The quantitative component of this study revealed no statistically significant decline or increase in the number of HIV tests or diagnoses among MSM after the media releases in question. These findings suggest that, after the media releases of interest, at the population level, HIV testing continued without significant change among MSM, who are persons in the second group of our population health HIV prevention framework; i.e., members of a sexual networks with high HIV incidence and prevalence.d See Figure 1. As Dodds  noted, however, significant changes in patterns of HIV testing may be unlikely because of the varying and conflicting influences that affect persons’ decisions to undergo HIV testing. Observing no changes should, accordingly, have been an expected outcome.
Moreover, this lack of statistically significant change in HIV testing and HIV diagnosis numbers should not have been surprising considering recent data on nondisclosure prosecutions. For one, Adam  reported that the majority of his Toronto-based participants supported nondisclosure prosecutions, thus highlighting that many persons’ HIV testing practices might not be affected by these prosecutions. Likewise, research from the UK generated similar results: Dodds  reported that, in a survey of 8252 gay men, the majority of participants (n = 4667 or 56.5%) felt that prosecutions were appropriate in situations of HIV transmission when HIV-positive serostatus disclosure had not occurred. More specifically, O’Byrne and colleagues , in their preliminary analysis of an Ottawa-based research that involved a convenience sample of 441 participants who identified as MSM, found that most respondents neither felt that “nondisclosure prosecutions made them afraid to talk to nurses or physicians about their sexual practices”, nor that these prosecutions “affected their decisions to undergo HIV testing”. In combination, these studies suggested, as was observed in our study, that one might not detect a population-level change in HIV testing and HIV diagnoses as a result of nondisclosure prosecutions or related media releases.
Burris and colleagues’  work about the behavioural effects of nondisclosure prosecutions further supported this conclusion. In their survey of 248 persons who lived in Chicago and 242 persons who lived in New York City, these authors found that criminal laws which “regulat[e] sexual behaviour of HIV-infected” persons do not appear to affect most people’s sexual practices (p468). Similarly, Galletly and colleagues’ work , which involved statistical analyses of anonymous survey responses from a convenience sample of 479 persons who live in New Jersey, corroborated the idea that the criminal law has little impact on sexual behaviour. Whether or not Burris and colleagues’  and Galletly and colleagues’  findings apply to both HIV testing and the situation in Canada, however, remains unclear.e
Furthermore, in light of our above-noted findings, two other aspects of O’Byrne and colleagues’  research are noteworthy. First, in O’Byrne and colleagues’  study, among participants who were HIV-negative or unsure of their serostatus, those who noted that nondisclosure prosecutions either (a) affected their decisions to undergo testing or (b) made them reluctant to speak with nurses or physicians were more likely to have noted that they engaged in unprotected penetrative anal intercourse (x
2, 5.47 (1), p = 0.019). Second, among the same group (HIV-negative or unsure of their serostatus), more respondents who noted they had never been tested for HIV indicated that nondisclosure prosecutions affected their testing practices (x
2, 12.19 (1), p < 0.001) .
Because our findings are based on people’s actions, and not their self-reported behaviour, our population-level testing numbers raise interesting questions about O’Byrne and others’  survey data. For one, the changes in our population-level HIV testing and diagnosis numbers might not have been significant because the persons who reported that these prosecutions affected their testing practices were so few in O'Byrne and colleagues'  study that they, therefore, would have had no impact on the population-level figures. In such a case, even if changes in testing occurred among a subset of MSM engaging in more unprotected sex and less frequent HIV testing, these changes may not have been noticeable at the aggregate-level. Alternatively, it could be that the persons who might were most influenced by nondisclosure prosecutions had previously decided not to be tested for HIV, thus indicating that such prosecutions are simply one additional reason for not undergoing HIV testing by persons already not accessing testing services As part of this, one must remember that, while the media releases of interest herein constituted what we believe are the first of such media releases in the local context, nondisclosure prosecutions from other Canadian regions have been occurring and have been publicized since the early 1990s. In this case, persons already not undergoing HIV testing due to nondisclosure prosecutions would not have induced any changes in our data. In addition, it is equally possible, in relation to nondisclosure prosecutions and HIV testing, that self-reported and actual behaviour do not align. In such case, what people say they will do and what they actually do may be different. It may simply be, furthermore, that people’s post hoc rationales differ from their initial, or at least originally stated, motivations. In any case, such possibilities warrant further investigation, which should elicit the information to either substantiate or refute the Supreme Court of Canada’s claims that nondisclosure prosecutions do not affect HIV testing practices.
