The association between the use of parents' and childrens' medication.

Background: Family background influences medical consumption. For example, 22% of the variance in frequencies of contact with the GP in The Netherlands can be ascribed to family influence. Up to now not much is known about family influence in one particular area of medical consumption: the use of (prescribed) medication. The difference between visiting the GP and the use of prescribed medication is that the role of the GP is more important in prescription: the ultimate decision to prescribe is taken by the GP, while the ultimate decision to visit the GP is usually taken by the parent(s). The aim of this study is to examine the association between the use of prescribed medication of parents and their children. Methods: Data were used from over 10 000 two-parent families with children under the age of 12 from the Second Dutch National Survey of General Practice. For a one-year period all contacts with the GP and all prescribed medication were registered in 104 general practices in 2001. Moreover, all patients filled out a census form including 10 questions on for example, educational level, health status, and type of insurance. These data were linked with the GP registration data. Data were analysed using multilevel analyses to take into account clustering of the data. The outcomes variable was the child’s number of prescriptions. Explanatory variables include characteristics of the children (e.g. age, gender, and health status) and parental characteristics (number of prescriptions and health status). Results: Differences in use beween different groups of children (f.e. girls versus boys) could mainly be explained by differences in consultation rates between these groups. The use of prescribed medication of parents and children was associated. Within the group of families where both parents did not use prescribed medication 31% of the children received prescriptions as against 61% of the children in families where both parents used prescribed medication. The correlation between the number of prescriptions of children and parents was stronger for mothers (r = 0.21) compared to fathers (0.15). Although the correlation was not strong it remained significant in the multilevel analyses where we controlled for child’s characteristics such as health status and contact frequency as well as for parental characteristics such as health status. This analysis was only performed for those children being at risk for prescription that is those children who visited their GP at least once during the year of registration. Conclusions: The use of medication by children is associated with their parents’ use. Health status, frequency of GP consultation and other child- and parent characteristics cannot fully explain this association. This finding is interesting since the GP is the one who ultimately decides whether or not to prescribe. Apparently GPs’ prescription behaviour is influenced by the family context of individual patients. (aut.ref.)