HIV Seroprevalence among Women Opting Out of Prenatal HIV Screening in Alberta, Canada: 2002–2004

The Canadian province of Alberta uses an “opt-out” strategy for prenatal HIV screening; testing is performed routinely for all pregnant women seeking prenatal care, unless they specifically choose not to be tested. Approximately 45,000 women undergo prenatal HIV screening annually in Alberta, representing >95% of all pregnant women accessing antenatal care. In 1999, 4.7% of women declined testing, and in 2000, 1.7% declined testing [1]. Edmonton and Calgary each account for approximately one-third of the province's 3.2 million inhabitants, with the remainder spread across the rest of the province.

From the inception of the opt-out program in 1998 through 2006, 4 perinatal infections were identified in Alberta (Alberta Health and Wellness, Disease Control and Prevention, unpublished data). Two infants were born to mothers who did not undergo prenatal testing; 1 child tested positive for HIV during immigration medical examination, and a fourth child was born to a mother who tested positive for HIV during pregnancy but did not receive medical care, because she was lost to follow-up.

Little is known about the HIV infection risk among the small proportion of women who opt out of testing. A study in 2000 indicated that older women and First Nations women (i.e., those with Indian treaty status) were most likely to opt out [2]. Older maternal age [3, 4], being married [3, 4], and a low perception of HIV infection risk [3, 5, 6] have been identified as factors associated with not being tested prenatally for HIV in other populations. Although these factors may be associated with lower HIV risk behaviors, it has also been demonstrated that pregnant women who decline HIV testing are often basing their risk assumptions on a poor understanding of HIV transmission risks [5].

We compared the epidemiological characteristics of women undergoing prenatal infectious disease screening who accepted prenatal HIV testing with those of women who declined prenatal HIV testing, in Alberta from 2002 through 2004. In addition, we anonymously tested prenatal specimens from woman who opted out of the HIV portion of testing to determine the seroprevalence of HIV in this patient subset.

Methods.The Alberta Provincial Prenatal Screening Program tests pregnant women for 5 infectious disease markers, including HIV. Since August 2002, all serum samples from the Alberta Provincial Prenatal Screening Program have been tested and stored at the Provincial Laboratory for Public Health (ProvLab; Edmonton, Alberta, Canada). The ProvLab provided data (derived from the prenatal requisition) on all women from whom the prenatal specimens were obtained, such as age, city of residence, and gravidity.

Women whose specimens belonged to the prenatal HIV testing “opt-out” group were linked back to a ProvLab database of known HIV-positive patients to identify women who previously tested positive for HIV. ProvLab has been the sole provider of HIV testing and confirmation in Alberta since 1986. Previous positive HIV test results were merged into the epidemiological database, and nominal information was then removed.

For anonymous serological testing of the specimens from women who opted out of HIV testing, serum samples were tested at the ProvLab in pools of 10 specimens using the AxSYM HIV1/2 gO system (Abbott Laboratories). All pools initially testing positive for HIV had their constituent serum samples retested as individual samples. Any serum sample that tested positive for HIV was confirmed by Genetic Systems HIV-1 Western blot (Bio-Rad Laboratories).

Specimens from individuals who accepted prenatal HIV screening were compared with specimens from women who declined prenatal HIV screening with regard to demographic characteristics of the women and HIV seroprevalence. HIV-positive and HIV-negative specimens were compared for both women who did and did not agree to prenatal testing. HIV-positive specimens were stratified into those from women who were previously recorded to be HIV positive on ProvLab records and those from women for whom HIV infection was a new diagnosis. All comparisons were made using descriptive statistics (e.g., χ² test and Student's t test) using Stata statistical software, version 8.0 (Stata). The Health Research Ethics Board at the University of Alberta, Edmonton, approved this study.

Results.In total, 110,408 specimens were submitted to the ProvLab for prenatal HIV testing from 1 August 2002 through 31 December 2004. Of these, 4343 (3.8%) were not tested for HIV, because the woman had opted out of testing.

From 2002 through 2004, there was a significant decrease in the proportion of submitted specimens from woman who had chosen to opt out of prenatal HIV testing (from 4.3% in 2002 to 4.1% in 2003 to 3.6% in 2004; P < .001) (table 1). Women who opted out of HIV testing were significantly older by 1 year of age (29 years vs. 28 years; P < .001) and were significantly more likely to have higher gravidity, compared with women whose specimens were tested for HIV (P < .001). There were significant differences in the proportion of opt-out specimens by region of the province. There was a significantly higher proportion of opt-out specimens from certain regions (Calgary and southern Alberta) than from other regions; this statistically significant increase was noted for all 3 years that were examined.

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 1

Comparison of prenatal specimens submitted for HIV testing and prenatal specimens not submitted for HIV testing (“opt-out” specimens), Alberta, Canada, 2002–2004.

