Epidemiology of cystic fibrosis: A review

Abstract

Three main reasons have been advanced for the degree of ignorance and confusion surrounding epidemiologic measures of the incidence and prevalence of Cystic Fibrosis, (CF) [83]. 1. 1.—Relatively recent recognition of CF as a specific disease entity. 2. 2.—Wide range of clinical manifestations, presentation as a multisystem disease, and great natural variation in the severity of involvement of various organs. 3. 3.—Past difficulty in establishing a diagnosis based on sound laboratory evidence, this being largely due to the lack of a screening test that had gained wide acceptance among diagnosticians [84], and inaccuracies in the performance and interpretation of the sweat test [85]. In addition to these factors, there are a number of methodological deficiencies and questionable assumptions in several of the papers that have here been reviewed. It may be noteworthy that the two sets of authors whose research design most closely matched [58, 74] being aimed at providing multiple ascertainment of cases and cross-checks of validity, have arrived at nearly the same estimated incidence of this particular disorder ( 1 3700 and 1 3800 respectively). The issue of selection and diagnostic criteria is one that has important implications for epidemiologic studies of cystic fibrosis. Particularly in a disease like CF, where the spectrum of clinical manifestations and severity of involvement is so broad, estimates of incidence will vary according to the criteria by which cases are ascertained and included. A greater degree of uniformity in standards than has been employed in previous studies is necessary before comparisons can meaningfully be made. There are good grounds for regarding most estimates of CF incidence to be minimum confirmed figures, and this is particularly so where the research design ensures that multiple ascertainments of the same case will be identifiable. Higher estimates are made where several sources of error (respondent bias, generalization of hospital data to the wider population, multiple ascertainments of the same patient) are not recognised.