Single-Day versus Once-Daily Transcranial Magnetic Stimulation in Depression: A Propensity-Matched Comparison of Treatment Outcomes

Recently, accelerated, pharmacologically augmented regimens for transcranial magnetic stimulation (TMS) have been described, allowing treatment completion in a single day. However, comparisons of outcomes versus once-daily treatment regimens are lacking. Here, we performed a propensity-matched retrospective analysis of individual patient data from a naturalistic sample of patients with major depression who received either a 36-day course of once-daily TMS or single-day of pharmacologically augmented TMS, at a large community practice. Following exclusions, the initial set of 879 patient records yielded a preliminary sample of 114 individuals who received one-day TMS and 330 individuals who underwent 36-day TMS. Propensity-score matching yielded a final analytic sample of N=191 36-day TMS patients, matched to N=106 one-day TMS patients. Remission (PHQ9<5) was defined as the primary outcome measure, with response (≥50% PHQ9 improvement from baseline) as the secondary outcome measure; similar analyses were performed for anxiety symptoms using the GAD7. PHQ9 remission was achieved in 49.1% (52/106) of the single-day TMS patients, versus 25.1% (48/191) of the 36-day patients (χ²=17.473, p <0.0001). PHQ9 response was achieved in 71.7% (76/106) of the single-day TMS patients, versus 56.0% (107/191) of the 36-day patients (χ²=7.084, p =0.0078). Regarding anxiety symptoms, remission (GAD7<5) was achieved in 62.3% (66/106) of the single-day TMS patients, versus 33.0% (63/191) of the 36-day patients(χ²=23.79, p<0.0001). Response was achieved in 72.6% (77/106) of the single-day versus 51.9% (98/189) of the 36-day TMS patients (χ²=12.16, p =0.0005). As previously observed, single-day TMS improvements were not immediate, but instead developed in a delayed fashion over the 6 weeks following treatment, and reaching a plateau at 6-12 weeks post-treatment. This pattern was well-modeled with an exponential-decay function, which yielded superior fits versus linear models in both groups, with the single-day TMS group showing a significantly greater magnitude of effect and rapidity of onset. Observed effects were robust across four slightly different versions of the single-day protocol, varying in target set (dorsolateral ± dorsomedial ± orbitofrontal stimulation), inter-session interval (20 versus 30 min), NMDA agonist (D-cycloserine versus D-serine), and presence/absence of lisdexamfetamine augmentation. Randomized controlled studies will be required for definitive comparisons of single-day to once-daily TMS and for optimization of the regimen. However, the present study suggests that single-day TMS regimens may be capable of achieving outcomes that are not only non-inferior, but indeed superior, to typical 36-day TMS regimens, when delivered under naturalistic conditions in the community.