Neuromodulation Device Use in Patients With Major Depressive Disorder

Experimental design 1. This study will be a randomized clinical trial to assess the effects of the device on depression as well as ancillary symptoms of depression. The study will involve 2 phases: Screening/Baseline and Treatment. In the screening phase patients will be assessed by clinical psychiatrists for depression using the Diagnostics and Statistics Manuel 5 (DSM-5) criteria. Baseline rates of depression, quality of life, rumination, irritability, anxiety and future orientation will be gathered prior to treatment. These metrics will be measured by HAM-D, QQLS, RRS, BIT, GAD-7, and FOS respectively. Patients will then be randomized to either the treatment group or the control group. Control group patients will be assigned sham devices by an investigator that is not involved in measurement. On Day 1, patients will be instructed on the use of the Device and will apply the device for the first time under clinic supervision; all other uses of the device will be conducted at home. During the treatment phase, patients will undergo 2 16 minute session of device treatment per day for 8 weeks. Recommended sessions occur once during the day and once before bed. Metrics described above will be measured at the end of week 2 (day 14), week 4 (day 28), week 6 (day 42), and week 8 (day 56). Adverse effects will be assessed at each of these junctures. Patient will return the device on day 56 or at time of discontinuation. Patients who choose to discontinue the study will have their last recorder measures carried forward. In an effort to increase external validity patients will not be asked to change their current treatment plan as outlined by their outpatient psychiatrist. Patients will may continue previously prescribed medication, but no medication changes will occur during the trial.

STATISTICAL METHODS AND DETERMINATION OF SAMPLE SIZE

1. Statistical and Analytical Plans
2. Before database lock, a statistical analysis plan (SAP) will be authored as a separate document and approved by the investigative team. The SAP will provide a more detailed description of the methods for the analyses described as follows. Any deviations from the planned analyses will be described and justified in the final study report. All study data will be presented in data listings by subject number and time point (where applicable). Summary tables will be presented by time point (where applicable).
3. Analysis Populations
4. A single set of all completed participants will be analyzed for the primary and secondary outcomes. Additional analyses may be performed on a case-by-case basis using descriptive techniques to enumerate and characterize any patients found to have experienced an Adverse Event (AE) during the trial. Any adjustments to the analysis populations will be based on procedures outlined in the SAP or modifications to the SAP, and also described in the final study report.
5. Disposition of Patients
6. Eligible patients who enter and complete each phase of the clinical study will be enumerated and included for analysis. The number and percentage of patients enrolled, completed, and discontinued from the study will be summarized to identify any non-random patterns of attrition and missing data at the item level.
7. Planned Analyses
8. Data analyses will proceed by first inspecting the psychometric characteristics of the primary, secondary, and predictive measures collected on each patient within each timepoint and across timepoints. For each measure, means, medians, and standard deviation will be calculated accounting for all available cases and then different assumptions of missingness. Overall, the primary hypotheses are centered on the demonstration of efficacy of the device, relative to baseline. Disposition for all enrolled patients will be summarized, and reasons for discontinuation will be tabulated. Tabular summaries and/or listings will be provided for baseline demographic and clinical characteristics.
9. Concomitant Medications
10. Concomitant medications taken during administration of the treatment will be summarized and classified by drug class and preferred name using the World Health Organization (WHO) Drug dictionary in the most current version when the study enrollment starts. The version of the WHO Drug dictionary will be noted in the final report.
11. Analysis of Efficacy Measures
12. The primary endpoints of change from baseline in depression severity as measured by the HAM-D. To compare the change between groups on their change from pre to post treatment, a Repeated Measures Anova will be conducted to test a pre-specified contrast that corresponds to the following null and alternative hypotheses: i. Null: [(Mean-Treatment-Post - Meant-Treatment-Pre) - (Mean-Sham-Post - Meant-Sham-Pre)] = 0 ii. Alternative: [(Mean-Treatment-Post - Meant-Treatment-Pre) - (Mean-Sham-Post - Meant-Sham-Pre)] != 0 b. Secondary outcome measures of quality of life, rumination, future orientation, anxiety and irritability will be studied similarly. Details of additional measures of efficacy and their analysis will be described prospectively, prior to final database lock, in the SAP for this study. From each model, means and 95% confidence intervals (CIs) will be provided for each treatment; and difference in means, one-sided 95% CIs of the difference, and p values (significance value of 0.05) determined.
13. If the primary objective meets statistical significance at 0.05, the secondary objectives will control for type-I error by adjusting p-values using a Hockberg correction. Both the adjusted and unadjusted p-values will be presented.
14. Analysis of Exposure and Adverse Events
15. Exposure (number of treatment sessions) will be summarized by treatment. Adverse event analyses will be performed using the ITT population. AEs will be coded by primary system organ class (SOC) and preferred term according to the Medical Dictionary for Regulatory Activities (MedDRA). Treatment-emergent AEs (TEAEs), overall and for those considered to be device-related, will be summarized by number and percent of patients in each primary SOC and preferred term by treatment. Summaries will also be presented for relationship to the study device, intensity, seriousness, AEs or Severe Adverse Events (SAEs) leading to discontinuation, deaths and hospitalizations, as well as device-specific AESIs. By-subject listings will be provided.
16. Determination of Sample Size
17. To determine sample size, preliminary data was used from Fibromyalgia trial titled, "A Single Arm Study Evaluating the Efficacy of the device in Patients with Severe Pain Due to Fibromyalgia". Results from the PHQ-9 depression scale administered in this trial will be used to estimate the effect size for the HAM-D depression scale in order to determine the appropriate sample size.
18. The Fibromyalgia trial had 2 phases, a screening phase that lasted 7 days in which no intervention was used and a treatment phase that lasted 15 days in which the participants used the Device at least twice daily (up to 4 times).
19. The effect sizes for both phases were calculated as a Cohen's Dz. The treatment phase had an effect size 0.79 (95% CI 0.25 - 1.33) and the screening phase had an effect size of 0.47 (95% CI 0.18 - 0.77). The effect size over the entire study was 1.26 (95% CI 0.61 - 1.9). In this trial, the screening phase can be used as an estimate for the sham/placebo effect that may be observed in future trials. In addition the observed effect size of 0.79 for the treatment period takes place after placebo improvements have occurred during the screening phase and are thus somewhat protected against containing a partial placebo effect. The assumption for this sample size calculation is that the treatment phase effect size in the current study may show a 1.26 total effects size and the sham arm may show a 0.47 effect size, and thus the correct effect size to determine the sample size of the current trial is the remaining true treatment effect above sham which is (1.26 - 0.47=0.79) 0.79.
20. For a parallel arm trial, in order to show statistical significance with an effect size of 0.79 at an alpha of 0.05, 2-Tailed, at a power of 80% on a between subjects T-Test the sample size needed would be 27 participants per arm. To account for potential dropout of participants, the final sample size will be adjusted upward to 40 participants per arm.
21. A simulation showing the similarities in results between a T-test and a Repeated Measures Anova is attached. Thus, the above sample size for a T-test was used for the Repeated Measures Anova (see attached file #1).
22. Blinding of study a. will provide randomized devices and give study staff a list of device ID numbers to assign in order. Study investigators will have no knowledge of which device is a sham device. will keep blinding information on a separate locked external hard drive that investigators will not have access to. Data attached to each device ID will sent to 's analytics team for statistical analysis.