This study is based on anonymized health services data, study population include US residents with medical insurance between 1 Jan, 2010 and 31 Dec, 2014 as described by the Truven MarketScan Medicaid (MDCD), Truven MarketScan Medicare Supplemental (MCDR), and Truven MarketScan Commercial Claims and Encounters (CCAE) databases. Within each database, pharmacologically treated depression (PTD) cases incident during 2011 will be followed for up to 4 years to ascertain their TRD status and 1-year incidence rates for PTD and TRD. Participants will not receive any intervention as a part of this study. To assure they are incident rather than prevalent cases, study participants are required to have 1 year without a dispensing of an antidepressant medication before they can join the cohort.

TRD will be defined as the failure of two medications, as marked by their replacement or supplementation with other medications using three alternative definitions. In one the time for a medication regimen to fail is limited to less than or equal to (<=) 3 months and in the second the time for a medication regimen to fail is limited to <= 6 months, both corresponding to the medication never providing adequate relief. In the third, the time for a medication regimen to fail will be limited to <= 2 years, corresponding to the medication not providing permanent or very long lasting relief.

An episode of PTD begins when a member of the study cohort receives a dispensing of an antidepressant (AD) medication and receives a diagnosis of depression between 180 days before that dispensing and 30 days after that dispensing.

TRD incidence will be stratified by age group and sex under each of the three definitions of failure of a medication regimen.

TRD incidence will be stratified by data population (database): Medicaid, Medicare, Privately insured.

Sensitivity analyses are intended to clarify how the results of the present study will compare to the results of other studies that may define as The maximum time for failure of a regimen is changed to: 60, 120, 150, or 365 days, or is not limited; the requirement that a regimen include greater than or equal to (>=) 28 days' supply of the antidepressant medication and antipsychotic medication (if present) is changed to >= 42 days and 14 days supply; participants who get an excluded diagnosis are retroactively excluded from the study cohort; or the definition of an incident case adds a requirement for no depression diagnosis in the prior year.