Does capture-recapture analysis provide more reliable estimates of the incidence and prevalence of leg ulcers in the community?

Abstract

Objective: To determine whether capture-recapture analysis provides more reliable estimates of the cumulative incidence and prevalence of leg ulcers in Auckland, New Zealand.

Methods: A population-based, cross-sectional study was conducted in the Central and North Auckland health districts of New Zealand in 1998. Cases were identified through health professional referral and by self-notification. All ages and ulcer types were investigated. Both traditional and capture-recapture methods of analysis were used to estimate the cumulative incidence and prevalence of leg ulcers in the study population.

Results: Four hundred and twenty-six people with current leg ulcers were identified during the 12-month study period. Using traditional methods of analysis, the annual cumulative incidence rate of leg ulcers in Auckland was 32 per 100,000, with a point prevalence of 39 per 100,000 and a period prevalence of 79 per 100,000 per year. Results from capture-recapture analysis, however, suggest an annual cumulative incidence rate of 252 per 100,000, with a point prevalence of 248 per 100,000 and a period prevalence of 530 per 100,000 per year.

Conclusions: The traditional method of calculating cumulative incidence and prevalence clearly under-estimates the frequency of leg ulcers in the Auckland region. Capture-recapture analysis provides a more reliable estimate of disease frequency, since cases that remain unidentified in the population are considered.