Supervised injecting facilities: how much evidence is enough?

Supervised injecting facilities (SIF) have been shown to be highly effective interventions in reducing the harms associated with injecting drug use. To date a total of 28 methodologically rigorous studies have been published in leading peer-reviewed medical journals (1). This growing body of evidence indicates that SIFs are associated with reductions in needle and syringe sharing, overdoses, public injecting and numbers of publicly discarded syringes (2 - 6), increased uptake of drug detoxification and addiction treatment pro- grammes (7) and have not led to increases in drug- related crime or rates of relapse among former drug users (8,9). (460%) Kings Cross residents agreed with the estab- lishment of the Sydney MSIC (68% in 2000, 78% in 2002 and 73% in 2005; p-trend ¼ 0.06). There is also significant support for the MSIC from the local business community. Among the 629 Kings Cross business operators surveyed at the three time-points, there was a statistically significant increasing trend in favour of the establishment of the service (i.e. 58% in 2000, 63% in 2002 and 68% in 2005; p-trend ¼ 0.03). A review of drug consumption facilities found that their establishment in local neighbourhoods led to major public debate in most of the 36 European cities where they operate (4). Results from our evaluation indicate that, in the Australian context, approximately three in five local residents and businesses agreed with the establishment of the Sydney MSIC prior to its opening and that this level of support has been sustained over time. Our results also suggest that local community members are cognisant of both potential public health (perceived reduction in blood-borne viral infections and overdose) and potential public amenity advantages of the Sydney MSIC (13). In a climate of ongoing political and policy debates surrounding harm reduction strategies, these data contribute to the growing evidence base supporting the benefits of SIFs. However, while community support may be an important determinant of political will, it is not a measure of efficacy and should not be the litmus test by which SIFs or, indeed, any health intervention, are evaluated. A well-designed and conducted randomised controlled trial (level 1 evidence) remains the best study design for determining a causal relationship between a public health intervention and its putative outcomes. The scientific, practical and ethical issues involved in applying this methodology to evaluating complex public health interventions such as SIFs mean that the