Predicting Lay Provider Competency in Addressing Suicidal Behavior in Nepal: ‘Let’s Talk About It’

Abstract

Suicide is a critical global public health issue in low- and middle-income countries (LMIC) in which the majority of the world’s suicides occur. Low provider competency in managing suicide risk, particularly, among non-specialist or “lay” providers with minimal health or mental health training, is a theoretically critical yet poorly studied aspect of global suicide prevention that has received limited attention in the global suicide arena. The current series of four studies examined barriers (provider stigma, competency, validation, and invalidation) to competent delivery of suicide prevention services by non-specialists in Nepal (N = 205 providers) as well as their impact on patient outcomes (depressive symptoms and suicidal ideation) (N = 96 patients receiving treatment over six months). In Chapter II, pre-training competency in delivering common therapeutic factors was significantly associated with suicide-specific clinical competency (β = .48, p < .001). A small percentage of lay providers (14.2%) assessed for suicidality, and one developed a safety plan. In Chapter III, 48.7% of providers assessed for suicide and 4.9% conducted safety planning 4-months post-training. These percentages increased to 57.5% and 11.9% 16-months post-training. Pre-training common factors competency significantly predicted suicide clinical competency 4-months post-training ((β = .24, p < .05). Contrary to hypotheses, pre-training provider implicit bias (β = .30, p < .001) and mental health knowledge (β = .23, p < .05) significantly predicted suicide clinical competency 16-months post-training. In Chapter IV, lay provider validation (β = 0.63, p < .001), but not invalidation (β = 0.22, p > .05) was associated with suicide clinical competency.In Chapter V, the relationship between lay provider suicide clinical competency and patient outcomes was mixed. An increase in post-training suicide clinical competency did not predict change in depressive symptoms (β = 3.14, robust Z = 1.86). However, there were significant products between pre- and post-training suicide clinical competency (β = -5.92, robust Z = -3.51), such that for providers with low pre-training suicide clinical competency, an increase in post-training suicide clinical competency predicted a worsening of depressive symptoms, relative to a reduction in depressive symptoms for providers with high pre-training suicide clinical competency. There was also a significant product between post-training common factors competency and suicide clinical competency (β = 3.49, robust Z = 3.79), such that for providers with higher suicide clinical competency, an increase in provider common factors competency predicted a reduction in depressive symptoms. Additionally, provider post-training explicit bias predicted a greater odds of improving SI (OR = 8.15; robust Z = 4.98), and there was a significant product between post-training provider implicit bias and suicide clinical competency (OR = 2.77, robust Z = 2.97). At high provider suicide clinical competency (vs. low provider suicide clinical competency), the odds of patient suicidality improving increased an additional 17.5%. Future studies should replicate the original investigation and conduct a deeper examination of modifiable processes that explain the link between common factors competency, validation, suicide clinical competency, and patient outcomes over time. This knowledge can be used to improve training of health workers, systematic detection of suicidal behavior within healthcare settings, and suicidal patient outcomes.