A structural equation modeling approach to the concepts of adherence and readiness in antiretroviral treatment

ObjectiveThe objective of this secondary analysis of data from a cross-sectional, nation-wide survey, was to test a hypothesized model with two latent concepts (readiness and adherence), based on the theory of trigger events. A secondary objective was to compare this model with two simpler models, without the concept of readiness.MethodsData consisted of a consecutive sample of 828 HIV patients ≥18 years on antiretroviral treatment at 30 out of 32 HIV Clinics in Sweden (response rate 97.5%). Structural equation modeling (SEM) was used to test the models against the empirical data. Chi2 test was used to compare fit between models.ResultsThe hypothesized model, with two latent concepts (readiness and adherence), fitted the data best (χ2 = 223.508, d.f. = 129, p-value < 0.0001, GFI = 0.970, CFI = 0.913, RMSEA = 0.030), and significantly better than the models with adherence as the only latent concept.ConclusionAlthough the SEM technique could not rule out that other models might also fit the data equally well, the hypothesized model, where readiness and adherence were two separate latent concepts, fitted data the best. This supports readiness as a distinct factor that influences adherence and hence treatment outcome. Increased attention should therefore be attached to interventions that focus on the individual’ readiness for behavioural change, i.e. factors amendable to change and that can be addressed by the patients themselves.Practice implicationsBased on these results it seems necessary to shift focus from adherence to readiness, especially in conditions where treatment can be postponed such as antiretroviral treatment.