ANALYZING CLAIMS DATA FOR HEALTH SERVICES USE IN CHILDREN AND ADOLESCENTS WITH ATTENTION DEFICIT HYPERACTIVITY DISORDER

Purpose: The aim of this study was to analyze pseudonymized Austrian public health insurance claims data of two years (2006-2007) in order to investigate the utilization of health services in children and adolescents with attention deficit hyperactivity disorder (ADHD).

Method: The data on drug prescription, stationary visits, extramural health services and individual information on patients like age, gender and residential district are stored in a relational database. As the data sets are linked to pseudonymized persons by their IDs, we were able to identify patients with certain interesting characteristics. We generally restricted evaluations to persons who were alive and under 20 years old on January 1, 2007 and analyzed the medical services that this subpopulation consumed in a second step.    In Austria only data on stationary visits contains ICD-10 coded information on diagnosis (for ADHD: F90 and F98.8). For identification of ADHD patients without stationary visits we used the results of a project that mapped drugs classified by their ATC code to corresponding diagnosis and found a correlation between ADHD and the ATC group N06B (“Psychostimulants, agents used for ADHD and Nootropics”), which clearly makes sense.

Result: The basic population consisted of 1 885 037 persons corresponding to persons under 20 years who are covered by Austrian public health insurance. From this population 5 707 patients filled a total of 62 850 prescriptions for a N06B drug during the two years period, with a clear concentration in the middle age group of the nine- to twelve-year-olds. The number of total prescriptions per quarter of year is nearly linearly increasing with the exception of the third quarter of each year, where there are far less prescriptions, possibly due to the summer holidays in Austria. On the other hand 1517 patients – of whom 802 had also filled prescriptions for a N06B drug – had 3901 stationary visits with ADHD diagnosis.

Conclusion: The linking of the data to persons makes it possible to filter patients for certain criteria (like the prescription of a drug indicating a disease, as in the case of ADHD) and analyze the rest of their medical claims. Future modeling studies will use this type of longitudinal data to map dynamic and seasonal relationships including treatment pathways.