It is widely developed the idea of communitary attention of serious mental disorders, instead of being recluded in long stay hospitals. But tomake this project real it is needed a complex sociosanitay support: psychosocial rehabilitation units, protected dwellings…and nowadays some people are still discharged to long term hospitals.

The objective of this work is to demonstrate that clinical characteristics are not determinant and that the main reason that supports the decision is the insufficent social resources spread.

279 inpatients that entered a Psychosocial rehabilitation unit from January 2005 to April 2009 are selected. There are two comparison groups depending on the derivation to long term units. There are compared different variables taking the HoNOS scale as measurement tool. Ths stathistical comparison test used is the Fisher exact test.

HoNOs scale items:

1. 1 Disruptive behaviour

2. 2 Non-accidental self-injury

3. 3 Alcohol/drug taking

4. 4 Cognition

5. 5 Physical illness

6. 6 Hallucinations/delusions

7. 7 Depressed mood

8. 8 Other mental problems

9. 9 Relationship

10. 10 Daily activity

11. 11 Living conditions

12. 12 Activities/occupation

There is not statistical significance when clinical items of the HoNOS scale (item 1 to 8) are compared, instead of item 5 (p=0,001). There appears significance when comparing social items and specially item 9 (p=0,031),11 (p=0,002) and 12 (0,026).

Nowadays there is no difference in the clinical profile of inpatients when derivation. The fact of not having adequate living conditions is determinant in the decision of derivation, showing that socio sanitary resources are insufficient for an optimal communitary system.