Methods for providing an earlier warning of emerging drug trends

Abstract

There is increasing international interest in the improved timely identification and reporting of new and emerging drug trends. Research directed at examining drug trends is particularly important in the contemporary context where a variety of new substances are available to increasingly diverse populations and in an expanding range of settings. Against this background, the primary aim of this study is to take up the theoretical and empirical challenge of developing a methodology and a model for earlier identification and earlier warning of emerging drug trends. Based on the Bergen Earlier Warning System (BEWS), the objectives of this thesis are to examine: key features required for a city level drug earlier warning system; drug measures or ‘indicators’ to the identify new and changing patterns in use; major challenges associated with data reliability, validity and triangulation; use of psychometrics to improve system validity; and how the model can report on the illicit use of medicines. The Bergen Earlier Warning System (BEWS) was established in 2002 drawing on principles derived and lessons learned from a European feasibility study conducted in 1998 (EMCDDA, 1999; Griffiths et al., 2000). The system is multi-source (drawing on a range of indicators of drug use, mortality, morbidity and market characteristics), incorporates a mix of approaches (routine data collection, media monitoring, key informant study) and utilizes mixed methods (both quantitative and qualitative). Results demonstrate that the use of the principles for psychometric scale development can assist with enhancing the validity of emerging drug trend monitoring systems (EDTMS), providing a coherent and explicit developmental framework, that has wider applications. A five stage review and refinement process is presented for the development of an EDTMS that focuses on strong construct, congruent and external validity. The use of a ‘pragmatic’ research paradigm is argued for in conjunction with a mixed method, multi indicator approach. Five criteria for indicator selection are presented and it is demonstrated how these may be utilized to attribute weightings to individual sources in the analysis phase. Key issues relating to reliability and validity in the monitoring of emerging drug trends are identified, specifically challenges at the level of system construction, as well as reliability challenges at the level of individual data sources. A general approach to analysis of mixed data is proposed based on overt triangulation practices. A specific example of analysis is provided with reference to BEWS’ use of an aggregated differential score. Finally, it has been shown how the BEWS can be used to monitor trends in alcohol, drugs and the street use of medicines and report back to policy makers and practitioners for early intervention purposes.