A survey of people who use stimulants can inform research involving methamphetamine without supplying a methamphetamine-specific overdose rate. Hughto and colleagues’ New England study illustrates why the denominator, event definition and drug category must remain visible when describing overdose experiences and emergency help-seeking.
Who entered the analysis?
The study recruited adults in Greater Providence, Rhode Island, and three Massachusetts communities between March 2022 and June 2023. Eligibility required illicit stimulant use within the previous 30 days. A modified respondent-driven approach at four community sites produced 230 survey participants; eight with missing primary-outcome data were excluded, leaving an analytic sample of 222.
Only 51/222, or 23.0%, reported methamphetamine use in the past 30 days, although 130/222 reported lifetime use. Recent crack cocaine use was reported by 206/222. Participants could report more than one substance. These are overlapping exposure histories, not separate randomized drug groups.
The headline percentages need their drug definition
Table 4 reports personal stimulant-overdose experience in 76/222 participants, or 34.2%, and witnessed experience in 94/222, or 42.3%. The abstract gives slightly different percentages. The table’s explicit counts provide the basis used here.
Among those describing their last personal event, 13/76 reported methamphetamine involvement. Among those describing their last witnessed event, 13/94 reported it. Drug categories were not mutually exclusive. The 34.2% overall figure therefore cannot be presented as the proportion of methamphetamine users who overdosed, much less the risk per methamphetamine dose or use episode.
What counted as an overdose?
All participants received the researchers’ broad stimulant-overdose definition before reporting personal and witnessed histories. The reports reflected participants’ recollections and interpretation of symptoms. The study did not clinically adjudicate every reported event or verify each event’s drug contents and concentrations with toxicology.
Questions about whether an overdose was possible, and how participants described one, were added later and asked of a 59-person subset at the final collection site. Responses from that subset cannot be attributed to all 222 participants. Open-ended notes were coded into categories and were not necessarily verbatim participant statements.
Emergency-call figures also have separate denominators
For the last personal event, 26/76 participants reported that 911 was called, or 34.2%. For the last witnessed event, the count was 45/94, or 47.9%. Of those 45 calls, nine were reportedly made by the participant personally: 20.0% of calls, rather than 20.0% of all 94 witnessed events.
These figures describe the reported last event, not every event a participant experienced. A report that no call was made does not establish that an event was medically mild or that informal responses were effective.
Association did not identify a causal barrier
The adjusted model associated extremely severe symptoms with greater odds that 911 was called for a personal event: adjusted odds ratio 5.25, with a 95% confidence interval of 1.42–19.40. That wide interval conveys uncertainty, and an odds ratio is not a risk ratio or proof that a particular symptom caused a call.
The study also found no statistically significant bivariate association between reported methamphetamine involvement in the last personal event and calling 911. That result does not prove equivalence between drugs; the subgroup was small. Broader explanations for help-seeking remained hypotheses requiring additional research.
What remains useful?
The study documents reported experiences and help-seeking gaps in a particular mixed-stimulant population. Recall bias, selection of survivors, regional recruitment, incomplete event verification and varying question denominators constrain generalization. It cannot estimate fatal-event risk or establish a dose-related threshold for methamphetamine toxicity.
Its practical scientific contribution is to improve how those questions are asked. Drug-specific inference needs drug-specific exposure and event data; a broad stimulant category should not become a methamphetamine conclusion through a change in wording.
Source
Hughto JMW, Kelly PJA, Vento SA, et al. Characterizing and responding to stimulant overdoses: findings from a mixed methods study of people who use cocaine and other stimulants in New England. Drug and Alcohol Dependence. 2025;266:112501; online 2024. doi:10.1016/j.drugalcdep.2024.112501. Historical mixed-methods survey; not a clinical toxicity trial or current legal guidance.
Leave a Reply