Increasing Voluntary HIV Testing by Offering Oral Tests in Incarcerated Populations

Abstract

Objectives. This study reports responses of incarcerated persons to voluntary blood and oral HIV testing.

Methods. Males and females in local detention and juvenile justice facilities in Maryland (n = 1314) chose oral or blood testing and reported reactions to the oral HIV test. The relationship of demographics and HIV risk factors to test choice was examined.

Results. Reactions to oral testing were very favorable; some participants reported that they would not otherwise have been tested. Participants who chose oral testing were more likely to be male and African American, but they did not differ from those who chose blood testing in most risk factors or in seroprevalence.

Conclusions. Oral HIV testing in correctional settings may promote voluntary testing among persons who otherwise would refuse or avoid testing, especially among groups (males and African Americans) disproportionately affected by HIV.

Incarcerated populations have a higher prevalence of HIV1 and AIDS2 than most other populations. Nationwide, among state prisoners tested in 1997, 2.2% of males and 3.4% of females were HIV positive.3 In Maryland in 1998, 3.0% of male inmates and 4.7% of female inmates were HIV positive (including those in state, local, and youth facilities); the prevalence at all other Maryland counseling and testing sites (CTSs) was 1.8% among males and 1.0% among females. Among Maryland inmates treated at prison health clinics for symptoms warranting HIV testing, 11% of males and 15% of females were HIV positive in 1998.4 Other states also report a high seroprevalence among inmates.5

High seroprevalence in correctional settings represents an opportunity for secondary prevention if HIV-positive individuals can be identified and provided appropriate treatment and postrelease referrals. In Maryland's state prisons, voluntary HIV testing is offered to all adults at intake, but fewer than half accept testing.5 Intake testing at local detention and youth facilities is not routine, but it is generally offered by local health departments or other contracted personnel at some point during incarceration. The proportion of inmates accepting testing is unknown, but without a routine offer it is presumably even less than among state inmates.

Given the high seroprevalence, ways of increasing testing among this population are urgently needed. We report here reactions of male and female inmates in Maryland to oral HIV testing, compare those who chose oral testing (“oral testers”) with those who chose blood testing (“blood testers”), and examine increases in testing at sites with and without available oral testing. Results are from a broader oral testing project that included nonincarcerated populations.

METHODS

Background and Procedure

Maryland's Orasure Demonstration Project introduced oral HIV testing at 32 counseling and testing sites between September 1997 and April 1998. Ten sites were local detention facilities for adults or youths. All sites provided standard CTS pretest HIV counseling. Most noncorrectional sites offered a choice of confidential or anonymous testing, but correctional sites offered only confidential testing. Clients heard descriptions of both blood and oral testing and signed written consents to participate.

Oral testers were invited to complete a posttest survey about their reactions. Demographic and risk factor comparisons between oral testers and blood testers were based on CTS records for all tests at participating sites during the project. CTS reports were linked to client satisfaction survey results by means of a unique peel-and-stick number label taken from the CTS form.

Instrument

The client survey included items related to issues such as previous testing, fear of needles, perceptions of the oral test, effect of oral test availability on the decision of whether to be tested, and perceptions of privacy and counseling. This report examines perceptions of the oral test (3 items), the impact of oral test availability on the decision to be tested (2 items), and past refusal or avoidance of blood testing owing to fear of needles (2 items).

RESULTS

The 10 correctional facility test sites reported 908 oral tests and 406 blood tests during the demonstration project. All comparisons of oral test reactions by sex or by sex and race are based on the 759 oral testers (84%) for whom complete information on sex and race was available. Those tested were primarily male, African American, and younger than 30 years (Table 1 ▶).

TABLE 1—

Characteristics of Incarcerated Men and Women Who Chose Oral or Blood Testing for HIV: Maryland, 1997–1998

	Oral Test (n = 908)	Blood Test (n = 406)	Odds Ratio (95% CI)	Total (n = 1314)
Sex, %
Male	80.3	73.6	1.46 (1.11, 1.92)	78.2
Female	19.7	26.4		27.8
Race/ethnicity, %
African American	72.2	51.5	2.63 (3.37, 2.05)a	65.6
White	23.7	44.3		30.2
Other	4.1	4.2		4.2
Age, y	26.5	25.9	NS (t test)	26.3
Seropositivity, %
Men	1.0	0.7	NS	0.9
Women	1.1	2.9	NS	1.8
Previously tested, %
Yes	67.7	63.5	1.20 (1.12, 1.29)	66.1
No	32.3	36.5		33.9

Note. CI = confidence interval; NS = nonsignificant.

