A Survey of American Psychiatrists’ Attitudes Toward Classic Hallucinogens

From the 1950s to the 1970s, classic hallucinogens (for example, LSD, mescaline, DMT, psilocybin, ayahuasca) were widely investigated and used in psychiatric practice and research, with roughly 10,000 patients and several thousand research subjects exposed during that period. Regulatory changes and rising recreational use in the late 1960s led to withdrawal of marketed supplies and restrictive legislation, after which clinical research in the United States largely ceased for several decades. Over the past decade, however, a resurgence of clinical research has occurred, driven by reduced stigma, a more favourable regulatory environment, and private research funding; recent studies cited by the authors include MDMA-assisted psychotherapy for PTSD, psilocybin trials for alcohol and nicotine use disorders and for depression/anxiety in life-threatening cancer, open-label work in treatment-resistant depression, and controlled trials of LSD for anxiety associated with life-threatening illness. Barnett and colleagues set out to gauge contemporary opinions among American psychiatrists about the safety and therapeutic potential of classic hallucinogens. The investigators focused explicitly on classic hallucinogens (excluding nonclassic agents such as MDMA) and hypothesised that many psychiatrists would remain guarded in their enthusiasm. They further hypothesised that younger age, male gender, more recent completion of training, and a psychotherapeutic treatment orientation would be associated with more favourable opinions toward these substances.

The study used an anonymous 14-item internet survey, approved by the Partners Human Research Committee, designed to assess attitudes toward classic hallucinogens. The instrument began with seven demographic items (age, sex, training level, decade of completion if applicable, primary daily activity—clinical care, research, or administration—primary treatment approach, and lifetime number of peer-reviewed publications). This was followed by seven opinion items querying agreement on five-point Likert scales (from "strongly disagree" to "strongly agree") about whether classic hallucinogen use: 1) increases risk of subsequent psychiatric disorders; 2) increases risk of long-term cognitive impairment; 3) should be illegal for recreational purposes; 4) is unsafe even under medical supervision; 5) shows promise as psychiatric treatments; 6) could improve psychotherapy outcomes when used adjunctively; and 7) deserves further research as potential psychiatric treatments. A random-number generator was used to select 1,000 psychiatrists from the American Psychiatric Association membership directory: 250 resident–fellows and 750 attending psychiatrists (including general members, fellows and distinguished fellows). The survey was e-mailed between 7 January 2016 and 17 March 2016, with reminder e-mails sent at 2 and 4 weeks. Data were captured using REDCap and analysed in Stata version 12.0. Descriptive statistics were calculated for each item. Fisher's exact test (two-tailed) compared demographic variables between trainees and attendings. Associations between the seven demographic variables and the seven opinion variables were tested using a nonparametric trend test across the five Likert levels. Statistical significance was set at p < 0.05. The authors noted that multiple comparisons were made and that they chose not to apply corrections such as Bonferroni because of the risk of inflating type II errors; they cautioned readers about the increased chance of type I errors as a result.

Of 1,000 psychiatrists contacted, 324 responded, giving an overall response rate of 32.4%. This included 93 of 250 trainees (response rate 37.2%) and 231 of 750 attending psychiatrists (response rate 30.8%). Among attending respondents, 97 (42.0%) had completed training before 1990. Trainees and attendings differed little on most demographic variables aside from age and predominance of clinical work. Respondents overall tended to view classic hallucinogens as potentially hazardous and appropriate to be illegal for recreational use. Nevertheless, a substantial minority expressed optimism about therapeutic potential: almost half of respondents agreed that hallucinogens show promise for treating psychiatric disorders, and nearly a third agreed they might improve psychotherapy outcomes when used adjunctively. Despite mixed views on therapeutic benefit and safety, a large majority strongly favoured further research on hallucinogens as potential psychiatric treatments. In subgroup analyses, three demographic variables were significant predictors of opinions: sex, trainee status, and age. Male respondents were generally more favourable toward hallucinogens than female respondents. Trainees and psychiatrists aged 40 years or younger were less likely to endorse concerns about risks and were more optimistic about therapeutic potential than more senior colleagues. No other demographic variable significantly predicted opinions, with one exception: respondents reporting no peer-reviewed publications were more likely to agree that hallucinogen use increases the risk for long-term cognitive impairment (p = 0.028). The extracted text does not report additional effect sizes, confidence intervals, or p-values for the other associations.

Barnett and colleagues interpret their findings as indicating that, although many American psychiatrists remain concerned about psychiatric and neurocognitive risks associated with classic hallucinogen use, a sizeable minority are optimistic about their therapeutic potential and most support further research. The investigators note that psychiatrists' concerns about risk parallel public opinion data from the 2014 National Survey on Drug Use and Health, where substantial proportions of respondents perceived great risk from LSD use. The authors contrast these attitudes with recent research suggesting that recreational hallucinogen use is not clearly associated with subsequent mental illness or cognitive impairment and with evidence that medically supervised administration can be safe. They also acknowledge documented adverse events from recreational use (for example, panic, psychosis, and hallucinogen persisting perception disorder) and rare case reports of suicide or violence, while noting that causal inferences from such reports are limited. The discussion addresses subgroup differences. Male respondents were more favourable toward hallucinogens; the authors suggest, speculatively, that this might relate to higher lifetime prevalence of hallucinogen use among men in population surveys, though they did not ask respondents about personal use. Trainees and younger psychiatrists were also more optimistic, a pattern the authors suggest could reflect greater exposure to contemporary positive literature and less familiarity with historically negative reports or with patients harmed by hallucinogens. The principal limitation highlighted is the modest response rate and the possibility of self-selection bias; the investigators caution that respondents with strong positive or negative views may have been more likely to participate. They argue, however, that internal comparisons between subgroups within the respondent sample (for example, men versus women, trainees versus attendings) should remain informative unless selection bias differentially affected those specific groups. Finally, the authors recommend further research into why sex differences in attitudes exist and suggest repeating the survey in 5 to 10 years to assess trends in psychiatric opinion.

Despite recent encouraging clinical findings, the surveyed sample of American psychiatrists remained generally concerned about potential adverse psychiatric and neurocognitive effects of classic hallucinogens, and a modest majority were sceptical that these agents have a role in psychiatric treatment, including as adjuncts to psychotherapy. Subgroup analysis showed that male and trainee psychiatrists were less concerned and more optimistic. Across the sample, however, there was substantial consensus that more research is warranted to assess the therapeutic potential of classic hallucinogens.