Invited commentary: asthma surveillance in US children.

Asthma is a highly prevalent and disabling disease of childhood. Data from national surveys indicate that 5–11 percent of children and adolescents in the United States— totaling over 4 million persons—have physician-diagnosed asthma (1–4). It is the most prevalent cause of childhood disability, with 1.4 percent of US children having disabling asthma (5). These data leave little room to doubt that asthma is one of the leading public health problems in the United States. Much of our knowledge about the burden of asthma in the US population comes from the data systems of the National Center for Health Statistics. The National Health Interview Survey (NHIS) has been used to track diagnosed asthma for over 30 years. Prior to 1997, the emphasis of the NHIS was to report the diagnosis of asthma. In 1997, the National Center for Health Statistics redesigned the NHIS. In this issue of the Journal, Akinbami et al. (6) try to make sense of the decline in the prevalence of asthma introduced by this redesign. The authors’ analyses are a credible attempt to correct the artifactual decrease in asthma prevalence due to the NHIS redesign. Still, one cannot help but question why the NHIS—a survey that is so central to our understanding of asthma trends and the allocation of resources to control asthma—was altered in the midst of what appears to be an epidemic. Updating the structure of surveys is important. Otherwise, our monitoring systems would quickly become obsolete. The redesign of the NHIS appears to be intended to measure asthma control rather than prevalence, as indicated by the change to a question about asthma “attacks” (6). However, without additional information about medication use, data on asthma control can be very misleading (7). For example, an asthmatic person who has not suffered from a recent attack because he or she uses inhaled steroids may not be counted in a survey that focuses on recent wheezing episodes or “attacks.” Thus, any observed decline may largely be a result of redefining asthma prevalence as the prevalence of uncontrolled asthma. Altering the design of the NHIS uncovered a largely unrecognized problem with our asthma surveillance system: It is a fragmented collection of data sets with no clearly stated roles and little coordination. Data collected in the various surveys conducted by the National Center for Health Statistics form the backbone of our asthma surveillance system. The National Death Index is used to measure rates and trends in asthma deaths (1, 4). The National Hospital Discharge Summary is used to monitor hospitalizations (1, 4, 8–10). The National Hospital and Ambulatory Medical Care Survey is used to monitor emergency room visits and ambulatory care (8, 9). The NHIS and the National Health and Nutrition Examination Survey (NHANES) are used to measure disease prevalence and disability (1–5, 11). These national surveys were not specifically designed for asthma, and they do not collect all the information we need to track asthma.