^aAfrican American vs White only.

Demographics

The odds of choosing oral testing over blood testing were almost 1.5 times greater for men than for women and more than 2.5 times greater for African Americans than for Whites. Because results differed significantly by both sex and race/ethnicity, test choice was compared by sex within each racial group. Only among African Americans did choice of test differ significantly between men and women, with men 1.8 times more likely to favor oral testing. Oral and blood testers did not differ on the basis of age (t₁₃₁₁) = 1.1, P = .27). HIV seropositivity did not significantly differ between oral and blood testers, for either men or women. Blood testers were slightly more likely to have been previously tested (odds ratio [OR] = 1.20; 95% confidence interval [CI] = 1.12, 1.29).

Risk Factors

Comparison of risk factors for oral and blood testers was done separately for men and women, because some risk factors (e.g., anal or oral intercourse between men) are sex specific. Among men, blood testers were significantly more likely than oral testers to report sexual activity during noninjecting drug use (OR = 2.03; 95% CI = 1.46, 2.83) or sexual relations with a person with “other” (unspecified) HIV/AIDS risks (OR = 4.67; 95% CI = 3.48, 6.26); oral testers were significantly more likely to report that they had exchanged sexual acts for drugs or money (OR = 2.25; 95% CI = 1.59, 3.19). No difference was found in other risk factors (such as sexual relations with other men or the use of injecting drugs). Among women, blood testers were significantly more likely to report sexual activity with a person with “other” risks (OR = 3.5; 95% CI = 2.15, 5.70), while oral testers were significantly more likely to report diagnosis of a sexually transmitted disease (OR = 1.69; 95% CI = 1.04, 2.74) or exchanging sexual acts for drugs or money (OR = 1.77; 95% CI = 1.08, 2.91). No difference was found in other risk factors.

Responses to the Oral Test

Responses to the oral test were favorable across all demographics. The majority of both men and women agreed that the availability of the oral test made it easier to decide to be tested and made them likelier to be tested in the future. Importantly, 16% of women and 18% of men indicated that they had put off or refused testing in the past owing to fear of needles, and 22% of men and 15% of women said that they would not have been tested without the oral option (Table 2 ▶).

TABLE 2—

Reactions of Incarcerated Men (n = 610) and Women (n = 149) Who Chose Oral HIV Testing: Maryland, 1997–1998

	Agree/Strongly Agree, %	Don't Know, %	Disagree/Strongly Disagree, %	χ2 (P)
In past, put off testing owing to fear of needles
Men	18.1	20.8	61.1	7.78 (.02)
Women	16.3	11.6	72.1
In past, refused test owing to fear of needles
Men	18.2	12.0	69.8	7.69 (.02)
Women	17.6	4.2	78.2
Oral test made it easier to decide to be tested
Men	58.6	16.8	24.6	1.98 (NS)
Women	62.8	12.2	25.0
More likely to be tested in future if test is oral
Men	54.7	23.9	21.4	10.26 (.006)
Women	55.3	13.5	31.2
Oral test is more difficult than blood test
Men	10.8	23.4	65.8	0.82 (NS)
Women	9.2	21.1	69.7
If there were no oral test, would have had blood test
Men	61.9	21.1	17.0	14.71 (.001)
Women	79.0	11.9	9.1
If there were no oral test, would not have been tested
Men	22.3	19.5	58.2	10.23 (.006)
Women	15.0	12.1	72.9

Some differences by sex did emerge (Table 2 ▶). Women were more likely than men to deny that they had put off testing in the past (χ² = 7.78, P = .020) or had refused to be tested (χ² = 7.69, P = .021) because of fear of needles. Men showed more uncertainty by responding “don't know” on these items. Women were also more likely to deny that the oral test made it more likely that they would be tested in the future, although majorities of both sexes agreed with this statement (χ² = 10.26, P = .006). However, there was no significant difference by sex regarding whether or not the oral test made it easier to decide to be tested. Women were more likely to report that they would have had a blood test if the oral test were unavailable (χ² = 14.71, P = .001) and to deny that they would not have been tested (χ² = 10.23, P = .006). Finally, differences by sex emerged in response to a question about which test was better on the basis of the respondent's experience (χ² = 9.98, P = .016; data not shown). Women were more likely to respond that the oral test was better (49.7% vs 20.8%), whereas men more often responded “don't know—never tested before” (19.8% vs 10.1%). Very few individuals of either sex (5.1 % of men and 6.7% of women) said that the blood test was better.

