Sampling in population-based cancer caregivers research

Abstract

Background: Awareness that cancer impacts not only the person with the disease but also the family has increased, yet existing data provide limited information, primarily because of reliance on small geographically restricted samples. The current study used population-based sampling to develop a formula to compute the probability of survivors completing a survivor survey and nominating their family caregivers.

Methods: Eleven SEER/NPCR state cancer registries participated in the American Cancer Society Study of Cancer Survivors survey, providing information about the survivors, including their age, race/ethnicity, gender, type of cancer, and stage of cancer. A total of 19,294 cancer survivors met the inclusion criteria (>/=18 years old and diagnosed with one of the 10 common cancers).

Results: Approximately 30% of survivors identified from state cancer registries completed the survivor survey, of whom 42% nominated a caregiver. Logistic regression analysis revealed that middle-aged, female, or non-black survivors and survivors diagnosed with breast or ovarian cancer were more likely to complete the survey and nominate a caregiver, whereas survivors with bladder or lung cancer and survivors with advanced-stage cancer were less likely to complete the survey and nominate a caregiver.

Conclusions: Using the formula based on the logistic regression analysis results, a number of certain groups of survivors to be recruited from state registry can be calculated in order to have a present number of caregivers to contact for participation into a caregiver study. This is practical and valuable information, which fosters research that uses state cancer registries and increases the generalizability of findings to multiple types of cancer and different stages of cancer.