Screening for depression, anxiety, and suicide risk in adults: A systematic evidence review for the U.S Preventative Services Task Force

Objective:
To review the benefits and harms of screening and treatment for depression, anxiety, and suicide risk, and the accuracy of instruments to detect these conditions among primary care patients.

Data Sources:
MEDLINE, PsychINFO, the Cochrane Central Register of Controlled Trials (CENTRAL), the Cochrane Database of Systematic Reviews through September 9, 2022, bridging from prior USPSTF reviews or other relevant reviews. Eligible studies included in the prior reviews were also included. We conducted ongoing surveillance for relevant literature through November 25, 2022.

Study Selection:
We reviewed 23,497 abstracts and assessed 1237 full-text articles against a priori inclusion criteria. We included English language studies of screening or treatment (compared to control conditions), or test accuracy of a priori selected screening instruments. Primary studies of screening and test accuracy were limited to primary care populations, as were primary studies of anxiety treatment. Primary studies of suicide prevention treatment that recruited from non-acute outpatient settings were included. Included study design varied by condition and key question; primary trials and test accuracy studies were used for smaller evidence bases, and existing systematic reviews (ESR) were used for large, mature bodies of literature. Observational studies and ESRs of observational studies were included for harms of pharmacotherapy. Critical appraisal was completed independently by two investigators for primary research. ESRs were appraised by a single reviewer and confirmed by a second reviewer if minimum quality standards were not met. Data were extracted from studies by one reviewer and checked by a second.

Data Analysis:
Where primary research evidence was sufficient for pooling, we conducted random effects meta-analysis using the DerSimonian & Laird or restricted maximum likelihood method with the Knapp-Hartung correction for a small number of studies. Where possible, subgroup analysis and meta-regression were used to explore effect modification. Pooled results from ESRs were presented in tables and forest plots.

Results:
185 studies (86 ESRs and 99 primary studies) were included, covering an estimated 13 million persons, across all conditions and key questions. Depression screening interventions, many of which included additional intervention components, were associated with a lower prevalence of depression or clinically important depressive symptomatology after six to twelve months (OR, 0.60 [95% CI, 0.50 to 0.73]; 8 RCTs [n=10,244]; I2=0%). Several instruments demonstrated adequate test accuracy (e.g., sensitivity 0.82 [95% CI, 0.76 to 0.86], specificity 0.87 [95 % CI, 0.84 to 0.89] for the patient health questionnaire (PHQ)-2 followed by the full PHQ-9 if the PHQ-2 is positive), and a large body of evidence supported benefits of psychological and pharmacologic treatment of depression. A pooled estimate from trials used for FDA approval data suggested a very small increase in the absolute risk of a suicide attempt with second generation antidepressants (OR, 1.53 [1.09 to 2.15]; N= 40,857; 0.7% of antidepressants users had a suicide attempt vs 0.3% of placebo users; median followup, 8 weeks). Two screening studies found no benefit for screening for anxiety. Among test accuracy studies, only the GAD-2 and GAD-7 were reported by more than one study and demonstrated adequate accuracy for detecting generalized anxiety disorder (e.g., sensitivity 0.84 [95% CI, 0.74 to 0.94], specificity 0.87 [95 % CI, 0.80 to 0.93] for the GAD-7 at a cutoff of 9). Evidence was limited for other instruments and other anxiety disorders. A large body of both primary and ESR evidence supports the benefit of treatment for anxiety. One RCT (n=443) of a suicide risk screening intervention found no reduction in suicidal ideation after two weeks; three studies of suicide risk test accuracy were included with no replication of any instrument; and suicide prevention studies did not demonstrate an improvement over usual care, and one large (n=18,883) trial found an increased risk of suicide attempts associated with a low-intensity online intervention in addition to usual mental health care, compared with usual mental health care alone.

Limitations:
Suicide prevention treatment studies typically used usual or optimized specialty mental health care as control groups, so could be considered comparative effectiveness. Limiting the examination of anxiety screening instruments to prespecified a priori instruments may have excluded some relevant studies. The use of ESRs may have limited our ability to examine effects in some specific patient populations.

Conclusions:
Both direct and indirect evidence support depression screening in primary care settings, including during pregnancy and postpartum. While evidence is insufficient to draw conclusions about the benefits or harms of anxiety screening interventions, there is clear evidence that treatment for anxiety is beneficial, and more limited evidence indicating acceptable accuracy of some anxiety screening instruments to detect generalized anxiety disorder. There are numerous important gaps in the evidence for suicide risk screening in primary care settings.