BACKGROUND
Governments aim to increase treatment participation by problematic drug users. In the UK this has been achieved by fiscal investment, an expanded workforce, reduced waiting times and coercive measures (usually criminal justice (CJ) led). No assessment of these measures on treatment outcomes has been made. Using established monitoring systems we assessed trends in 'dropped out' and 'discharged drug free' (DDF), since the launch of the national drug strategy, and rates of treatment re-presentation for these cohorts.

METHODS
A longitudinal dataset of drug users (1997 to 2004/05, n = 26,415) was used to identify people who dropped out of, and were DDF from, services for years 1998 to 2001/02, and re-presentations of these people in years to 2004/05. Trends in drop out and DDF, baseline comparisons of those DDF and those who dropped out and outcome comparisons for those referred from the CJ system versus other routes of referral were examined using chi square. Logistic regression analyses identified variables predicting drop out versus DDF and subsequent re-presentation versus no re-presentation.

RESULTS
The proportion of individuals dropping out has increased from 7.2% in 1998 to 9.6% in 2001/02 (P < 0.001). The proportion DDF has fallen from 5.8% to 3.5% (P < 0.001). Drop out was more likely in later years, by those of younger age and by CJ referrals. The proportion re-presenting to treatment in the following year increased from 27.8% in 1998 to 44.5% in 2001/02 (P < 0.001) for those DDF, and from 22.9% to 48.6% (P < 0.001) for those who dropped out. Older age and prior treatment experience predicted re-presentation. Outcome (drop out or DDF) did not predict re-presentation.

CONCLUSION
Increasing numbers in treatment is associated with an increased proportion dropping out and an ever-smaller proportion DDF. Rates of drop out are significantly higher for those coerced into treatment via the CJ system. Rates of re-presentation are similar for those dropping out and those DDF. Encouragingly, those who need to re-engage with treatment, particularly those who drop out, are doing so more quickly. The impact of coercion on treatment outcomes and the appropriateness of aftercare provision require further consideration.