Predictors of self-care: a test of Orem's model.

Despite advancements in self-care for people with cancer, a profile of individuals who perform self-care behaviors has not been determined. Without these data, interventions cannot be targeted to specific patient groups that demonstrate the most need. This study's investigators used Orem's model of self-care to determine predictors of self-care behavior. The sample consisted of 127 adults with cancer who were to receive their first dose of chemotherapy (e.g., doxorubicin, methotrexate, 5-fluorouracil, cyclophosphamide, or cisplatin, given alone or in combination). Data were collected from the subjects prior to the first cycle and at drug nadir for each of the four chemotherapy cycles. Potential predictors of self-care included demographic variables, performance status, affective state, social support, ability to manage a situation, self-care ability, and prior health-promoting activities. Self-care was measured using a 40-item behavior checklist. Multiple regression techniques were used to obtain four significant predictors of self-care, which accounted for 47% of the variance. Subjects with lower performance status, higher anxiety, less social support, and more education performed more self-care. These findings clarify Orem's model with this clinical population and provide a patient profile of self-care for practice.