Predictors of Recurrence in Bipolar Disorder; STEP-BD
From MindLinc Wiki
Bottom Line: This is study is an analysis of data collected from the STEP trial, examining recurrence among pts with bipolar d/o who have achieved a response. Predictors of recurrence include high number of prior episodes, current substance use or anxiety d/o, hx of eating d/o, and residual sxs present during recovery phase. Study is limited in generalizabilty, and findings may be influenced by different f/u periods between pts.
Method: This study is derived from the STEP trial, a large multicenter prospective study funded by the NIMH. Enrollment was offered to all patients with bipolar spectrum disorders seeking outpatient treatment. -Eligibility criteria: for STEP-BD: age 15 or greater, meet DSM-IVcriteria for bip I d/o, bip II d/o, bip NOS, cyclothymia, and schizoaffective d/o (confirmed by ADE and MINI). For this study: not recovered at initial assessment, achieved recovery within 2 yrs of follow-up -Exclusion criteria: inability or unwillingness to comply with assessments, inability to provide informed consent, current inpatient status (deferred enrollment) -Sample: first 2000 pts entered into STEP-BD -Assessment: initial measures included the Affective Disorder Evaluation (ADE) and the Mini-International Neuropsychiatric Interview (MINI) – performed by separate clinicians. Clinical Monitoring Form (CMF) completed (scaled 0-2: 1 or greater = syndromal) at each follow up appointment. The MADRS and Young Mania Rating Scale. The MADRS and Young Mania Rating Scale was completed q 3 mos for 1 yr, then q 6 mos; -Treatment: standard care as decided by treating psychiatrist, guided by Standard Care Pathways (SCP) – a treatment approach designed for the STEP trial, “menus of reasonable choices” formed from treatments considered first line in prior published trt guidelines; a psychosocial intervention was also included in standard care. -Follow-up: pts were seen as often as clinically indicated -Definitions: -Recovery: 0-2 syndromal features for 8 wks -Subsyndromal mood sxs: at least 3 syndromal features, not meeting dsm-iv criteria -Recurrence: meeting dsm-iv criteria -time to event: # of days from 1st recovered visit to event -Analysis: a priori outcomes: prior # mood episodes, axis I comorbidity, presence of residual manic / depressive sxs; 28 secondary outcomes were identified, including sociodemographic data, recent clinical course, lifetime clinical course, comorbid psychiatric illness, and family hx. Association between indiv predictors and time to recurrence was analyzed using Cox regression which was followed with a stepwise model.
Validity: By nature of this sub-study, pts were not randomized to trts but all received trt guided by the SCP, although adherence is uncertain. Full information regarding the baseline characteristics of the study sample is not provided; the sample consists of predominantly caucasian participants (93.6%). Pts were all symptomatic at time of entry into the study. Two separate clinicians rated the MINI and ADE, with inter-rater reliability ranging from .83-1.0. The main follow-up measure (CMF) used in this study was performed only by the treating clinician, with no blind raters or input from collateral. Follow-up was 2 yrs from study entry, time of f/u after recovery may differ between patients. Little information is provided regarding missing pt information or pts lost to f/u, so it is unclear whether LOCF was utilized.
Results: 2000 pts enrolled, all but one included: 530 (26.5%) were recovered at study entry. Of those not recovered 1469, 858 (58.4%) achieved recovery within two years of follow-up. The following results are based on information from this group of 858. (see table 1, and table 2). 416 (48.5%) experienced a recurrence – depressive epis 298 (35%), mania 50 (6%), 41 (5%) hypomania, and 27 (3%) mixed, with a median time to recurrence 44.9 wks (95%CI: 37.6-53.1). A prior outcomes: 1: # mood episodes: increased hx of phases associated with increased risk of recurrence 2. axis I comorbidity: current but not lifetime hx of substance use d/o associated with increased risk of mania recurrence. Current anxiety d/o and lifetime hx of eating d/o associated with an increased risk of depressive recurrence. 3. presence of residual manic / depressive sxs: residual manic sxs associated with recurrence of depression and mania; residual depressive sxs associated with depressive recurrence. See table 1 for secondary outcomes.
Comments: This multicenter study is the largest prospective study to assess predictors of recurrence across a spectrum of bipolar disorders. Only 58% of pts symptomatic at enrollment reached response, of which 48.5% experienced a recurrence during this study period. This may be a low estimate given that pts were followed 2 yrs from study entry and not from time of response. This study does a nice job in identifying factors associated with relapse. Limitations include f/u concerns as expressed above, limited generalizability, and limited information available regarding baseline char of sample and site differences.
Presented by Jedidiah Perdue, MD on 5/1/06
