The relevance of a good therapeutic alliance development between patient and therapists in the treatment success has been documented in more than 3 decades of empirical research. In the case of the treatment of severely disturbed patients, the alliance construction process involves particular characteristics determined, in part, by the patients inability to form safety bonds with others and because of, usually, various therapeutic figures are engaged in their treatments. The present work offers a general review of the most important empirical evidence about the therapeutic alliance process in institutional context treatments (i.e., hospitalization, therapeutic community), introduces the concept of Institutional Therapeutic Alliance (ITA) - clinical and empirical phenomenon that accounts for the working bond between the patient and the therapeutic staff perceived as a whole - and reports the major results of a longitudinal study conducted to assess the ITA and explores the relationship with treatment outcomes.

55 day-hospital patients take part in the research and were evaluated at admission, before one week, at discharge and after 3 months. The assessment battery included: Symptom Check List (SCL-90), Global Assessment Scale (GAS), Multidimensional Social Perceived Support Scale (MSPSS), Institutional Working Alliance Inventory (IWAI) and Subjective Distance Scale (SDS). The results show that ITA is positively correlated with symptomatic reduction at discharge and negatively associated with patient's re-hospitalization after 3 months.

The work concludes by discussing, from a clinical point of view, the promoting and obstructing alliance factors linked with the patient, the staff and their relationships.