Beyond stages of change: Dynamic indicators of readiness for follow-up care in long-term childhood cancer survivors

Background: Adult survivors of childhood cancer do not participate in recommended follow-up health screenings. Recent studies show that survivors' level of readiness for follow-up care, i.e. stages of change, is an indirect predictor of their screening behaviors. To inform intervention strategies, we sought to: a) classify survivors based on potentially modifiable affective, cognitive, and motivational indicators; b) identify background variables likely to predict class membership; and c) determine the extent to which readiness for follow-up care varied across the classes. Methods: The Interaction Model of Client Health Behavior guided a latent class analysis of data from 862 participants (ages 18-52 years; mean, 31; SD, 7) in the Childhood Cancer Survivor Study, North America's largest cohort of pediatric malignancy survivors. Class indicators included self-reported cancer-related worries, feelings, attitudes, and motivation. A three-category stages of change variable (pre-contemplation, contemplation, action) represented readiness for follow-up care. Results: Four unique survivor classes represented the best fitting model and were labeled: 1) concerned (18%); 2) indifferent; (38%); 3) secure (34%); and 4) resigned (10%). These classes differed significantly in gender, race/ethnicity, income, self-rated health, presence of late effects, and social influence. Using the resigned class as the reference group, survivors were less likely to be in the indifferent class if they read about the long-term effects of childhood cancer therapies (OR 0.29, p=0.003), reported late effects (OR 0.11, p<0.0001), and reported parental worries about their cancer (OR 0.32, p=0.007). They were more likely to be in the indifferent class if they reported higher income (OR 2.70, p= 0.027) and good to excellent health (OR 10.82, p <0 .0001). Survivors in the pre-contemplation stage of readiness for follow-up care had an estimated 83% probability of being in the Indifferent class. Stages of readiness were evenly distributed among the other three classes. Conclusions: To our knowledge, this is the first typology of childhood cancer survivors differentiated by affective, cognitive, and motivational determinants of readiness for follow-up care. Interventions tailored to background variables and modifiable behavioral indicators are more likely to increase survivor health screening than interventions that target pre-defined stages of readiness for follow-up care.