Screen for Depression Before Surgery in Older Adults: GDS-15 Predicts Postoperative Delirium (OR = 2.50)
- Preoperative depression significantly predicts postoperative delirium (POD): OR = 2.50 (95% CI: 1.90–3.28) across 42 studies and 9,701 patients — a 2.5x increased risk
- Even subclinical depressive symptoms (GDS-15 scores below the standard depression cutoff) predicted elevated delirium risk — the threshold for concern is lower than current screening standards suggest
- GDS-15 (Geriatric Depression Scale, 15-item) emerged as the most consistent predictor tool across studies, outperforming other depression measures for POD risk stratification
- Surgical cohort type and depression assessment tool significantly moderated the association — the risk is not uniform across all surgical populations
Postoperative delirium in older adults is common (10–50% depending on surgery type), dangerous (associated with prolonged hospitalization, cognitive decline, and increased mortality), and partially preventable — if risk factors are identified preoperatively. This meta-analysis of 42 studies establishes preoperative depression as one of the strongest modifiable risk factors, and it does something guidelines have not yet done: it identifies the specific screening tool that predicts it best.
The GDS-15 signal
Multiple depression instruments were used across the 42 studies. The GDS-15 consistently outperformed others for POD prediction. This is not because the GDS-15 is a better measure of depression — it is because it captures the specific depressive features in older adults (somatic complaints, cognitive slowing, apathy) that overlap with delirium vulnerability. The 15-item format is brief enough for preoperative assessment, and the finding that even subsclinical scores predict risk means the current binary approach (depressed/not depressed) is too coarse.
Below the cutoff matters
The most clinically important finding: GDS-15 scores below the conventional cutoff for depression were still associated with increased delirium risk. This means a preoperative patient who scores 3 on the GDS-15 (below the 5-point depression threshold) is already at elevated risk compared to a patient scoring 0. The implication is a dimensional, not categorical, approach to preoperative depression screening.
For your practice
If you work with older adults facing surgery — or consult on surgical services: add the GDS-15 to preoperative assessment. It takes 5–7 minutes. Use the score dimensionally, not just as a binary screen. Patients above the depression cutoff need psychiatric optimization before elective surgery. Patients below the cutoff but above zero need enhanced monitoring during the postoperative period. For liaison psychiatry and geriatric consultation services: this meta-analysis strengthens the case for routine preoperative mental health screening — a practice that remains optional in most surgical pathways.
A 5-minute depression screen before surgery could prevent postoperative delirium. The GDS-15 is the tool, and even subclinical scores matter.
Heterogeneous surgical populations (cardiac, orthopedic, general). Observational designs — cannot confirm causality. MMSE (cognitive function) did not significantly influence outcomes, which may reflect its insensitivity to depression-delirium pathways. The specific mechanisms linking depression to delirium (vascular, inflammatory, neurochemical) remain speculative.