Longitudinal Patterns of Mexican and Puerto Rican Children's Asthma Controller Medication Adherence and Acute Healthcare Use.

Rationale: Researchers tend to study Latinos as a single group, but recent asthma research confirmed differences among Latino subgroups. Variations in controller medication adherence may be a factor in the observed health disparities between Mexican and Puerto Rican children. Adherence is not a stable phenomenon; however, there is a paucity of data on patterns of adherence, sociodemographic predictors of patterns, and variations in asthma-related acute healthcare use by adherence pattern among Latino subgroups. Objectives: To identify patterns of inhaled corticosteroid medication adherence over 12 months among Mexican and Puerto Rican children with persistent asthma, to examine sociodemographic predictors of adherence patterns by ethnicity, and to investigate asthma-related acute healthcare use based on these patterns. Methods: We analyzed controller medication Doser data from Mexican and Puerto Rican children (n = 123; ages 5-12 yr) with persistent asthma who participated with their caregivers in a longitudinal nonintervention study (Phoenix, AZ, and Bronx, NY). Interview and medical record data were collected at enrollment and at 3, 6, 9, and 12 months after enrollment. Results: Forty-seven to fifty-three percent of children had poor adherence (<50%) over each of the follow-up periods (cross-sectional). Children with lowest adherence were Puerto Rican, from nonpoor families, or female. Longitudinal latent class analysis yielded four adherence classes: poor, moderate, decreasing adherence, and increasing adherence. Puerto Rican children had significantly higher odds of "decreasing" (odds ratio [OR], 2.86; 95% confidence interval [CI], 0.40 to 20.50) and "poor" (OR, 5.62; 95% CI, 1.44 to 21.90) adherence than Mexican children. Females had significantly greater odds of "decreasing" (OR, 4.80; 95% CI, 0.73 to 31.74) and "poor" (OR, 5.20; 95% CI, 1.77 to 15.30) adherence group membership than males. The "decreasing" adherence group was comprised of only poor children. Children in the "poor" adherence class had the highest mean number of acute visits and emergency department visits/hospitalizations across all assessment periods. Conclusions: This study demonstrated that unique ethnicity within Latino populations may be associated with different risk levels for suboptimal controller medication adherence, which may be a factor in the observed asthma health disparities between Mexican and Puerto Rican children. Increased understanding of and attention to children's controller medication adherence patterns will provide evidence needed to identify children at highest risk for acute healthcare use and offer more-intensive intervention using less-intensive approaches for those at low risk. Clinical trial registered with www.clinicaltrials.gov (NCT01099800).