OBJECTIVES: Multiple challenges exist in accurately estimating the birth prevalence of rare diseases. In addition to difficulties in accurately identifying and classifying patients, epidemiological estimates are often based on a small number of cases. There is therefore the potential for substantial differences in estimated rates based on very small changes in the number of cases observed. The objective was to investigate the sensitivity of epidemiological estimates of rare disease birth prevalence, to the underlying number of cases identified.

METHODS: As a case study, estimates of birth prevalence of Pompe disease were identified from a published literature review of newborn screening studies for this rare condition. The impact on birth prevalence of varying the number of identified cases, assuming ≤2 additional or fewer cases were identified, was calculated and explored graphically, based on studies including representative population-based samples.

RESULTS: Birth prevalence estimates from included studies ranged from 1:4,447 (9 cases among 40,024 newborns) to 1:27,886 (4 cases; 111,544 newborns) live births (LB). In the largest study (28 cases; 473,738 newborns), birth prevalence was estimated at 1:16,919 LB. With two additional or fewer cases identified, estimates would range from 1:15,791 to 1:18,221 LB. Comparatively, in the smallest study (3 cases; 27,724 newborns; birth prevalence 1:9,241 LB), with two additional or fewer cases identified, estimates would range from 1:5,545 to 1:27,724 LB.

CONCLUSIONS: Accurate and precise measurement of the birth prevalence of rare diseases is important to understanding their true clinical and economic burden at the population level. This case study demonstrates the dramatic differences that might be observed in estimates of rare disease prevalence based on a difference of a few cases observed, particularly in studies with small sample sizes. Using standardized definitions in large representative samples in studies of rare diseases is critical to understanding the true burden of these conditions.