Family's impact on health : a critical review and annotated bibliography / Thomas L. Campbell.

  • Campbell, Thomas L. (Thomas Lothrop)
Date:
1986
Catalogue details

Licence: Attribution-NonCommercial 4.0 International (CC BY-NC 4.0)

Credit: Family's impact on health : a critical review and annotated bibliography / Thomas L. Campbell. Source: Wellcome Collection.

    19/255 (page 9)
    Methodology of Family Studies Research Design There is a complex and recursive interaction between family and health that is difficult to study. As the family affects an illness, the illness may be changing the family simultaneously. Some radical system theorists argue that attempting to determine cause and effect or the direction of causality is both impossible and meaningless within a systems framework (Dell 1980, pp. 328-329). Yet the predominant belief in medicine is that the observed changes in the family and other psychosocial variables are simply the result of illness. Proponents of the family's influence on health need to demonstrate a causal relationship in the other direction as well. Case studies are useful for generating hypotheses, which must be tested in controlled studies. The description of a case may be highly influenced by the author's theoretical background or biases. Series of cases are frequently used to demonstrate the effectiveness of a particular treatment. However, without an adequate control group, no conclusions can be made. Only a few c^se studies are included in this review. The cross-sectional or prevalence study is a fairly easy and inexpen¬ sive way to examine associations between family factors and health. Large numbers of subjects can be examined, thus increasing the power of the study. These studies are particularly good for examining whether certain types of family structure, communication, or interactions (e.g., psychoso¬ matic family) are associated with certain illnesses (e.g., diabetes, asthma, or anorexia nervosa). To determine whether family variables are specific for certain illnesses, control families should include not only normal fami¬ lies but families with other acute or chronic illnesses. For example, while it has been demonstrated that parental communication deviance occurs in normal families and families with psychopathology, it is most commonly associated with schizophrenia. A frequent problem in cross-sectional studies is the overinterpretation of their results. No causal relationships can be determined from these studies. Correlations may be due to the family affecting the illness, vice versa, or a third, confounding variable (e.g., social class, genetics) affecting both. Prospective cohort (incidence) studies, while expensive and time con¬ suming, allow one to determine whether family factors precede and potentially affect an illness. Studies that attempt to determine the role of the family in the onset or etiology of an illness need either very large sam¬ ples (e.g., Framingham Study [Haynes et al. 1983], Alameda County Study [Berkman and Syme 1979]) or samples of subjects at high risk for the dis¬ order (e.g., UCLA Family Study [Doane et al. 1981]). Prospective studies of the course or treatment of an illness are much easier to conduct and have been underutilized. Most of the social supports studies use a variant of this design, called a retrospective cohort study, in which data have been 9