Key Takeaways
Extended Takeaways
- Baseline BMI materially changed the ranking of aripiprazole: A-ARI was preferred over S-BUP only in patients with nonelevated weight, while S-BUP was preferred over A-ARI in patients who were overweight at baseline.
- In adults under 65 years, A-ARI provided 27.3 additional gross DFDs of depression efficacy versus S-BUP, but TD alone offset 56%–81% of that benefit and, in overweight patients, weight-related harms were 95%–131% as large as the depression benefit.
- For the oldest nonelevated-weight subgroup, treatment choice was less stable than in younger adults: among adults aged 85–89 years of nonelevated weight, A-ARI, C-BUP, and S-BUP were favored in 58.8%, 36.0%, and 5.2% of probabilistic sensitivity analysis runs, respectively.
- The modeled penalty for falls with C-BUP increased with age but remained smaller than its efficacy advantage in the base case; the expected harm was described as about half a week of depression-equivalent QALY loss in adults 65–84 and a week and a half in adults 85–89.
- The weight signal for long-term aripiprazole was supported by external validation data: clinically significant weight gain in the VAST-D continuation phase was 30.6%, similar to the 36.6% reported in an earlier study, and estimated weekly gain was 0.05 (0.04–0.06) kg per week versus 0.12 (0.03–0.21) kg/week in meta-analysis.
- The model's treatment ranking was robust to several assumptions clinicians often question: changing the discount rate to 0% or 3%, shortening the horizon from lifetime to 20 years, extending fall-related disability to 2 years, or modeling weight gain as fully reversible after 3 years did not change the preferred treatment for any subgroup.
