Nine-month predictors and outcomes of SSRI antidepressant continuation in primary care☆☆☆
Received 26 January 2005; received in revised form 31 March 2005; accepted 4 April 2005.
Abstract
Background
This study aimed to identify the predictors and outcomes of SSRI antidepressant continuation, discontinuation and switching over a 9-month period of naturalistic observation.
Methods
Primary care patients (n=573) with physician-diagnosed depression from 37 practices were randomized to an open-label trial of one of three selective serotonin reuptake inhibitors (SSRIs) managed in their primary care setting. Psychiatric characteristics and treatment course were assessed at baseline and at 1, 3, 6 and 9 months after medication initiation.
Results
Nineteen percent of patients switched SSRIs, which occurred significantly sooner than discontinuation (median: 41 vs. 100 days). Time to discontinuation was primarily explained by baseline patient skepticism about taking an antidepressant (62% increase in discontinuation risk). In contrast, time to switch was associated with greater impairment at baseline and lesser improvement in impairment during the first month on medication. Patients who discontinued were significantly less likely to be depressed 9 months after starting medication than those who either continued or switched medication, and were less symptomatic and impaired than patients who switched.
Conclusions
Baseline impairment may increase the risk for SSRI antidepressant switching. Additionally, patient skepticism about antidepressants predicts early SSRI discontinuation and may predict rapid recovery. Intent-to-treat analyses in nonrandomized clinical trials may paradoxically inflate antidepressant effect sizes.
aDepartment of Family Medicine, University of Michigan, Ann Arbor, MI 48109, USA
bDepartment of Psychiatry, University of Michigan, Ann Arbor, MI 48109, USA
cDepartment of Medicine, Division of General Internal Medicine and Geriatrics, Indiana University School of Medicine, Indianapolis, IN 46202, USA
dRegenstrief Institute, Indianapolis, IN 46202, USA
eUS Outcomes Research, U.S. Medical Division, Eli Lilly and Company, Indianapolis, IN 46285, USA
fDepartment of Psychology, Indiana University-Bloomington, Bloomington, IN 47405, USA
gDepartment of Medicine, Division of Biostatistics, Indiana University School of Medicine, Indianapolis, IN 46202, USA
Corresponding author. Department of Family Medicine, University of Michigan, Box 0708, Ann Arbor, MI 48109, USA. Tel.: +1 734 998 7120x320; fax: +1 734 998 7335.
☆ Competing interests: James E. Aikens, PhD, is an employee of the University of Michigan and has served as an independent consultant to Pfizer, Inc. Kurt Kroenke, MD, has received research funding and honoraria as a consultant from Eli Lilly and Company, Pfizer, Inc., and Wyeth Pharmaceuticals. Ralph W. Swindle, PhD, is currently an employee of and stockholder in Eli Lilly and Company. Dr. Swindle has not received funding from any other source than Eli Lilly and Company over the last 5 years. Eli Lilly and Company funded the original ARTIST study, a randomized effectiveness trial of its SSRI antidepressant fluoxetine (Prozac), with paroxetine (Paxil) and sertraline (Zoloft). However, Eli Lilly and Company did not provide any funding for this manuscript.
☆☆ Author's contributions: JEA took primary responsibility for formulating the research questions, analyzing the data and writing all sections of the manuscript. KK and RWS each participated in the design and execution of ARTIST, assisted in interpreting the measures and results, and assisted JEA in writing the manuscript. GJE provided consultation to JEA on data analysis, managed the data and assisted in writing the manuscript. All authors read and approved the final manuscript.
ABSTRACT