Search

The Duke Social Support and Stress Scale (DUSOCS) was validated in 249 adult family practice patients using the Family Strengths, the Family Inventory of Life Events, and the Duke Health Profile (DUKE) as comparison instruments. Validity was supported in that the DUSOCS family support measure had the clinically expected positive associations with DUKE health measures (regression coefficients of +7.4 to +18.7) and negative associations with DUKE anxiety and depression measures (-2.0 to -17.2). DUSOCS family stress had negative associations with the health measures (-11.6 to -34.5) and positive associations with anxiety and depression (+18.9 to +32.1). Family and non-family stress contributed more than severity of illness to elevated levels of anxiety and depression and lowered levels of mental health, social health, and self-esteem, while these same stresses contributed only half as much as severity of illness to lowered physical health.

This review assesses past progress, current practices, and future needs in research and clinical practice involving the social support needs of cancer patients. A review is given of the various conceptualizations of the social support/stress paradigm and of the state of the art of measuring social support. Then the current work in the field of social support and cancer is considered and an argument is made for the use of social support measures, which are relevant to the experiences of the cancer patient. Potential adaptations of an existing instrument (the Duke-UNC Functional Social Support Scale) are demonstrated, and a taxonomy of stages of cancer that would require additional types of social support measures and interventions is outlined. Interventions are discussed in terms of the traditional support groups as well as interventions by the oncologist and primary care physician. An argument is made for the inclusion of quality of life or functional measures as outcomes in clinical trials and the care of the cancer patient. Finally, the need to address the existential, philosophic, or religious issues surrounding cancer and its treatment is discussed.

A 14-item, self-administered, multidimensional, functional social support questionnaire was designed and evaluated on 401 patients attending a family medicine clinic. Patients were selected from randomized time-frame sampling blocks during regular office hours. The population was predominantly white, female, married, and under age 45. Eleven items remained after test-retest reliability was assessed over a 1- to 4-week follow-up period. Factor analysis and item remainder analysis reduced the remaining 11 items to a brief and easy-to-complete two-scale, eight-item functional social support instrument. Construct validity, concurrent validity, and discriminant validity are demonstrated for the two scales (confidant support--five items and affective support--three items). Factor analysis and correlations with other measures of social support suggest that the three remaining items (visits, instrumental support, and praise) are distinct entities that may need further study.

Three hundred forty-three family-practice patients were surveyed by questionnaire and medical record audit to evaluate the relationships between social support and medical care utilization. Social support was not associated with laboratory test ordering. The mean number of office visits per year was higher for patients with low versus high confidant support (4.71 vs. 3.81, P less than 0.10) and affective support (5.21 vs. 3.60, P less than 0.05). Mean total charges in 1 year were higher for patients with low versus high confidant support ($232 vs. $148, P less than 0.05) and affective support ($244 vs. $154, P less than 0.05). Poor confidant and affective support were both associated with longer visits. Structural measures of social support were not related significantly to any utilization indicator. These findings were maintained in multiple-regression models controlling for physical health and seven demographic characteristics. Second-order regression models revealed interaction by race, employment status, and sex. Blacks showed no effect of confidant support on office visits. Poor confidant support resulted in $201 more in total charges for the unemployed (P = 0.003) versus $49 more for the employed (P = 0.15). Women with low affective support had $119 more in charges (P = 0.001) versus $16 less for men (P = 0.82). The results suggest that low functional social supports are important determinants of increased medical service utilization and that they may have differential effects by race, sex, and employment status, all of which should be considered independently in future studies.