The present results pose a different issue. We have noted the possibility that the intervention improved health-care utilization through some mechanism other than ETS exposure reduction. However, the latter interpretation cannot be ruled out, especially given the relatively large effect size and the relatively small sample. The dose-response relationship between reductions in ETS exposure and improvements in asthma outcomes is unknown at present. Modest reductions in ETS exposure actually may have important benefits in children with hyperreactive airways, and decreases sustained during the winter months may be especially important since that is the time when children and smokers tend to be indoors, ventilation is poorer, and viral respiratory tract infections are most prevalent. Further, prenatal exposure and the duration of prior exposure may modify the dose-response relationship. If these considerations are important, the failure of most ETS reduction studies to measure asthma outcomes as well as exposure, coupled with a relatively high variance in the measure of exposure and the rather low statistical power in some instances, may result in an underestimation of the value of such interventions. Future studies need to be appropriately powered to detect modest differences in exposure, and the measurement of exposure should span several seasons of the year. Conversely, a very small reduction in exposure or indoor concentration might prove statistically significant in a study with a very large sample but may have little clinical significance. The only way to guard against the misinterpretation of the results of such studies is to gather concurrent data on ETS exposure and disease outcomes. my canadian pharmacy
There is an additional measurement issue. While cotinine is the best available biomarker of ETS exposure, single, intermittent urine samples provide a relatively crude index of both typical and maximal exposure. We observed that 30% of the children from homes where smoking was not restricted had a measured cotinine level at baseline < 10 ng/mL. Similarly low levels were obtained at baseline in 47% of the children whose parents reported that they had moderate or heavy exposure. While it is possible that some of these parents may have overestimated the child’s exposure, it seems more likely that these children actually had exposure that was much higher at certain times than would be suggested by the measured value. If that is the case, the exposure and exposure-reduction signals conveyed by the results of isolated urine cotinine measurements might be relatively weak. This sampling problem could be remedied by more frequent or more carefully targeted measurements (eg, more frequent measurements during the fall and winter seasons). Adherence to such frequent measurements may be low, however, except in more motivated families, and the measurements themselves may be reactive.