Background

Depression is an important consequence of stroke that influences recovery yet often is not detected, or is inadequately treated. This is an update and expansion of a Cochrane Review first published in 2004 and previously updated in 2008.

Objectives

The primary objective is to test the hypothesis that pharmacological, psychological therapy, non‐invasive brain stimulation, or combinations of these interventions reduce the incidence of diagnosable depression after stroke. Secondary objectives are to test the hypothesis that pharmacological, psychological therapy, non‐invasive brain stimulation or combinations of these interventions reduce levels of depressive symptoms and dependency, and improve physical functioning after stroke. We also aim to determine the safety of, and adherence to, the interventions.

Search methods

We searched the Specialised Register of Cochrane Stroke and the Cochrane Depression Anxiety and Neurosis (last searched August 2018). In addition, we searched the following databases; Cochrane Central Register of Controlled Trials, CENTRAL (the Cochrane Library, 2018, Issue 8), MEDLINE (1966 to August 2018), Embase (1980 to August 2018), PsycINFO (1967 to August 2018), CINAHL (1982 to August 2018) and three Web of Science indexes (2002 to August 2018). We also searched reference lists, clinical trial registers (World Health Organization International Clinical Trials Registry Platform (WHO ICTRP); to August 2018 and ClinicalTrials.gov; to August 2018), conference proceedings; we also contacted study authors.

Selection criteria

Randomised controlled trials (RCTs) comparing: 1) pharmacological interventions with placebo; 2) one of various forms of psychological therapy with usual care and/or attention control; 3) one of various forms of non‐invasive brain stimulation with sham stimulation or usual care; 4) a pharmacological intervention and one of various forms of psychological therapy with a pharmacological intervention and usual care and/or attention control; 5) non‐invasive brain stimulation and pharmacological intervention with a pharmacological intervention and sham stimulation or usual care; 6) pharmacological intervention and one of various forms of psychological therapy with placebo and psychological therapy; 7) pharmacological intervention and non‐invasive brain stimulation with placebo plus non‐invasive brain stimulation; 8) non‐invasive brain stimulation and one of various forms of psychological therapy versus non‐invasive brain stimulation plus usual care and/or attention control; and 9) non‐invasive brain stimulation and one of various forms of psychological therapy versus sham brain stimulation or usual care plus psychological therapy, with the intention of preventing depression after stroke.

Data collection and analysis

Review authors independently selected studies, assessed risk of bias, and extracted data from all included studies. We calculated mean difference (MD) or standardised mean difference (SMD) for continuous data and risk ratio (RR) for dichotomous data with 95% confidence intervals (CIs). We assessed heterogeneity using the I² statistic and assessed the certainty of evidence using GRADE.

Main results

We included 19 RCTs (21 interventions), with 1771 participants in the review. Data were available for 12 pharmacological trials (14 interventions) and seven psychological trials. There were no trials of non‐invasive brain stimulation compared with sham stimulation or usual care, a combination of pharmacological intervention and one of various forms of psychological therapy with placebo and psychological therapy, or a combination of non‐invasive brain stimulation and a pharmacological intervention with a pharmacological intervention and sham stimulation or usual care to prevent depression after stroke. Treatment effects were observed on the primary outcome of meeting the study criteria for depression at the end of treatment: there is very low‐certainty evidence from eight trials (nine interventions) that pharmacological interventions decrease the number of people meeting the study criteria for depression (RR 0.50, 95% CI 0.37 to 0.68; 734 participants) compared to placebo. There is very low‐certainty evidence from two trials that psychological interventions reduce the proportion of people meeting the study criteria for depression (RR 0.68, 95% CI 0.49 to 0.94, 607 participants) compared to usual care and/or attention control.

Eight trials (nine interventions) found no difference in death and other adverse events between pharmacological intervention and placebo groups (RR 1.25, 95% CI 0.32 to 4.91; 496 participants) based on very low‐certainty evidence. Five trials found no difference in psychological intervention and usual care and/or attention control groups for death and other adverse events (RR 1.18, 95% CI 0.73 to 1.91; 975 participants) based on very low‐certainty evidence.

Authors’ conclusions

The available evidence suggests that pharmacological interventions and psychological therapy may prevent depression and improve mood after stroke. However, there is very low certainty in these conclusions because of the very low‐certainty evidence. More trials are required before reliable recommendations can be made about the routine use of such treatments after stroke.