The BRFSS in 2004 had responses from 303,822 people. In total, 227,326 (75%) met the required criteria for age and state of residence. Among these, 209,903 (92%) were willing to respond to questions about HIV testing. Of these respondents, 87,120 (42%) reported having ever been tested for HIV outside of blood donations, and 28,612 (14%) reported having been tested in the 12 months prior to the survey. After taking into account sampling weights, about 17% of respondents reported that they had been tested for HIV in the prior 12 months.
As of 2004, ten states had enacted a statute requiring written informed consent prior to a patient’s undergoing HIV testing: Hawaii (which was excluded from this analysis), Maine, Maryland, Massachusetts, Michigan, Nebraska, New York, Pennsylvania, Rhode Island, and Wisconsin. According to the 1998 and 2004 compilations of HIV-related state statutes published by Law & Sexuality
, none of the statutes relating to written informed consent changed between 1998 and 2004.19,20
In bivariate analysis, women who were aged 18–24 years; described themselves as black, Asian, multiracial, or of Hispanic ethnicity; had a lower education; had a lower income; were without health-insurance coverage; were not in a committed relationship; and had at least one risk factor for HIV were relatively more likely to report being tested for HIV in the 12 months prior to the survey. States were also ranked according to the prevalence of AIDS per capita. The sample was then split into two groups (high and low AIDS prevalence per capita). Respondents living in states with a higher prevalence of AIDS per capita were more likely to report a recent HIV test than respondents living in states with a lower prevalence of AIDS per capita (18.3% vs 12.5%, respectively).
Overall, respondents living in a state with a statute requiring written informed consent were slightly less likely to report a recent HIV test than respondents living in states with no such statute (16.3% vs 16.8%, respectively). While this finding was not significantly different from no association in unadjusted analysis (OR=0.96; p=0.081, 95%CI=0.91, 1.10; ), its significance increases with adjusted analysis. Incorporating all of the individual- and state-level covariates into the model leads to a firmer inverse association between the presence of state-mandated written consent and the likelihood of an individual’s reporting an HIV test in the prior 12 months (OR=0.85; p<0.001, 95% CI=0.80, 0.90).
The average marginal effect—the average of the discrete change for the full sample—was also estimated for residing in a state with a statute requiring written informed consent compared to residing in a state without such a statute. The average partial effect was −0.02 (p<0.001, 95% CI= −0.03, −0.01)—that is, the probability that an individual reports having had a recent HIV test will be 2 percentage points lower in states with a statute requiring written informed consent than in states without such a statute. If the baseline percentage (or probability) of individuals reporting no recent HIV testing is 16.8% in states without a written informed-consent statute, then written informed-consent statutes are associated with an 11.9% reduction in HIV testing, based on the estimated average marginal effects.
The association between written informed-consent requirements and testing differed according to the respondent’s self-reported HIV risk factors, race/ethnicity, and educational attainment (p-values for interaction=<0.001). Those living in a state with a requirement for written informed consent were significantly more likely to report a recent HIV test if they self-reported having an HIV risk factor compared to those who did not report such risk factors. Living in a state with a requirement for written informed consent was also associated with significantly lower rates of testing among respondents who were non-Hispanic white (OR=0.77; 95% CI=0.71, 0.82) or Asian (OR=0.73, 95% CI=0.53, 0.99) but not among respondents who were non-Hispanic black (OR=0.99; 95% CI=0.86, 1.14) or Hispanic (OR=1.14; 95% CI=0.94, 1.40). Similarly, the association between living in a state with a requirement for written informed consent and lack of testing varied according to a respondent’s level of education: The effect of a requirement for written informed consent was greater among respondents with higher levels of educational attainment (OR=0.79; 95% CI=0.72, 0.88 among college/technical school graduates) than respondents with lower levels of educational attainment (OR=1.06; 95% CI=0.86, 1.31 among non–high school graduates).
The results from the sensitivity analysis were similar. When states requiring written informed consent in more restricted medical settings were added to the group of states requiring written informed consent in all settings, the association remained negative between the presence of state-mandated written informed consent and the likelihood of an individual’s reporting an HIV test in the prior 12 months (OR=0.87; p<0.001, 95% CI=0.82, 0.92).
Finally, the analyses were also conducted including data for Washington DC (which was excluded because data were not available on federal funding for HIV prevention). Including these additional 2979 respondents in the analyses did not change the main results.