Indicators of Prognosis Independent of Treatment for Adults with Depression in Primary Care, Going Beyond Baseline Symptom-Severity: A Systematic Review and Individual Patient Data Meta-Analysis

Abstract

Background: There is evidence that depressive symptom severity is associated with prognosis but existing studies have usually been for people on a single treatment so we lack evidence for this effect independent of treatment. There are other factors that we will call ‘disorder severity’ such as duration and comorbidity that have been reported as associated with prognosis, but current evidence does not indicate if these associations are independent of treatment and symptom severity.

Methods: Medline, Embase, PsycINFO and Cochrane Central were searched from inception to 20^th March 2019 for randomised controlled trials (RCTs) of adults seeking treatment for depression from their general practitioners, that used the Revised Clinical Interview Schedule (CIS-R) so that there was uniformity in the measurement of disorder severity factors. Individual participant data were gathered from 12 RCTs. Two-stage random-effects meta-analyses were undertaken calculating effect estimates within each study and then pooling across these to ascertain the independent association between each potential prognostic factor and depressive symptoms at 3-4 months post-baseline, remission and depressive symptoms at 6-8 months post-baseline. Risk of bias was calculated using QUIPS and quality was assessed using GRADE.

Findings: Baseline depressive symptom severity was strongly associated with prognosis independent of treatment; there was a 27∙5%(95%CI: 25∙0 to 30∙1) difference in depressive symptoms at 3-4 months per one standard deviation increase in baseline score, equivalent to about eight Beck Depression Inventory points or five PHQ-9 points. The duration of anxiety, duration of depression, comorbid panic disorder, and a history of antidepressant treatment were also associated with prognosis independent of depressive symptom severity, though the association with antidepressant treatment was less robust. For participants with all four of these factors, there was a difference of 39∙9%(95%CI: 12∙6 to 74∙0) in their score at 3-4 months. Risk of bias was low in all studies, quality was high and heterogeneity was within acceptable limits for all but two prognostic indicators. A number of sensitivity analyses did not alter our conclusions.

Interpretation: Depressive symptom severity had a strong association with prognosis independent of treatment. The duration of depressive symptoms, duration of anxiety symptoms, panic disorder and past antidepressant use were all associated with prognosis independent of depressive symptom severity and treatment. Consideration of these prognostic indicators could be clinically important for determining prognosis and inform patients and clinicians about likely outcomes in the clinical management of depression.

Funding Statement: This work was supported by the Wellcome Trust through a Clinical Research Fellowship to JB (201292/Z/16/Z), Medical Research Council (Programme for IW: MC_UU_12023/21), MQ Foundation (for ZC: MQDS16/72), the Higher Education Funding Council for England, the National Institute of Health Research (NIHR), NIHR University College London Hospitals Biomedical Research Centre (RS, KC, PB, and SP), University College London (GA, GL), University of Pennsylvania (RDR), Vanderbilt University (SDH), University of Southampton (TK), University of Exeter (EW), and University of York (SG).

Declaration of Interests: All authors declare no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; and no other relationships or activities that could appear to have influenced the submitted work.

Ethics Approval Statement: PROSPERO registration: CRD42019129512.

Keywords: Depression; Prognosis; Individual Patient Data Meta-analysis; Primary Care; Severity of Illness Index