There are several limitations to this study. Since we did not include a control group, some of the observed changes over the study period may have been due to natural improvement in asthma management skills, or changes in health care delivery. It is important to note though that in Rhode Island, ED use and hospitalization rates have increased over the last decade. 24 Second, we used a “payer” perspective, which does not include “spillover effects” for family members other than the child with asthma. Potential spillover effects include, for example, decreased healthcare use for other family members, or missed work/school days for caregivers. These effects may be particularly relevant for our group of older children, who had increased controller medication use. Although this use was associated with increased medication cost, it may have been associated with improvements in asthma control, not measured in the “payer” framework of the current R01. As a result, our conservative ROIs may have actually underestimated the true returns that could be achieved with the PAQS project. A third limitation is that the study population was predominantly female with relatively low numbers of cigarettes smoked per day, both factors which may have contributed to the relatively high smoking cessation rates.