Clarifying Selected Citations and Interpretations in ‘Underage deaths associated with substance‐related harm in Taiwan: A national 10‐year study’

Wei et al. [1] report a valuable national post-mortem series of 47 substance-detected deaths among preschool-age children in Taiwan (2013–2022), highlighting methamphetamine detection (66%) and frequent attribution to drug intoxication, lung disease and head trauma. Because this study may inform child-protection and harm-reduction policy, we would like to request clarification on several statements where the cited references appear to support a more nuanced interpretation. First, the Introduction states that ‘32.7% of individuals who use substances continue substance use during pregnancy’ and cites Chasnoff et al. [2]. In that study, 32.7% refers to the proportion screening positive on the 4P's Plus across 7818 prenatal-care attendees, whereas continued use after pregnancy recognition was reported at a lower prevalence (15% of the total sample). A clearer distinction between any periconceptional use and continued use after pregnancy recognition may improve accuracy. Second, the Discussion recommends that mothers who use methamphetamine avoid breastfeeding for at least 48 hours after use or until methamphetamine is undetectable in the urine for at least 24 hours, citing Chomchai et al. [3]. However, that study concluded that methamphetamine disappears from breast milk approximately 1 day before maternal urine becomes negative and suggested breastfeeding may be considered safe when urine has been negative for ≥24 hours. If the 48-hour recommendation derives from another guideline, explicit citation would strengthen clarity. Third, the manuscript links substance use with choking and sudden infant death syndrome (SIDS), citing Saccomanno et al. [4] and Makarious et al. [5]. These sources primarily address broader epidemiological associations and prenatal opioid exposure. Given that the present case series is based on post-mortem detection rather than defined prenatal exposure cohorts, tempering causal language may enhance interpretability. Fourth, the absence of deaths among 5-year-olds is interpreted as consistent with a bimodal paediatric poisoning distribution, citing Riordan et al. [6], which focuses specifically on painkiller poisoning. Broader surveillance data might more robustly support general age-pattern conclusions. These clarifications may strengthen interpretability while preserving the important contribution of this forensic dataset. Jia-Min Yang: Writing—original draft; investigation. Lien-Chung Wei: Conceptualization; writing—review and editing; supervision. None.