This thesis aims to improve the early identification of mental healthproblems (MHPs) in children by developing a prediction model for MHPs inchildren based on readily available information from... Show moreThis thesis aims to improve the early identification of mental healthproblems (MHPs) in children by developing a prediction model for MHPs inchildren based on readily available information from electronic patient recordsfrom general practice.The prediction models for child MHPs, based on the data from the electronichealth records of general practitioners (GPs), have not yet performed wellenough to be used safely in daily practice. A number of relevant predictivecharacteristics have been identified: characteristics such as physicalcomplaints (e.g. abdominal pain or headache) and characteristics related tohigher health care use (e.g. more than two GP visits or a laboratoryexamination in the previous year) were age-independent predictors of MHPs.Awareness of (a combination of) these characteristics can help GPs to identifyMHPs at an early stage.To investigate whether merging information from preventive youth healthcare(PYH) and GPs in one algorithm can improve the identification of MHPs, wecombined information from the electronic files of PYH and GPs. However, themodels based on these combined data did not outperform the models based on GPdata alone. Several individual characteristics measured in PYH turned out to bepredictors for MHPs in general practice. Show less
