Perceived helpfulness of treatment for major depressive disorder: findings from the WHO World Mental Health Surveys

Importance: Perceived helpfulness of treatment is an important patient-centered outcome that is a joint function of (i) whether treatment providers are perceived as helpful and (ii) whether patients persist in help-seeking after previous unhelpful treatments.
Objective: To examine the prevalence and predictors of the two main components of perceived helpfulness of treatment in a representative sample of individuals with a history of depression.
Design: The World Health Organization World Mental Health Surveys, a coordinated series of community epidemiological surveys of non-institutionalized adults.
Setting: Seventeen surveys in 16 countries (8 surveys in high-income and 9 in low/middle-income countries).
Participants: 2,726 people with a lifetime history of treated DSM-IV major depressive disorder (MDD).
Exposures: Socio-economic characteristics, lifetime comorbid conditions (anxiety and substance disorders), treatment type, treatment timing, and country income level.
Main Outcomes and Measures: (i) Conditional probabilities of helpful treatment after seeing between 1 and 5 professionals; (ii) persistence in help-seeking after between 1 and 4 unhelpful treatments; and (iii) ever obtaining helpful treatment regardless of number of professionals seen.
Results: Cumulative probability of helpful treatment after seeing up to 5 professionals was 92.9%, but only 39.4% of patients persisted this long (i.e., beyond 4 unhelpful treatments), resulting in 68.2% of patients ever receiving treatment they perceived as helpful. Probability of perceiving treatment as helpful increased with 4 predictors: older age at initiating treatment, higher education, shorter treatment delay, and receiving medication from a mental health specialist. Decomposition showed that the first 2 of these 4 variables predicted only conditional probability of an individual treatment provider being helpful, whereas the latter 2 variables predicted only persistence. A time trend was found for higher probability of an individual treatment provider being helpful in recent years, but a counteracting trend existed for persistence being lower in recent years, resulting in a non-significant association between historical time and patient-level probability of receiving helpful treatment.
Conclusions and Relevance: The probability that MDD patients obtain treatment they consider helpful might increase, perhaps dramatically so, if they persisted in help-seeking after unhelpful treatment with up to 4 prior professionals.