While the quantitative component of our study did not point to any significant relationship(s) between nondisclosure prosecutions and public health HIV prevention efforts, our qualitative findings highlighted that nondisclosure prosecutions may have detrimental effects on both PHAs and prevention. Specifically, our data suggested that such criminal proceedings could negatively affect people aware they are HIV-positive, and could potentially undermine their HIV prevention abilities. This assertion arose from the qualitative finding that nondisclosure prosecutions have made some PHAs less willingly to both seek assistance from, and speak candidly with, public health workers about how to diminish onward HIV transmission. According to participants in this study, the reluctance to seek out and speak with such health professionals related to their belief that the local public health department shares information with police about instances of serostatus nondisclosure without requiring production notices, e.g., warrants.
When we considered this finding using the population health HIV prevention framework in Figure 1, it raised important questions about the extent to which nondisclosure prosecutions undermine the abilities of local public health officials to work with PHAs who wish to discuss what one of our participants called the “issues” he has had with safer sex and HIV prevention. Our qualitative data, therefore, suggested that nondisclosure prosecutions and their related media releases likely undermine HIV prevention involving the second highest priority group within our HIV prevention framework; i.e., PHAs who know they are HIV-positive, and who engage in practices that transmit HIV. See Figure 1. This scenario is undesirable from an HIV prevention perspective because, while Marks and colleagues argued that persons unaware of being HIV-positive are likely implicated in most HIV transmission, persons aware of their serostatus—particularly when they are asking for assistance to prevent HIV transmission—constitute an important group for HIV prevention [12, 13].
This finding that nondisclosure prosecutions may hinder health professionals’ abilities to provide accurate, open, and supportive care for PHAs aligns with previous research. It corroborates Mykhalovskiy’s findings about how, on the one hand, public health workers have begun to undertake HIV prevention counselling with PHAs with an “eye to the law”, and how, on the other hand, this approach has impaired the abilities of these health professionals to, first, discuss serostatus disclosure and, second, counsel PHAs about how to address the “challenges they may be facing disclosing their HIV-positive status” (p.672) . Similarly, our findings correspond with the issues raised during O’Byrne and Gagnon’s  discussion groups with frontline health staff in Ottawa: Nurses perceived that nondisclosure prosecutions have negatively affected their therapeutic relationships with PHAs, and, consequently, have undermined their abilities to provide appropriate and timely nursing care for these patients.
Moreover, other indications from our HIV-positive participants that criminal prosecutions made them wish they had undergone anonymous HIV testing, which would have resulted in the local public health department not possessing an identifiable record of their HIV-positive serostatus, might align with the current reality in Ottawab. Local HIV testing and diagnosis figures indicated that anonymous HIV testing has the highest positivity rate (1.1% for anonymous testing, versus 0.26% for nominal testing), thus suggesting that people may select anonymous HIV testing if they believe the test will be positive . Such findings may, furthermore, align with Lowbury and Kinghorn’s  arguments that nondisclosure prosecutions constitute a “clear disincentive to testing” (p.666), with the outcome not being an outright refusal to undergo HIV testing, but rather, the selection of anonymous rather than nominal HIV testing. Before absolute conclusions can be drawn about anonymous HIV testing using our data, however, further research is required to more precisely discern the relationship(s) between nondisclosure prosecutions and this HIV testing method. While our study examined if testing and diagnoses changed, more specific research should explore if nondisclosure prosecutions affect who undergoes testing, when they do so, and by which means. The answers to these questions could both reconcile the seemingly paradoxical findings which emerged in this study, and explain the partial differences between this study and those which preceded it.
