Background: Novel psychoactive substances (NPSs) contribute to the public health impact of substance misuse. This report provides research evidence addressing 11 research questions related to NPSs, covering types, patterns and settings of use; supply sources; and implications for policy and practice.

Methods: The study used a conceptually linked three-phase mixed-methods design with a shared conceptual framework based on multiple-context risk and protective factors. Phase 1 was a quantitative phase involving secondary data analysis of the longitudinal Belfast Youth Development Study (BYDS), a latent class analysis using the 2039 BYDS participants. Phase 2 was an extensive qualitative analysis via narrative interviews with participants, sampled from BYDS, drug/alcohol services and prisons, to explore NPS use trajectories. Phase 3 was the final quantitative phase; generalisability of the shared risk factor part of the model was tested using the manual three-step approach to examine risk factors associated with latent class membership. The quantitative and qualitative analyses were integrated, thus allowing emerging findings to be further explored.

Results: The data suggest that NPSs have a place within a range of polydrug use trajectories. Models showed no distinctive NPS class, with no clear evidence of differential risks for NPS use compared with the use of other substances. From the qualitative analysis, a taxonomy of groups was derived that explored how and where NPSs featured in a range of trajectories. This taxonomy was used to structure the analysis of factors linked to use within a risk and protective framework. Drivers for use were considered alongside knowledge, perceptions and experience of harms. Suggestions about how interventions could best respond to the various patterns of use – with special consideration of synthetic cannabinoids (SCs), including how they relate to the use of heroin and the potential for NPSs to operate as a ‘snare’ to more problem use – were also presented.

Limitations: The study was conducted during 2016/17; generalisability beyond this sample and time point is limited. The level of missing data for some of the BYDS analysis was a limitation, as was the fact that the BYDS data were collected in 2011, so in a different context from the data collected during the narrative interviews. The Psychoactive Substances Act 2016 (Great Britain. Psychoactive Substances Act 2016. London: The Stationery Office; 2016) came into force during qualitative fieldwork and, although not particularly influential in this study, may be influential in future work. It is acknowledged that many of the data related to SCs and mephedrone. Although drug use was measured by self-report, the strength of rapport within interviews, reflective diaries and methodological acceptability checks helped to mitigate self-report bias.

Conclusions: NPSs continue to present significant challenges for legislation and monitoring, researching and developing interventions. Understanding of usage patterns remains poor, with most information based on populations and settings where problems have already occurred. This research contributes to the evidence base by providing much needed further empirical data on the lived experiences of NPS users across a range of settings. In the light of these data, implications for policy and practice are discussed.

Future work: Future research must generate improved epidemiological data on the extent, patterns and motivations for use longitudinally. The uniqueness of the information concerning SC use points to a specific set of findings not evidenced in other literature (e.g. intensity of SC withdrawal). Future research should focus on the symbiotic link between SC and heroin use.