The recent article by Salize & Dressing (Reference Salize and Dressing2004) reported that frequencies of compulsory admissions vary remarkably among countries in the European Union, from 6 per 100 000 citizens in Portugal to 218 per 100 000 in Finland. These findings are not surprising given the large differences in the laws, mental health acts, and legal instruments of the countries but they are astonishing given the much smaller differences in psychiatric morbidity. These differences show that the number of involuntary admissions is a result of a complex set of still poorly understood legal, political, economic, social and multiple other factors (Reference Salize, Dressing and PeitzSalize et al, 2002). However, data on the effectiveness of coercion measures are lacking and there is no evidence base for involuntary commitment. The few studies have focused mainly on out-patient commitment and show mixed results (Reference Swanson, Swartz and WagnerSwanson et al, 2000; Reference Steadman, Gounis and DennisSteadman et al, 2001; Reference Swanson, Swartz and ElbogenSwanson et al, 2003).

The absence of an evidence-based model for the use of coercion in psychiatry is partly due to ethical difficulties in studying coercion measures, for example, using randomised controlled trails. We need to find ways to overcome these difficulties, for example by assessing the effectiveness of involuntary admission in those who pose relatively little danger to themselves and others. Results of these studies need to be taken into account in the current debate on the use of coercion measures. It is likely that certain groups of patients benefit more from specific coercion measures than others. Patients with psychotic disorders with severe social breakdown and lack of motivation for treatment probably benefit more from cerocion measures than those with personality disorders. International comparative studies are needed to assess the effects of different laws on outcomes, for example laws using criteria of danger v. those using need for treatment criteria. Valid and reliable instruments are needed when deciding to use coercion; these should include assessment of the severity of psychiatric disorder, danger to self or others and motivation for treatment. Researchers active in this field could form collaborative (inter)national working groups on pressure for treatment and coercion in psychiatry.