No significant differences were found on the basis of race/ethnicity. Significant differences were found for 7 of 8 items when participants were divided into 4 age groups (younger than 20, 20–29, 30–39, and 40 and older), but every comparison indicated that those younger than 20 were more likely to respond “don't know” to each question.

Increases in Total Testing

At the 10 participating detention facilities, HIV testing during the demonstration project increased by 63% compared with the same time period a year earlier. The increase was 28% across all 32 participating sites. In contrast, there was only a 0.5% increase across all other Maryland testing sites combined.

The greater likelihood of previous HIV testing for oral rather than blood testers appears contradictory to the attitudes reported in the client survey. However, when blood and oral testers from all correctional sites were compared for calendar year 1998 (when oral testing became more widely available), the opposite pattern occurred. Of 12 709 tests in 1998 (including the last months of the demonstration project and introduction of oral testing at other correctional sites), 10 659 were blood and 2050 were oral; 25.2% of blood testers, but 29.3% of oral testers, had not been previously tested (χ² = 722.5, P < .0001). This finding in the larger data set is consistent with the survey finding that oral testing appeals to clients who would have refused blood testing. Although speculative, one possible reason for this inconsistency between earlier and later oral testers is that oral testing was newly introduced to Maryland in the demonstration project, and clients who had been previously tested may have been more receptive to HIV testing in a new modality.

DISCUSSION

The oral HIV test is perceived favorably by inmates across all sex, race/ethnicity, and age categories. Inmates report that it is easier than the blood test, makes the decision to be tested easier, and increases the likelihood of future testing. Both male and female inmates are 7 to 8 times more likely to favor the oral test than the blood test. Substantial minorities of both sexes agreed that they would not have been tested if there were no oral test, and that they had put off or refused testing in the past because of fear of needles. A study on oral testing in New York (not based on an inmate population) also reported a strong preference for oral testing among those offered a choice.6 These findings support the view that the availability of oral testing has the potential to significantly increase acceptance of voluntary HIV testing among both male and female inmates and to enable identification of seropositive inmates.

Men and African Americans were disproportionately more likely to choose the oral test. Importantly, incarcerated women appeared to be less averse to needles than men and more receptive to testing, regardless of the availability of the oral test. Thus, the oral HIV test may be particularly useful in facilitating testing among incarcerated men—especially those who are averse to needles—and among African Americans, who are disproportionately affected by HIV. A study in Michigan found that, as in Maryland, relatively more African Americans chose oral testing than did those of other racial/ethnic backgrounds.7 These findings indicate that oral testing can enhance the reach of testing to at-risk individuals in the incarcerated population.

Despite some differences in reported risk factors (e.g., oral testers of both sexes were more likely to report exchanging sexual acts for drugs or money), oral and blood testers did not differ in the occurrence of major risk factors such as injection drug use or, for men, sexual intercourse with another man. Importantly, they also did not differ in seroprevalence. Thus, oral testing appeals to persons with a wide variety of risk factors, and those choosing this option are just as likely as blood testers to be seropositive. Some laboratory concerns about the sensitivity and specificity of oral testing were reported, but routine procedure for a Maryland CTS calls for a second testing by blood specimen following any positive result from an initial blood or oral fluid specimen. The sensitivity of oral testing exceeds 99% in clinical trials comparing oral and blood tests.8

Identification of more seropositive individuals at intake or later during incarceration would allow the provision of clinical services in the detention setting and appropriate postrelease referrals. More testing could also raise awareness of HIV/AIDS in seronegative inmates through exposure to pre- and posttest counseling. Given the favorable evaluation of oral testing by the incarcerated participants, and its potential to increase voluntary testing among groups more affected by HIV (such as African Americans), oral testing continues to be offered at local detention centers and juvenile justice facilities in Maryland. Discussions are under way to increase the availability of oral testing at other types of testing sites.