Despite the strengths and informativeness of the density measures over commonly used dichotomous measures, several limitations prevail. First, an important caveat is that these density measures may not be utilized by studies where information about biological relatives are not typically available, admittedly making the general applicability of these measures somewhat limited. Second, the AUC for density measures, while greater than those for dichotomous measures, were nevertheless low for diagnostic/screening contexts. Indeed, it has been observed that measures of diagnostic accuracy are very sensitive to the characteristics of the population in which the test accuracy is evaluated. Accordingly, in our analyses we show that diagnostic accuracy is substantially variable across groups, particularly for dichotomous measures. Third, more work is needed to express confidence in the diagnostic accuracy of an FH density score based on how many and which kind of relatives and their affectedness are known for a given individual. Moreover, it has been observed that some measures largely depend on the prevalence of the disorder/condition, while others are highly sensitive to the spectrum of the disorder/condition in the studied population.