Lastly, our other qualitative finding that some participants described PHAs as sexually promiscuous latent criminals raised a few important points for consideration. First, our participants’ linkage of HIV and the criminal law cast light on a potentially new aspect of the contemporary understanding that surrounds HIV and persons living with this virus. While the association between HIV and sexuality has long been documented [18, 25–28], there has yet to be empirical accounts about people assuming a natural association between HIV, the criminal law, and latent criminal tendencies of PHAs. More research is needed on this topic, particularly from a sociological perspective.
The second aspect of our finding, that some participants linked HIV and the criminal law was that this belief added a new dimension to current knowledge about HIV stigmatization. From an HIV prevention perspective, because the extant literature identified that HIV stigmatization compromises PHAs’ abilities to maintain safer sex efforts with serodiscordant partners, nondisclosure prosecutions may be an unwanted addition [25–29]. Indeed, it may be a new item that can exacerbate HIV transmission [25–29]. In addition to the stigmatization that is already associated with an infection that is primarily sexually transmitted,  in our research, PHAs now appeared to be considered latent criminals; our participants used language, such as, “dangerous” and “murderer” to label PHAs, and made these associations based on serostatus, not sexual practices or behaviour. Therefore, because HIV stigmatization often corresponds with an increased probability of HIV transmission, our findings raise important questions about whether our participants’ association of PHAs with the criminal law, i.e., assumptions regarding latent criminality, could undermine PHAs’ psychosocial wellbeing, and, in turn, exacerbate HIV transmission at the population level.
Our research results thus corresponded with Dodds’ findings , which suggested that a hurdle to contemporary HIV prevention is the “direct conflict between the attitude changes that are necessary … for men to avoid acquisition of HIV and the way that these same men use popular narratives about criminal prosecutions to support their view about the way the world should be” (original emphasis, p.513). Our participants’ perceptions that PHAs were dangerous criminals supported Dodds’  assertions that some MSM prefer to “blame” PHAs for HIV transmission, rather than focus on their personal responsibilities in HIV prevention. As noted by Dodds , such perceptions likely hinder population health HIV prevention efforts. Further research must clarify the precise linkages between this potentially new aspect of HIV-related stigmatization, and persons’ actual HIV prevention efforts. In the meantime, however, a potentially problematic situation exists for PHAs and for HIV prevention.
These conclusions must be interpreted in light of certain limitations. Results cannot be generalized to all groups vulnerable to HIV described in the population health HIV prevention framework because this study only analyzed qualitative and quantitative data for MSM. This study recruited interview participants at venues frequented by these men, resulting in self-selection of men more connected to the gay community. Thus, participants did not fully represent the opinions of all MSM in Ottawa. Additionally, an inclusion criterion for the qualitative component of this study was awareness of the media release, whereas the extent of awareness of the media release was not assessed among the MSM who were tested for HIV after the media release. Thus, causation cannot be attributed to the media release or nondisclosure because other factors may have affected testing volume. Furthermore, the population of MSM vulnerable to HIV acquisition is unknown and so the data cannot be contextualized with a denominator. The unadjusted numbers for MSM tests are subject to reporting on the laboratory requisition, while the adjusted numbers are subject to the methodology used by the LEP. The 2008-2010 average annual response rate for the LEP survey sent to physicians with newly diagnosed patients from Ottawa was 55%, resulting in 69% of tests with known risk factor information prior to adjustment (2011 data not yet available).
Need for further research
More studies are needed to explore the effects of nondisclosure prosecutions on population level HIV prevention. Researchers must undertake larger quantitative cohort studies that examine fluctuations in testing patterns over time, and in relation to legal changes and the testers’ attitudes and behaviours regarding nondisclosure and sexual practices. Data should also be gathered about the actual practices of individuals who seek testing, rather than just about their exposure information. This would allow for discrimination between, to use public health terminology, the low-risk and high-risk testers within each surveillance population; indeed, not all MSM who undergo HIV testing are at high-risk for HIV. In combination, such data could inform both population level HIV prevention strategies and the current legal context. Moreover, further research is needed to determine why approximately half of persons tested for HIV in Ottawa have no identified risk factors on the laboratory requisition because this risk-factor information helps establish HIV prevention policy, programming, and practice within public health departments.