The adequacy of suicide statistics for use in epidemiology and public health.

In spite of persistent criticisms regarding their validity, official suicide statistics continue to be used in epidemiologic studies that have substantial public health implications. We ask which epidemiologic findings might be affected by underreporting in the suicide rate. We estimate the extent of potential underreporting by comparing Canadian suicide rates with and without deaths of undetermined origin (UDs) added. Our results tend to confirm findings from other jurisdictions, specifically that underreporting is probably higher among females and cases of poisoning and drowning. The highest underreporting in the 1950-82 period was in 1977-1978. Overall, the average potential underreporting was found to be 17.5% for females and 12% for males. Still, underreporting is not sufficiently large that reasonable reformulations of the suicide rate substantially alter findings, suggesting that most epidemiologic conclusions based on official rates are essentially correct.