to the outcome in CCBT, and both credibility and expectancy tend towards significance in the TAU group. Credibility and expectancy do not contribute to the outcomes of the combined treatment. CONCLUSIONS: Patients’ initial belief about the success of their depression treatment can influence the outcome. Taking the patient’s pre-treatment expectancy and credibility into account may contribute to a more effective treatment. MENTAL HEALTH—Health Care Use & Policy Studies PMH51 DECLINE IN DEPRESSIONTREATMENT PERSISTS AFTER FDA ANTIDEPRESSANT WARNINGS Libby AM, Orton HD,Valuck RJ University of Colorado Denver, Aurora, CO, USA OBJECTIVES: To measure the persistence over time of intended and unintended effects on community based depres- sion treatment in the U.S. associated with the FDA warnings on antidepressants and suicidality. In October 2003 the U.S. FDA issued a Public Health Advisory about the risk of suicidality for pediatric patients on SSRI antidepressants; a boxed warning and medication guide were implemented in February 2005, and later extended to young adults aged 19–24. Unintended declines in diagnosis and non-SSRI substitute treatment have been shown immediately following the advisory both for pedi- atric patients, and for adult patients not targeted by the warn- ings. Whether those changes persisted is unknown. METHODS: Pediatric, young adult, and adult cohorts with newly diagnosed episodes of depression were created using a national, integrated managed care claims (commercially avail- able from PHARMetrics®, a Unit of IMS, Inc.) from July 1999-June 2006 (n = 55,218 youth; 44,141 young adults; 394,524 adults patients with new episodes). Time series analy- ses compared post-FDA advisory trends to expected trends based on pre-advisory patterns. RESULTS: Young adult and adult populations mirrored changes in pediatric depression care after the FDA advisory. Reductions in national rates of depres- sion treatment were substantial, returning national diagnosing rates to 1999 levels for pediatric patients and to 2003 levels for adults. Primary care providers continued significant reductions in new diagnoses of depression (50% lower for pediatric, 40% lower for young adult, 30% lower for adult). Substitute care by psychiatrists or psychologists, psychotherapy, and anxiolytic and atypical antipsychotic medication use did not significantly change from pre-advisory trends. CONCLUSIONS: Declines in depression diagnosis continued after the advisory. Substitute care did not compensate. Spillover to adults continued. Results suggest that the unintended effects were substantial, diffuse, and non-transitory in a large national population. Policy actions are required to counter the unintended consequences of reduced depression treatment. PMH52 PILOT SURVEY OFADVERSE EVENT MANAGEMENT ASSOCIATED WITH ANTIPSYCHOTIC USE: THE NEGLECTED DIMENSION IN RESOURCE USE IMPLICATIONS (A UK PERSPECTIVE) Meier G, Malcolm B AstraZeneca, Luton, Bedfordshire, UK OBJECTIVES: A pilot survey was conducted to determine which NHS services and resources may possibly be impacted in the management of adverse events (ADEs) related to antipsychotic use. METHODS: Ten UK mental health specialists were inter- viewed. The six-question survey advised participants to assume that patients’ psychotic symptoms were controlled or that it was too early in therapy to determine efficacy. The ADEs considered were limited to those identified by NICE as most troublesome, i.e. sedation, extrapyramidal symptoms (EPS), weight gain and sexual dysfunction. RESULTS: The most common routine tests performed were: full blood count, glucose, cholesterol and weight. Action was prompted for sedation and sexual dysfunc- tion only if the patient raised a concern. For weight gain and EPS, clinician and patient concerns were given equal consideration. Actions taken in order of preference were: sedation—decrease dose, change timing, switch treatment; EPS—decrease dose, add anticholinergic, switch treatment; weight gain—switch treat- ment, decrease dose; sexual dysfunction—decrease dose, switch treatment, add sildenafil. Referrals were mentioned in all responses but were most commonly associated with weight gain and sexual dysfunction. Most common tests requested as a result of an ADE assessment were: glucose, weight, cholesterol and prolactin. Switching treatment was the most common action in response to positive tests results for any ADE. CONCLUSIONS: This pilot survey highlights that ADEs could play a substantial role in treatment costs given the potential use of resources linked to their observation, diagnosis and management. As such, in order to comprehensively calculate the cost-effectiveness of any antipsychotic treatment, acquisition costs as well as costs asso- ciated with potential adverse event management should be con- sidered. To confirm the findings of this survey, further in-depth research is warranted. PMH53 PREDICTIVE FACTORS OF RECURRENCE AND BIPOLAR DISORDER MANAGEMENT IN SPAIN: A PROSPECTIVE COHORT STUDY BASELINE ASSESSMENT Tafalla M 1 ,Vieta E 2 , de Dios C 3 , Goikolea J 2 , Saiz-Ruiz J 4 , González-Pinto A 5 , Montes JM 6 , Diez T 1 1 AstraZeneca, Madrid, Spain, 2 Hospital Clínic de Barcelona, Barcelona, Spain, 3 Centro de Salud Mental Fuencarral-El Pardo, Madrid, Spain, 4 Hospital Ramón y Cajal, Madrid, Spain, 5 Hospital de Santiago Apóstol, Vitoria, Spain, 6 Hospital del Sureste, Arganda del Rey, Spain OBJECTIVES: To fully describe patients’ characteristics, man- agement patterns, predictive recurrence factors and economic impact of health care attention of bipolar patients population in Spain. The description of the sample of an ongoing cohort study is being presented. METHODS: Prospective observational cohort study with a follow-up of 12 months including consecu- tive outpatients diagnosed of Bipolar Disorder I or II (DSM-IV- TR), stabilized for at least 2 months, who had at least a mood disorder episode (depression, mania, hypomania or mixed) in the last year. RESULTS: A total of 571 patients were included, 60.1% women. Mean age was 47.4 years (SD 13.1). Only 37.5% were active workers, up to 5.4% were having temporal disability leave and 16.2% were permanent disabled for working. Regard- ing study disease, 75% were bipolar I and mean time since diagnosis was 12.3 years (SD 10.5), although time since first mood episode compatible with bipolar diagnose was 16.3 years (SD 11.2). Mean mood episode number since disease onset was 10.6 (SD 9.3). Up to 74.1% of patients had been hospitalised during the disease evolution, mean times were 3.6 (SD 3.3). 21.3% of patients have had a suicide attempt. Regarding the baseline evaluation, 29.9% of patients were not free of mood symptoms even being stabilized for at least two months, present- ing more than two mania symptoms (10.6%) and more than 2 depression symptoms (18.2%). CONCLUSIONS: Even in the stabilization phase of the disease there is an important symptoms load. Although depressive symptoms are more frequent than manic symptoms, both subsyndromal symptoms are present in euthimic bipolar patients. Bipolar disorder is a relevant chronic A596 Abstracts