A study can find associations with both ventricular arrhythmias and death without demonstrating that the deaths were caused by the arrhythmias. A large 2025 methamphetamine cohort illustrates why a composite outcome and a separate mortality outcome must retain their original definitions.
What were the study’s endpoints?
Noubiap and colleagues analyzed 29,593,819 adults receiving hospital-based care in California during 2005–2019; 690,737 had methamphetamine-use coding at some point. Their ventricular-arrhythmia composite combined ventricular tachycardia, ventricular fibrillation and cardiac arrest. All-cause mortality was a separate endpoint. Noubiap et al., 2025.
Methamphetamine coding was associated with the composite at adjusted HR 1.90 (95% CI 1.85–1.95) and with mortality at HR 1.51 (1.47–1.54). The analyses updated exposure and covariates across encounters and treated death as a competing risk for arrhythmia outcomes. Administrative misclassification, healthcare-based selection and unmeasured confounding limited causal conclusions. Methods and limitations.
Why cannot the two hazard ratios identify a fatal pathway?
A mortality endpoint records that a person died. An all-cause endpoint does not require that a particular arrhythmia initiated the death. Showing that exposure predicts both outcomes in separate analyses does not demonstrate that the first outcome mediated every death contributing to the second.
The two hazard ratios also answer different questions and have different outcome definitions. Subtracting them, multiplying them or treating one as the fraction of deaths explained by the other would create an analysis the investigators did not perform.
Why does cardiac arrest belong beside the composite label?
Readers may hear “ventricular arrhythmia” and assume every included event was a documented ventricular rhythm. The study’s operational composite was broader because it included cardiac arrest. Reporting that component makes the result reproducible and prevents a group label from implying more detailed rhythm documentation than an administrative endpoint provides.
A composite can improve event capture while combining outcomes with different clinical meanings. Its overall association does not require every component to have the same frequency, effect size or causal pathway. Component-specific claims therefore need their own results.
What does a competing risk add?
A person who dies can no longer develop a later newly observed event in the same way as a person who remains under follow-up. Accounting for that competing event makes the time-to-event analysis more appropriate. It does not turn death into an arrhythmia diagnosis or resolve the cause of an individual fatality.
The longitudinal exposure record likewise differs from an acute toxicology concentration. A time-updated code can support following associations over years. It cannot specify the dose, route or physiological state immediately before one cardiac event.
What can the study support in a case review?
The results strengthen the population-level rationale for examining stimulant exposure in cardiac investigations. They do not diagnose a fatal arrhythmia from methamphetamine presence alone, establish a concentration threshold or transfer directly to all patients receiving prescribed amphetamine.
A case-specific opinion should distinguish rhythm strips, resuscitation records, toxicology, structural heart findings and independently established exposure timing. The cohort supplies context for those records; it does not replace them. This is observational evidence published in 2025, not a controlled exposure experiment or a current treatment rule.
Source
Noubiap JJ, Dewland TA, Montenegro GC, Oo HH, Tseng ZH, Marcus GM. Illicit stimulants and ventricular arrhythmias: a longitudinal cohort study. European Heart Journal. 2025;46:3639–3647. doi:10.1093/eurheartj/ehaf282.