In Vancouver, Canada the popularity of crack cocaine and ease of administration through smoking has made public crack cocaine use a common feature of the streets in the city's drug use epicentre, known as the Downtown Eastside [17]. Public crack cocaine smoking is posing a growing burden for law enforcement agencies responsible for maintaining public order [18]. In addition, the health and social harms associated with crack cocaine smoking are extensive. Compared to other drug user populations, crack users are more likely to engage in risky behaviors [19-21] and illegal activities [22,23] and to experience homelessness [14] and health problems [14,24-27], yet are less likely to access health and social services [28]. It has also been recently documented that daily crack cocaine smokers are at a four-fold greater risk of contracting HIV compared to their drug using peers who smoke crack cocaine less often or not at all [16].

Given the dramatic rise in crack cocaine smoking and the public order and public health concerns associated with it, the need for targeted interventions for people who smoke crack cocaine is unambiguous. One potential intervention that is receiving increasing attention from public health officials, health researchers and local community groups is supervised drug consumption facilities analogous to supervised injection sites but that accommodate crack smoking and distribute drug consumption materials specific to safer inhalation (such as sterile crack pipes, mouthpieces, and screens) [16,29-33]. The Canadian Institutes of Health Research recently approved funding to conduct a randomized control trial to evaluate the impact of a supervised inhalation facility on access to medical and social services, particularly addiction treatment, among Vancouver-based crack cocaine smokers [34].

In our first analysis among crack cocaine smokers, the outcome of interest was using drugs (non-injection) in public areas in the last six months. As in previous analyses, public areas included city streets, parks, public washrooms, parking lots, clubs or bars and abandoned buildings. To characterize our outcome of interest we a priori selected a range of socio-demographic and behavioural variables we hypothesized might be relevant to smoking crack cocaine in public areas. This selection was informed by the 'risk environment framework' and previous analyses among street-involved drug users highlighting connections between social, structural and environmental level factors and risky drug consumption practices [14,31,39-41]. Variables included: age (per year older); gender (female vs. male); Aboriginal Ancestry (yes vs. no); limited access to private space, defined as answering "no" to the question: "Do you have a private indoor space for socializing with friends and acquaintances?" or reporting that the number of guests they were allowed to have in their residence at one time was restricted to less than three (yes vs. no); daily cocaine injection (yes vs. no); daily heroin injection (yes vs. no); daily crack cocaine smoking (yes vs. no); non-fatal overdose, self identified by participants (yes vs. no); encounters with police in the last month, defined as being questioned, searched or stopped by police (yes vs. no); being a victim of violence defined as being physically assaulted (yes vs. no); sex trade involvement, defined as exchanging sex for money, shelter, drugs or other commodities (yes vs. no), and participation in drug dealing (yes vs. no). Unless otherwise stated, all drug use and behavioural variables refer to the previous six month period.

In a second analysis, we sought to assess and identify predictors of willingness to use a supervised inhalation room. Because we were particularly concerned with public drug use we restricted our sample to crack cocaine smokers that reported recently using non-injection drugs in public areas. To measure willingness we asked participants "If there was a safe place to smoke your drugs (ventilated inhalation room), close to where you buy or use, would you use it?"

Variables of interest for our second analysis were also selected a priori based on factors we hypothesized might be associated with willingness to use an inhalation room. These included age (per year older); gender (female vs. male); Aboriginal Ancestry (yes vs. no); limited access to private space, as defined above (yes vs. no); drug scene exposure, defined as spending an average of seven or more hours on the street each day in Vancouver's drug use epicentre in the previous six months (yes vs. no); most drug use in public areas, defined based on reports that public locations were where they most frequently used drugs (yes vs. no); daily crack cocaine smoking (yes vs. no); risky pipe sharing, defined as reporting sharing a crack pipe or mouthpiece in the same six month period as having burns or sores on their mouth (yes vs. no); encounters with police in the last month (yes vs. no); and being a victim of violence (yes vs. no). As above, unless otherwise stated, all drug use and behavioural variables refer to the previous six month period.

For our second analysis, the demographic and behavioural characteristics of public crack cocaine smokers stratified by willingness to use a supervised inhalation room are presented in Table ​Table3,3, and the univariate results of factors associated with willingness to use a supervised inhalation room are presented in Table ​Table4.4. The results of the multivariate logistic regression for factors associated with willingness to use a supervised inhalation room are also shown in Table ​Table4.4. Factors that remained independently associated with willingness included: female gender, risky pipe sharing and recent encounters with police (see Table ​Table44.).

