Cystic fibrosis: a system for assessing and predicting progression

This study presents a radiography-based database scoring changes over time in a large population of patients with cystic fibrosis. The purpose of this database is to provide comparison for groups of patients undergoing experimental treatment to assess effect of the treatment. The data may also be used to compare individuals with their age-matched cohorts with cystic fibrosis.

From 230 patients, 3038 chest radiographs were scored using the Brasfield system. The scores from radiographs from all the patients were individually plotted for age, and a single age-based severity curve was created. The age-based severity curve was compared with similar curves derived from pulmonary function studies of a subset of the same patient population.

We found high inter- and intraobserver reliability. The difference between the observers averaged 1.3 Brasfield points, the scale of which ranges up to 25 points. The age-based severity curve was presented as mean Brasfield scores versus age (birth to > 30 years) plotted with 95% confidence limits; the curve was also plotted in percentiles. The rate of decline of this curve was similar to the decline of pulmonary function studies in this patient population.

The age-based curve, a structural anatomic parameter, differs from pulmonary function studies, which are functional. Thus the age-based severity curve provides an additional, independent basis for comparison between groups and individuals. It may be used for the initial assessment of lung disease and for gauging and predicting the rate of decline. The curve may be used as a long-range outcome criterion to evaluate new treatments in groups of patients with cystic fibrosis.