clustering of individuals within counties, hierarchical modeling wasused to generate unbiased and efficient estimates, as well as properstandard errors. This modeling takes into account (a) the influenceof different sample sizes across counties and (b) the dependenceamong individual outcomes clustered within the same county.Results: HLM showed both significant individual and contextualeffects. Race (i.e. African American), homelessness, and co-morbidity of a substance abuse significantly decreased the odds ofconforming to standards of continuity of care. Co-morbidity ofsubstance abuse provided the most substantial decrease in ratesof conformance. There was very little consistency in odds ofconformance rates for contextual level variables. Conformance tostandards of quality continuity of care decreased if the communityhad higher rates of poverty, had fewer community mental healthcenters, and was classified as rural.Discussion: The overall findings show that basic standards of qualitycontinuity of care are not being met for some adults withschizophrenia in parts of the USA. The results highlight substantialdisparities in basic standards of quality continuity of care. Con-formance rates are significantly lower for African Americans and thehomeless. Particularly striking were the results for those with a co-occurring substance abuse disorder as they had the lowestconformance rate of any individual based factor. Contextual factorssignificantly associated with lower conformance rates were numberof available mental health centers, level of poverty, and whether thecommunity is rural. Such disparities leave certain individuals andcommunities vulnerable to a host of negative consumer outcomes.The results demonstrate a clear need for quality of care monitoringparticularly for specific groups and communities. The results canserve as guidance for future research in disparities in quality of care.

doi:10.1016/j.schres.2010.02.745

Poster 251IMPACT OF INITIAL HOSPITALIZATION ON 3-YEAR OUTCOME INPATIENTS WITH FIRST EPISODE PSYCHOSIS IN HONG KONG

May M.L. Lam1, Jennifer Y.M. Tang1, Cindy P.Y. Chiu1, Christy L.M.Hui1, Gloria H.Y. Wong1, C.W. Law2, Dicky W.S. Chung3, Kathy P.M.Chan4, Steve Tso5, S.F. Hung4, Eric Y.H. Chen1

1University of Hong Kong, Hong Kong; 2Queen Mary Hospital, HongKong; 3Tai Po Hospital, Hong Kong; 4Kwai Chung Hospital, Hong Kong;5Castle Peak Hospital, Hong Kong

Background: Psychotic disorders are among the 10 leading causesof disability-adjusted life years (DALYs). Increasing effort has beendevoted in identifying factors and interventional strategies asso-ciated with better outcome. The Early Assessment Service for youngpeople with psychosis (EASY) was launched in Hong Kong in 2001aimed to provide early detection and intensive intervention forpeople with first episode psychosis. This study compared the 3-yearoutcome of patients receiving the Early Intervention (EI) servicewith those who received standard care (SC) prior to the launch ofthe EI service. For the Early Intervention (EI) group, the impact ofinitial hospitalization was studied for effect on subsequent outcome.Methods: Seven hundred patients with first episode psychosisconsecutively enrolled in EASY programme were compared with700 matched historical control.Results: Fewer patients in the Early Interventionprogrammehad beenhospitalized at the first month of treatment. Patients in the EarlyIntervention programme had few number of compulsory admissions,fewer number of subsequent admissions and had shorter duration ofhospital stay compared with patients in historical control. Amongstpatients in Early Intervention programme, those who were not

hospitalized initially were found to have better functional outcome,fewer days in hospital, less suicide attempts.Discussion: The current study suggests that Early InterventionProgramme in Hong Kong is successful in reducing the number ofadmission and duration of hospitalization. Early Interventionprogramme should aim for community treatment and to minimizethe initial hospitalization in order to bring better outcome.

doi:10.1016/j.schres.2010.02.746

Poster 252ASSERTIVE COMMUNITY TREATMENT (ACT) AS PART OFINTEGRATED CARE VERSUS STANDARD CARE: A 12-MONTHTRIAL IN PATIENTS WITH FIRST- AND MULTIPLE-EPISODESCHIZOPHRENIA-SPECTRUM DISORDERS TREATED WITHQUETIAPINE IR (ACCESS TRIAL)

Martin Lambert1, Dieter Naber1, Benno Schimmelmann2

1University of Hamburg, Hamburg, Germany; 2University of Bern, Bern,Switzerland

Background: The ACCESS trial examined the 12-month effective-ness of continuous therapeutic assertive community treatment(ACT) as part of integrated care (IC) compared to standard care (SC)in a catchment area comparison design in patients with schizo-phrenia spectrum disorders (SSD) treated with quetiapine IR.Methods: Two catchment areas in Hamburg, Germany (UKE &AWR) with similar population size and health care structures wereassigned to offer 12-month ACT (& IC; UKE, n=64) or SC (AWR,n=56) to 120 first-episode patients and multiple-episode patientswith a history of relapse due to medication non-adherence. Primaryoutcome was the time to service disengagement. Secondaryoutcomes comprised medication non-adherence, improvements ofsymptoms, functioning, quality of life, satisfaction with care frompatients and relatives perspectives, and service use data.Results: 17 of 120 patients (14.2%) disengaged with service, 4 patients(6.3%) in the ACT and 13 patients (23.2%) in the SC group. The meanKaplanMeierestimatedtime in servicewas50.7 weeks in theACTgroup(95% CI 49.1-52.0) and 44.1 weeks in the SC group (95% CI 40.1-48.1).This difference was statistically significant (p=.0035). Mixed ModelsRepeatedMeasures indicated larger improvements for ACTcompared toSC regarding symptoms (p<.01), illness severity (p<.001), globalfunctioning (p<.05), quality of life (p<.05), and client satisfaction asperceived by patients and family (both p<.05). Logistic regressionanalyses revealed that ACT was associated with a higher likelihood ofbeing employed/occupied (p=.001), of living independently(p=.007), of being adherent with medication (p<.001), and a lowerlikelihood of persistent substance misuse (p=.027).Discussion: Compared to standard care, intensive therapeutic ACTas part of integrated care could improve 1-year outcome. Future studiesneed to address, inwhich settings these improvements canbe sustained.

doi:10.1016/j.schres.2010.02.747

Poster 253"TOO MUCH": A MULTIVARIATE PREDICTIVE MODEL OFATTITUDES TO USE OF LAI ANTIPSYCHOTICS

Tim J. LambertCERP, University of Sydney Concord, NSW, Australia

Background: Recent research has found that in many countries/regions LAIs are underused. One aspect of this pattern has

Abstracts408