Many patients with depression do not respond well to medication and are diagnosed with treatment refractory depression (TRD). Sometimes medications don't work because an individual metabolizes the drugs in an atypical manner (too fast/slow). Alternatively, drugs may fail to work because the underlying sub-type of depression is not effectively targeted by a medication. This study will use genetic testing of subjects with TRD to personalize the drug treatment of depression and guide the patient to a better clinical outcome. In the guided group, the clinician will receive a pharmacogenetic report to help individually tailor medication selection for TRD patients, potentially allowing the clinician to pick more effective medications right away, and when necessary, use drug combinations that are well-tolerated and less likely to cause unwanted side effects. The control group will receive a sham genetic report and be treated according to typical standards of care. The investigators will conduct our study in a "real world" setting, with few restrictions on which TRD patients can participate. In this way, the findings may be more likely to reveal how useful genetic testing will be when applied more broadly in psychiatry.

Study Design and Methods. The investigators will use prospective methods to assess the utility of pharmacogenetic guided treatment (PGT) of TRD. The use of genetic testing for PGT subjects, and clinical care guided by the individual's genetic results are the only subject activities in this study that represent a departure from standard clinical care. Other activities carried out within the context of this study are conducted in accordance with accepted clinical practice by a trained and licensed psychiatrist. These include (but are not limited to) the use of prescription medications, laboratory testing, vital sign measurements, mood scales and diagnostic interviews, electrocardiograms, and imaging studies. Standard clinical activities will be conducted at the discretion of the attending psychiatrist, determined by the clinical needs of the individual patient. All laboratory testing conducted for the purposes of clinical treatment (diagnosis, treatment, and prevention of illness) will be conducted at the VA San Diego or VA Palo Alto accredited clinical laboratories. No investigational medications or devices will be used in this study.

Setting: Clinical care will be conducted at the Veterans Affairs San Diego Healthcare System and Veterans Affairs Palo Alto Healthcare . Patients with TRD will be recruited from multiple sites in the local VA Systems. Clinical care will be provided by a VA psychiatrist.

Genotyping: After obtaining written consent, patients will give a saliva sample from which genomic DNA is prepared using established protocols. DNA samples will be assigned a coded subject identification number that will not contain any sensitive personal information. The code key will remain at the VASDHS at all times. Samples will be sent to a CLIA certified laboratory (Pathway Genomics) for genotyping and analyzed using the commercially available Mental Health DNA Insight Test. Unused DNA sent for genotyping will be destroyed. No DNA storage outside the VASDHS will be permitted. No other clinically relevant information will be uncovered by the genetic testing.

Randomization & Study Design: Patients will be randomly assigned to one of two groups: Pharmacogenetic guided treatment (PGT) and treatment as usual (TAU). All patients will provide a saliva sample at the start of their participation 5-7 days before their initial clinic appointment: All subjects will be genotyped and analyzed using the Mental Health DNA Insight Test. The treating physician for the PGT group will receive the report before the first clinic visit. The treating physician for the TAU group will receive a sham report.

Patients will remain blind into which treatment group they are assigned until completion of the study. The study physicians will not be blinded due to the need to review the PGT report. However, bias will be reduced by using patient self-reports of their mood using the 16-item quick inventory of depressive symptoms self report (QIDS-SR16) scale. All patients will then be seen by a study psychiatrist at 4 weeks and 8 weeks follow-up. All patients will complete self assessment inventories at each visit. In addition they will assess tolerability using a standardized side effect checklist. Overall functional impairment and symptom severity will be assessed using the Clinical Global Impressions Scale (CGI). Primary study outcome will be the CGI score at the end of the study (8 weeks). A Comprehensive Suicide Risk Assessment will be done at study initiation and a clinical suicidality assessment will be conducted at every visit according to usual clinical practice.

Inclusion Criteria:

• Age >18
• Veteran
• Psychiatric diagnosis
• Clinically significant depressive symptoms
• Failed 1 or more adequate treatment trials

Exclusion Criteria:

• No clinically significant symptoms of depression
• No previous medication trials
• Pregnancy
• Inpatient medical or psychiatric hospitalization

patient blinded to condition