Abstract
Objectives—
To examine the efficacy of population screening for haemochromatosis by analysing the screening performance of seven intervention trials, and to compare this with the expected performance derived from family studies.
Setting—
Seven population intervention trials carried out between 1983 and 1995 in Australia, Scandinavia, Iceland, and the United States.
Methods—
Seven of 23 English language trials identified were suitable for the meta-analysis. Transferrin saturation and serum ferritin measurements derived from family studies were used to predict detection and false positive rates for each trial.
Results—
The seven trials used various screening and diagnostic criteria. A total of 18 396 men and 12 254 women were screened. Because some cases were not detected by screening, and some screen positive individuals did not complete diagnostic testing, the prevalence of homozygous individuals was underestimated in all the trials. The reported and predicted percentages of screen positive individuals nearly always agreed. The homozygote prevalence was estimated to be 34 men and 40 women per 10 000 (prevalence predicted from family studies is 53 per 10 000). Clinical manifestations were present in 50% of male and 44% of female homozygotes.
Conclusions—
False positive rates, homozygote prevalences, and frequency of clinical manifestations were in general agreement with predictions from family studies. However, incomplete understanding about a number of issues requires that further pilot trials be carried out before screening can be considered part of routine medical practice.