The profile of public crack cocaine smokers as higher-intensity drug users who have interactions with the criminal justice system is reflective of previous findings describing public injection drug user populations [1,5]. The association between drug dealing and public crack use may reflect the increased amount of time individuals spend on the street when engaged in street-level drug dealing. It may also be a function of the accessibility of drugs and additional resources gained through drug dealing which may lead to greater drug consumption and hence a greater likelihood for consuming in public areas [23].

Our finding that 71% of public crack cocaine smokers are willing to use an inhalation facility also supports previous willingness estimates conducted among the general population of Vancouver-based illicit drug users and suggests that an intervention of this nature will likely reach the target population [36]. The high degree of willingness that this study found among public crack cocaine smokers to use an inhalation facility suggests that, like supervised injection facilities, these interventions are likely to successfully encourage public drug users to relocate to indoor venues.

The increased likelihood of being willing to use an inhalation facility among female participants may reflect heightened vulnerability of women involved in street drug use and it is noteworthy that Vancouver's supervised injection facility has had success in attracting vulnerable female drug users and providing them with safer alternatives to street-based drug using venues. In previous research female IDU have described the unique role that Vancouver's supervised injection facility has played in promoting their physical security and health safety [42].

Interestingly, one of the common features among both public crack cocaine smokers and those who are willing to use a supervised inhalation facility is their elevated likelihood of recently having encounters with law enforcement. This suggests that public crack cocaine smokers who are the subject of law enforcement attention are very willing to relocate to alternative off-street and health-focussed environments if they were made available. Indeed, our data indicate that 81% of public crack cocaine smokers who have had a recent encounter with police are willing to use a supervised inhalation facility.

A key implication of these findings is that there is a large demand for supervised inhalation rooms among individuals that are potentially key contributors to drug-related street disorder. The association between public crack smoking and encounters with police suggests that interventions of this nature are likely to target a critical sub-population of drug users and could be a valuable tool for police in the management of street disorder. Previous studies have found that Vancouver police regularly refer public injection drug users to the local supervised injection facility [43]. Since our analysis indicates that local police are already frequently interacting with public crack smokers the establishment of a supervised inhalation facility could provide a unique opportunity for police to direct this vulnerable group to a low-threshold service where they will have opportunities to be linked with appropriate health and social services.

It is critical to note that although this study suggests that supervised inhalation facilities could aid in the reduction of public disorder, drug consumption facilities do not address the route causes of street disorder and are not appropriate substitutes for other essential health and social interventions such as supportive housing and addiction treatment. To be effective supervised inhalation facilities should be integrated into broader comprehensive approaches to addressing the problems associated with illicit drug addiction.

This study has a number of limitations. Firstly, VIDUS is a community recruited non-randomized sample and therefore our findings may not be generalizable to other settings. If supervised inhalation facilities are being considered in other settings, willingness studies should be conducted among the local target population and should not rely on the findings emerging from our setting. The generalizability of our findings is also limited by our study sample which was restricted to individuals with a history of injection drug use. Crack cocaine smokers who did not have a history of injection drug use were not eligible for our study. Given the harms associated with injection drug use we anticipate that if a selection effect were present it would likely bias our sample towards high risk drug users, suggesting that this group would be an appropriate target population for public health intervention. We should also note that among our study sample daily crack cocaine smoking was significantly more common than daily injecting, suggesting that despite the requirement of a history of injecting, our sample represents a primarily crack cocaine smoking population. Secondly, some of our measures relied on self-report and could be vulnerable to socially desirable reporting. This would have likely been of most relevant to our measure of willingness, since respondents might perceive a pressure to report being willing to engage with low-threshold services of this nature given the widespread activism among local drug users in our study setting to implement supervised drug consumption facilities [32]. While it is possible that some respondents may over-report willingness, a previous study comparing measures of willingness to use a supervised injection facility before it was established with later reports of actual attendance after an injection facility was established suggests that willingness measures are good predictors of later behaviour among this population [44]. Lastly, socially desirable reporting could have influenced reports of stigmatized behaviour, such as public drug use leading to an underestimation of public crack smoking. If social desirability was an issue in our study we suspect our finding would be a conservative indication of the prevalence of and harms associated with public drug use among crack cocaine smokers.