Abstract
Background: Researchers have been conducting efficacy and effectiveness studies on interventions that address medication-taking behavior in patients with HIV infection. However, they also must examine the fidelity to the research protocol for the delivery of the intervention.
Objectives: The aims of this study were to determine the level of fidelity to each of the specified end points (number of intervention sessions, session duration, number of telephone calls per session, the length of time between sessions, sessions with multiple interventions, and attrition from the intervention protocol) in the delivery of the 12-week, nurse-delivered, structured telephone intervention designed to improve medication adherence and to determine the factors (race, gender, and socioeconomic status) related to participation in the intervention.
Methods: This descriptive correlational study was a substudy of a larger randomized clinical trial (N = 200) testing the effectiveness of a 12-week nurse-delivered intervention designed to improve medication adherence to antiretroviral medications in persons with HIV infection. The Intervention Tracking Form was used to collect data on all intervention delivery end points during the delivery of the structured intervention. A sociodemographic questionnaire was used to collect the sociodemographic characteristics of the sample.
Results: Two thirds of the sample (n = 66) were male, and slightly more than half were White. The average age was 39.68 (SD = 7.98) years. The average number of intervention sessions delivered was 8.1 (SD = 4.07). Participants were more likely to receive the first five intervention sessions (n = 77, 77.8%), and 21 (21.2%) dropped out of treatment before it was completed. Nearly one quarter (n = 24, 24.2%) of the sample had doubled-up interventions. Intervention sessions lasted, on average, 11.3 min. Typically, more than one telephone call was needed before the participant was reached (M = 2.2). The mean number of days between sessions was 11.5 days. Women were more likely to have doubled-up interventions (p = .036). There was a marginally significant difference (p = .075) in the number of sessions received between the Whites and non-Whites: Whites (M = 8.8, SD = 3.9) received slightly more sessions compared with non-Whites (M = 7.24, SD = 4.2). When examining the interaction effects between the demographic factors considered, a race-by-income effect was observed for the mean number of attempts to contact the participant (p = .044).
Discussion: These results demonstrate a lack of fidelity to the research intervention protocol. Factors beyond the researchers' control may have influenced fidelity to the intervention protocol.