recommends that all adolescents/adults between the ages of 13–64 years old be tested for HIV, regardless of risk. To meet this goal, the CDC proposed that an HIV test be offered routinely as part of other routine diagnostic and screening procedures. Our state’s success with Pap testing may represent an opportunity to increase HIV testing among women. This is of particular importance to AA women living in SC who are disproportionately burdened by both cervical cancer and HIV/AIDS.^30–43

Our findings suggest additional research is needed to further examine the relationship between Pap and HIV testing behaviors. Given SC’s overall success in cervical cancer screening, linking Pap testing with HIV testing during routine health visits may facilitate meeting the CDC’s goal of increasing the number of persons who know their HIV status.¹⁹ Given that older women (50+ years old) in this study had a lower HIV testing rate, studies^18,20 and epidemiological data²¹ have shown that this age group is more likely to receive a late HIV diagnosis. Healthcare providers should ensure that their HIV test offering practices also include this population. Increasing HIV prevention efforts aimed at meeting the specific cultural needs of older women are warranted given the growing incidence and prevalence of HIV/AIDS among older adults²¹ and the dearth of HIV prevention programs that target this population.⁴⁴ The fact that the older women in our study had similar Pap test screening rates for women 25–49 years old suggest further that these routine gynecological exams may represent potential missed opportunities to target this group for HIV prevention messages.