Of 4343 opt-out specimens, 11 (0.25%) were known to be previously positive for HIV and were removed from testing, and 8 (0.18%) did not have sufficient serum for testing. Therefore, a total of 4324 specimens were tested, and of these, 3 (0.07%) tested positive for HIV. These 3 specimens were from women from different regions of the province, and 2 of the women had previous pregnancies. Of all specimens prenatally tested for HIV, 46 (0.04%) tested positive, 22 (0.02%) of which were from women for whom HIV infection was a new diagnosis. The proportion of newly diagnosed HIV cases among the opt-out group was greater than the proportion of newly diagnosed cases among the group who opted to undergo prenatal HIV testing (0.07% vs. 0.02%; prevalence ratio, 3.3; 95% CI, 1.0–11.2).

Correlates of HIV seropositivity were similar for specimens that were tested prenatally for HIV and for the opt-out specimens. There were no statistically significant differences with regard to maternal age or year of specimen submission between the groups. Gravidity was statistically significantly higher among women with specimens testing positive for HIV than among women with specimens testing negative for HIV. There was a higher proportion of specimens from previously HIV-positive women than from women with newly diagnosed HIV infection in the northern region of Alberta and Edmonton, whereas specimens from Calgary and the southern region of the province were more likely to be from women with newly diagnosed HIV infection than from women with previously diagnosed HIV infection.

Conclusions.The proportion of specimens from women who opted out of prenatal HIV testing decreased from 4.3% in 2002 to 3.6% in 2004. Despite the small proportion of opt-out specimens, HIV seroprevalence among these specimens was 3.3 times higher than that among specimens prenatally screened for HIV. This may indicate that women at higher risk for HIV infection are opting out of testing.

Women opting out of HIV testing were significantly older, supporting previous literature [2–4], and had higher gravidity and parity (data not shown) than the women who agreed to HIV testing. A higher proportion of specimens from the southern part of the province were not screened for HIV, which may indicate a difference in prenatal screening protocol among some clinicians in those regions. Women from Calgary were significantly older than women from other regions (30 years vs. 28 years; P < .001), which in part may explain the higher proportion of opt-out specimens from this city.

Prenatal HIV screening programs represent an important opportunity for preventing mother-to-child transmission. Archibald et al. [7] estimated that, by screening 90% of pregnant women across Canada, a 65% reduction in the number of HIV-infected infants could be achieved. Policies and programs for prenatal HIV testing vary across Canada; however, the highest testing rates are reported for provinces using an opt-out approach [8]. Interestingly, our seroprevalence results closely mirror results from the province of Ontario, which uses an “opt-in” approach for prenatal HIV screening. Remis et al. [9] reported rates of HIV infection of 0.6 cases per 1000 pregnant women not tested for HIV and 0.3 cases per 1000 pregnant women screened for HIV in 2002.

We feel that our data support routine prenatal HIV testing, which has been endorsed by the Centers for Disease Control and Prevention since 2001 [10]. Acceptance of this recommendation has been limited by concerns that pregnant women will be tested without their consent or knowledge [11]. Proponents of universal testing argue that it is unethical not to test pregnant women, given our ability to reduce the rate of mother-to-child transmission of HIV among pregnant women receiving optimal care to <2% [12]. In addition, requirements for testing pregnant women for other communicable diseases—such as syphilis and hepatitis B, for which mother-to-child transmission is less preventable with currently available interventions, compared with HIV infection—do not require such stringent testing and counseling requirements. In a study conducted in the United States, prenatal HIV testing was not perceived differently than other prenatal infectious diseases screening programs; in fact, 69% of women at an antenatal clinic supported routine HIV screening [6].

This study was limited by the lack of information available on prenatal specimens. Additional information, such as knowledge of the womens' HIV status, ethnicity, or risk behaviors, would have been important in fully examining the characteristics and reasons for women opting out of prenatal HIV testing. For example, it is not clear whether the 3 women for whom HIV infection was a new diagnosis who opted out of HIV testing were unaware of their HIV status or whether they had previously tested positive for HIV outside Alberta and, therefore, opted out of testing. In addition, ProvLab information was specimen-based; therefore, women whose specimens were tested multiple times during 1 pregnancy may have biased seroprevalence estimates, although this proportion is thought to be low. And finally, the low number of HIV-positive specimens limited the analyses and strength of the conclusions that can be drawn from this research.

Although our data support routine HIV testing during pregnancy, further discussion is warranted regarding the ethical implications of enacting broad screening policies for HIV during pregnancy. At the same time, policies and programs that enhance our ability to reach high-risk women who may not access antenatal care are warranted.

• Received June 4, 2007.
• Revision received August 30, 2007.
• Accepted November 12, 2007.

• © 2007 Infectious Diseases Society of America