Journal Article Annotations
2022, 1st Quarter
Annotations by Scott A. Simpson MD MPH
There remains significant ambiguity in selection of first line agents for agitation in emergency psychiatry. Intramuscular haloperidol remains widely used, but comparative studies with this agent and how to mitigate its risks are not common. The investigators conducted a retrospective chart review to compare the efficacy and safety of haloperidol+lorazepam with haloperidol+lorazepam+diphenhydramine for the treatment of agitation in the ED. Four hundred charts were reviewed from among the health system’s 21 EDs across 6 states. There was no difference in need for repeat medication administration between the combinations (14% v 20% of patients respectively), but patients also receiving diphenhydramine were more likely to have hypotension, oxygen desaturation, or require physical restraints. Adding diphenhydramine more than halved the risk of requiring treatment for extrapyramidal symptoms (EPS) (2% v 4%).
Strength and weaknesses:
The investigators identified potential cases electronically across multiple sites and regions. Their inclusion of undifferentiated agitation improves applicability of these findings to practice. While the chart abstraction methodology was well-described, charts were coded only once which may have introduced error. The outcomes are fairly dichotomous so interrater reliability would likely be high. Statistical testing was not adjusted for multiple comparisons, and differences between medication selections may be overstated. EPS were only captured if they required medication treatment; however, EPS may be delayed after intramuscular haloperidol, and EPS occurring after ED departure (eg, on an inpatient unit or after discharge) may not have been captured. No analyses for confounding, including by race or other social determinants, were included. There is likely confounding by indication (as likely reflected in the higher rate of restraint among the diphenhydramine cohort); the authors do not well describe why one providers in their system would select one of these medication options versus the other.
This real-world study suggests that adding diphenhydramine to haloperidol and lorazepam is unnecessary in most instances. Generally, excluding diphenhydramine is preferable given that the drug must be delivered in a different syringe than haloperidol (thereby producing unneeded discomfort and safety concerns) whereas haloperidol and lorazepam can be administered together. However, patients with high risk of EPS may benefit from the addition of diphenhydramine or alternative therapy, such as an intramuscular atypical antipsychotic.
This randomized trial recruited 756 adolescents with risky alcohol use from 3 EDs in the United Kingdom. After screening, adolescents were randomized to receive no intervention; brief feedback with personalized advice, or brief feedback with a brief electronic intervention. The brief electronic intervention was a phone app that “uses the concept of game-playing, in which users explore, navigate, learn facts about alcohol, record alcohol consumption, receive personalised feedback and set goals in an engaging city-scape format with the aim of supporting users to reduce or stop alcohol consumption.” The interventions were not associated with a decrease in alcohol consumption, and alcohol consumption increased among all three cohorts.
Strength and weaknesses:
This is a rigorously designed and reported trial. The primary outcome was clearly identified, and secondary outcomes on behavioral change and cost effectiveness add further validity and nuance to the findings. (There was no impact of the intervention on these outcomes, either.) The authors suggest that screening alone may raise awareness of problematic alcohol use among adolescents, but the increase alcohol consumption in even that control group belies the suggestion that screening alone is an effective intervention. There may be selection bias, as the highest risk drinkers may have been unwilling to participate in the study, but this would likely bias towards an effect and is unlikely to have changed the findings. The outcomes were primarily based on self-report and alcohol consumption rather than other functional measures, such as academic progress or legal involvement, that might better signify the presence of a use disorder.
Adolescent substance use is a common issue treated by emergency clinicians. These findings confirm that these youth remain at risk of escalating use. While brief ED-based intervention are feasible, their utility remains unclear. Among adults, brief interventions in the ED appear most effective for lower-risk users of alcohol rather than patients with more severe use or non-alcohol use disorders. Perhaps substance use among adolescents is reflective of more severe disease, or adolescents require more intensive treatments that include robust support for care transitions and treatment retention after an ED visit. This study highlights the need but, alas, not the solution.