15581-m-nanninga

CHAPTER 4 82 with a Mediterranean background were more likely to be referred to specialized care [8]. This may indicate that in the Netherlands attempts to refer ethnic minorities to specialized care have been more successful than in the U.S., but evidence for this is lacking. The family factor with the greatest impact was living in another than a biological two-parent family, suggesting a universal effect across countries. The effects of SES varied. The effect of SES may be different between health care systems of different countries, with for example income being more important in case of high-out-of pocket payments in health care. In Dutch studies, lower SES increased the likelihood of passing through the care system –i.e., professional problem recognition and referral–, whereas in U.S. studies higher SES did so. Our findings reflect those of Sayal [15] regarding the role of psychosocial problems, age, gender, and SES. Regarding life events, past treatment, academic problems, ethnicity, and living in a single-parent family we found these to determine the passing of more than one filter or use of care, whereas Sayal found this for the passing of only one filter. In addition, we found evidence concerning the role of urban area and place of residence. Finally, Sayal also reported professional-related determinants, but these came from studies of moderate or weak quality and were thus excluded by us. Our findings mostly reflect those of Zwaanswijk and colleagues [9] regarding the role of psychosocial problems, age, and academic problems on help-seeking. However, Zwaanswijk and colleagues also reported some determinants of help-seeking that we did not identify, that is, family burden, parental attitudes and beliefs, family educational level, family stress, parental mental health problems and parental receipt of treatment, family size, and type of abuse. In contrast, we found that living in a single-parent family increased the likelihood of help-seeking, whereas Zwaanswijk and colleagues did not identify this determinant. Differences in the definition of help-seeking might explain this: Zwaanswijk and colleagues included use of care in their definition of help seeking whereas we strictly distinguished help-seeking from use of care to provide more insight. Our more strict quality assessment of studies may be an additional explanation. Regarding problem recognition by a professional, the determinants we identified are mainly similar to those of Zwaanswijk and colleagues [9]. In addition, though, we found ethnicity, urban residence, and lower SES to be related to professional problem recognition. This might be explained by our inclusion of studies on problem recognition by a wide range of professionals instead of by only the general practitioner.