To study whether an individualized intervention focused on problem-solving could improve adherence for vulnerable patients, we conducted a randomized controlled trial in adults from low-income inner city neighborhoods with moderate or severe asthma. Even with a relatively large sample size, a long observation period, and the use of electronic monitoring rather than self-report, we could not detect any differences in outcomes between the group that received PS and the group that received standard AE. Adherence declined over the 6-month monitoring period, but both groups demonstrated improved quality of life, FEV1, and asthma control. These observed improvements could have been due to monitoring, attention, provision of medications, or regression to the mean. Additionally, there were no statistically detectible differences or changes over time with respect to ED visits or hospitalizations, either overall or asthma-related. To isolate the factors most responsible for the favorable outcomes, future work should explore the contribution to outcomes from the common elements in both treatment groups: attention, electronic monitoring, and the provision of medications, keeping in mind that adherence declined.
Despite an intervention aimed at adherence, there was no difference between PS and AE in asthma outcomes, nor was there a difference in the decline of adherence. It is noteworthy that adherence was already relatively good at baseline and remained so throughout the observation period at a mean of 61%.7
This is especially remarkable considering a recent study by Williams et al56
which measured ICS adherence of less than 30% using prescription fill data for asthma patients in southeastern Michigan. Actual total adherence was likely lower than their estimates because not all filled medications are taken.56
In our study we monitored both groups from the beginning, informing participants that monitor use was recorded which could have resulted in higher adherence than would have occurred had medicine-taking behavior not been monitored (Hawthorne Effect). Additionally, the relatively high adherence found in both groups may help explain the improvements in quality of life and asthma control observed for both PS and AE.6
However, hospitalizations and ED visits did not decline. It is possible that within the context of our particular sample, while attention, monitoring and providing medications can improve asthma status to some degree, they are not sufficient to mitigate the use of other medical resources, such as hospitalizations, over the 6 month period. This may be a result of the vulnerability of our sample of adults who, in addition to their unusually severe asthma (mean baseline FEV1= 66%), are high baseline users of hospitalizations and ED visits, have many comorbidities, have significant tobacco exposure, and have limited financial and social resources.
It is noteworthy that our interventions, PS and AE, are focused solely on the patient. Medical services and the patient’s larger context were not addressed. Patient concerns, detailed in their choice of the Other Problem (Table E-1
), indicate that additional significant health and social problems are common. These Other Problems suggest that it is important for practices to take into account of these “external” factors which may be extremely difficult.
Some support for this argument comes from . Patients readily admitted that there were times when they did not take medications. Self-efficacy was relatively good and did not change over the course of the study suggesting that patients were not motivated to achieve better adherence. Thus, this intervention is too far “downstream”; instead motivation needs to be addressed and addressed better considering patients other priorities and problems.
The limitations of our study are informative. As noted, electronic monitoring of adherence cannot be achieved divorced from the Hawthorne Effect. Because of monitoring, many members of the AE group (66%) thought researchers were teaching about adherence. Thus, the interventions may have been perceived similarly by participants. Although the intervention was complicated and labor intensive, it focused only on patient behavior and did not consider the environment of the practice site or that of the patient’s larger social context. The study design had important elements of pragmatic research, although its complexity limits as a “real world” pragmatic intervention.57
The pragmatic elements that can inform subsequent studies are: inclusion of patients with comorbidities and those with significant tobacco exposure; utilization of a flexible individualized intervention; and delivery of PS and AE by research staff who were not trained medical personnel. The latter factor demonstrates that interventions such as attention and provision of medications could be delivered by a variety of health care workers.
In summary, our study indicates that problem-solving does not improve adherence or decrease asthma morbidity in this population. Provision of medications, monitoring, and attention are associated with improvement in some asthma outcomes but research is needed to identify interventions to alleviate the need for ED visits and hospitalizations for asthma and to improve the health of high morbidity patients living in low-income inner city neighborhoods. Such interventions likely should be incorporated into practice procedures58
and include consideration of the larger social context.
Clinical Implications/Key Summary
- Problem-solving was not better than asthma education in improving adherence or asthma outcomes in low-income inner-city adults with moderate or severe asthma.
- Monitoring inhaled steroid use with provision of medication was associated with improvement in FEV1, asthma-related quality of life, and asthma control but did not reduce asthma-related hospitalizations or ED visits for these patients.