Grant: Operating Grant

While 1.2 million Canadian children and youth live with a mental illness, only 1 in 6 receives necessary health services. Without formal mental health care, children and youth are vulnerable to illness deterioration that can escalate to crises involving aggressive or life-threatening behaviours and emotional distress. For acute crises, there are limited immediate unscheduled mental health care options. On any given day, Emergency Departments (EDs) are a key service point for families hoping to access care, and are considered a 'safety net' in response to the lack of inpatient and outpatient services. To create health care improvements, a better understanding of the nature of ED visits for pediatric mental health issues is needed. Canada's public health system provides an opportunity to examine a variety of health care services through administrative data sets. Using different data sources, we can get a comprehensive picture of patient needs, if the statistical and computational techniques exist to turn these data into usable information for stakeholders. In this project, we will develop new statistical analysis methods. We will apply these new methods to Alberta administrative data sets on children and youth presenting for mental health care between April 2002 and March 2008. We will invent new methods, called multivariate marked point processes, that will allow us to combine data in a novel and more accurate way. Estimates from the methods will quantify the extent and direction of the relationships between ED visits and physician visits in non-ED settings. In addition, we will conduct extensive simulation studies (computer experiments) to investigate the properties of our new methods. With our results, we will be able to inform health services planning and ultimately help improve the care of children and youth with mental health concerns. In addition, our methods will be more widely applicable for researchers modeling different illness processes.