28 Chapter 2 Data sources We used health records linked to administrative data. The information on mental health specialist care was obtained from the provider payment system datasets (Diagnose Behandeling Combinatie). These were available from 2011 to 2018 and were organised in terms of treatment records. For each treatment record, we observed the opening and closing information, diagnosis, and type and amount of care provided. Additionally, the Global Assessment Functioning (GAF) score at the start and end of the treatment record was available between 2011 and 2016, the period used to define our study population. Data from 2017 and 2018 were used to observe additional treatment in subsequent records. Treatment records were linked at the patient level using a pseudonymised patient identifier to three administrative data sources available for the entire study period: the Dutch Personal Records Database, tax records, and health insurance claims. The Dutch Personal Records Database contained demographic information on all residents of the Netherlands, including gender, migration background, date of birth, and place of residence. The tax records provided data on annual net equivalised household income. Finally, the health insurance claims contained data on annual health care expenditures disaggregated by types of care. Detailed information about the datasets is available in the Table A1. This research was exempt from medical ethical review according to the Dutch Medical Research Involving Human Subjects Act, because patients were not subjected to procedures or required to follow rules of behaviour. For the same reason, consent to participate was not required. The data are accessible through Statistics Netherlands [24], which reviews all output to minimise disclosure risk. Study population A total of 3,630,595 specialist mental health care records for adults living in the Netherlands were opened between Jan 1, 2011, and Dec 31, 2016. To evaluate economic gradients in the first interaction with specialist care, we restricted our population to the first record observed in this period. Additionally, we used data on annual specialist mental health care expenditures to exclude records with positive expenditure in the calendar year before the record opening, to ensure that the study population only included initial treatment. We also excluded records with zero treatment minutes, because they corresponded to episodes where no treatment was provided. Finally, we excluded records with missing treatment minutes, GAF score, or income, which mostly corresponded to records terminated after intake, diagnosis, or a crisis episode. Our study population consisted of 951,530 patients with a first specialist mental health treatment record opened between Jan 1, 2011, and Dec 31, 2016 (Table A2 and Figure A1).
RkJQdWJsaXNoZXIy MjY0ODMw