Background: Four cases of sudden death in children 12 years or younger during desipramine treatment were identified between 1986 and 1992. We evaluated whether these events support the hypothesis that exposure to therapeutic doses of desipramine contributes to the risk for sudden death in otherwise healthy children.
Method: The National Center for Health Statistics provided the baseline number of sudden unexplained deaths in children 5 to 14 years old. Data from the National Disease and Therapeutic Index were used to estimate the exposure to desipramine in children in the same age group. Since two of the four deaths were identified by 1987, we used the post-1987 experience as if it were a prospective period in which a causal association could be examined.
Results: The number of sudden deaths in desipramine-exposed children did not increase from 1986 to 1992 despite a marked increase in exposure. By using 4 to 6 months as the average lifetime of a desipramine prescription and a baseline rate of sudden death of 4.2 deaths/million/year in this population, the post-1987 period would account for 162,000 to 242,000 person-years of desipramine exposure. Although not statistically significant, this level of exposure corresponds to a relative risk of 2.1 (95% CI = 0.5 to 15) to 3.1 (95% CI = 0.8 to 22).
Conclusion: Although, based on our estimates, the evidence for an association between desipramine and sudden death in children aged 5 to 14 years appears weak, replication of our findings is needed with a more precise numerator (total number of deaths) and denominator (the appropriate conversion from drug appearance to actual exposure) before a firm conclusion on this subject can be drawn. Until then, even if remote, the possibility of an association between desipramine and sudden death in children stresses the importance of assessing risks and benefits when desipramine is used in pediatric patients.