Factors associated with nonresponse to a computer-tailored asthma management program for urban adolescents with asthma.
Joseph CL, Havstad SL, Johnson D, Saltzgaber J, Peterson EL, Resnicow K, Ownby DR, Baptist AP, Johnson CC, Strecher VJ.
Department of Biostatistics and Research Epidemiology, Henry Ford Health System, Detroit, Michigan 48202, USA. cjoseph1@hfhs.org
Abstract
BACKGROUND: The ability to identify potentially resistant participants early in the course of an intervention could inform development of strategies for behavior change and improve program effectiveness.
OBJECTIVE: The objective of this analysis was to identify factors related to nonresponse (i.e., lack of behavior change) to an asthma management intervention for urban teenagers. The intervention targeted several behaviors, including medication adherence, having a rescue inhaler nearby, and smoking.
METHODS: A discriminate analysis was conducted using data from a randomized trial of the intervention. Included in this analysis are participants who reported a physician diagnosis of asthma, completed a baseline questionnaire, were randomized to the treatment group, completed >or=2 of 4 educational sessions, and completed >or=2 of 3 follow-up questionnaires. Ninety students met criteria for inclusion in this subgroup analysis.
RESULTS: In logistic regression models for medication adherence, nonresponse was related to low baseline asthma self-regulation, odds ratio = 3.6 (95% confidence interval = 1.3-9.5). In models for having an inhaler nearby, nonresponse was related to low baseline self-regulation and to rebelliousness, OR = 4.7 (1.6-13.2) and 5.6 (1.7-18.0), respectively. Nonresponse to smoking messages was related to rebelliousness, low emotional support, and low religiosity, ORs = 7.6 (1.8-32.3), 9.5 (1.4-63.5), and 6.6 (1.5-29.8) respectively.
CONCLUSIONS: Certain variables had the ability to discriminate the likelihood of response from that of nonresponse to an asthma program for urban, African American adolescents with asthma. These variables can be used to identify resistant subgroups early in the intervention, allowing the application of specialized strategies through tailoring. These types of analyses can inform behavioral interventions.