[Are the responses of clients with psychiatric and addiction disorders using services for the homeless valid?]

To verify the validity of self-reported data on service use from clients with mental or substance abuse disorders in Montreal and Quebec services for homeless individuals.

To compare the self-reported data from the Enquête chez les personnes itinérantes (Fournier, 2001) on health service use with official data from Quebec health services (MEDECHO and RAMQ).

The analysis shows a moderate-to-high level of concordance between the self-reported and the official data. Almost every item analyzed presents moderate but significant intraclass correlation coefficients for general and psychiatric hospitalization and use of psychiatric medication, but lower and nonsignificant coefficients for medical hospitalization. Participant characteristics such as mental disorders, homeless status, and substance abuse problems do not seem to have an impact on data validity.

The answers on health service use from individuals with mental health problems, homeless status, or substance abuse problems are generally valid in the results presented. Thus the self-reported data from these individiuals seems to be a generally valid source of data and an affordable one for research on service use or other domains.