Lysergide Treatment of Schizophrenic and Nonschizophrenic Alcoholics; A Controlled Evaluation

Earlier research on lysergide (LSD) as a treatment for alcoholism began in the 1950s with Hoffer and Osmond and emphasised producing a controlled altered state that might motivate abstinence. Over the next decade a number of reports—many from Canadian groups—described promising therapeutic effects, but several were criticised for weak methodology, chiefly lack of appropriate controls, blind assessment, and standardised outcome measurement. Subsequent controlled projects produced mixed results, with some early controlled trials finding no differences and others reporting benefits, highlighting uncertainty about whether observed changes were transient or durable once patients returned to ordinary life. Tomsovic and colleagues initiated this study (data collection September 1965 to April 1968) to provide a controlled evaluation of adding a single, high-dose lysergide experience to an existing 90-day Alcoholic Rehabilitation Programme. Although the original plan was to have comparable proportions of schizophrenic and nonschizophrenic alcoholics across treated and control groups, preliminary analyses drew attention to differential effects by psychiatric diagnosis and prompted a focussed comparison of outcomes for schizophrenic versus nonschizophrenic patients. The investigators acknowledged substantial methodological challenges in evaluating a spectacular mind‑altering drug and aimed to report one-year follow-up outcomes, psychometric measures of the acute experience, and the incidence of post‑treatment recrudescence.

Design and setting: The study recruited volunteers from a 90-day Alcoholic Rehabilitation Programme. Volunteers were paired with contemporaneous admissions and, when participants were about halfway through the programme (≈45 days), one of each pair was assigned to receive lysergide by coin toss. The project also used two types of controls: control group I (volunteers who did not receive lysergide or received a placebo substitute) and control group II (a larger group of routine programme participants who either had completed the programme before the lysergide project began or did not volunteer). Participants and diagnostic ascertainment: Seventy-five volunteers received lysergide; among these, 23 were classified as schizophrenic. The extracted text reports 30 schizophrenics in control group I and 60 in control group II. Diagnoses of schizophrenia were based on past medical history, MMPI profile and clinical appearance; about 65% of those labelled schizophrenic were judged to have a high level of diagnostic certainty, with the remainder described as more ambiguous cases (possibly schizoid personality by other clinicians). Intervention and setting of administration: Treated patients received a single oral dose of 500 µg lysergide early in the morning in a specially furnished, quiet room designed to support a psychedelic experience (colourful drapes, pictures, music, props). During the first seven hours a patient‑chosen nursing assistant or nurse remained with the patient for non‑directive support and made written observations. The administering psychiatrist made intermittent visits. Patients were given preparatory readings and a predischarge debrief the next day in group therapy. No additional drugs were routinely used to terminate the experience or induce sleep. Outcome assessment and psychometrics: Acute and subacute reactions were measured with the Blewett and Chwelos Scales. Scale 1, a 66‑item self‑inventory, was completed the day after the session and provided 22 items used to gauge the conventional therapeutic response; Scale 2 was administered 30–45 days later to assess persistence of attitudinal changes. Investigators also made a subjective classification of each experience ("good", "moderate", "poor") based on session observations, narrative and written accounts, and post‑session behaviour, and rated the spiritual quality when present. Postdischarge follow-up and validation: Drinking control was assessed with a Drinking Adjustment Scale administered at discharge and by postal or telephone follow-up at 3, 6 and 12 months. The scale used six drinking control categories that were later collapsed for analysis. Validation calls were made to spouses or other informants in many cases; at 12 months validation rates reported in the extracted text were 50% for the lysergide group, 37% for control group I and 54% for control group II. Patients who were lost to contact, died, rehospitalised for similar treatment, or incarcerated for prolonged periods were removed from the analysis according to case‑by‑case adjudication. Additional measures and analysis: After early spontaneous reports of post‑experience recurrences, the investigators added items to follow-up forms to record recrudescence (defined as a vivid perceptual, thought or feeling experience clearly related to the lysergide session). Primary comparative analyses used chi‑square tests; the authors report dichotomous comparisons of complete abstinence versus non‑abstinence to amplify differences that appeared concentrated in the abstinent category. The extracted text does not present comprehensive modelling or other inferential methods beyond the chi‑square tests described.

Sample composition and follow-up: Seventy‑five patients received lysergide, of whom 23 were diagnosed as schizophrenic. Control groups included schizophrenic patients as reported above. There was notable attrition over the 12‑month follow‑up due to inability to contact patients, deaths and methodological removals (for example, rehospitalisation or prolonged incarceration); the authors state the percentages lost were approximately similar across groups. Validation of self‑reports at 12 months was partial: 50% validated in the lysergide group, 37% in control group I and 54% in control group II. Psychometric and subjective experience: Scale 1 scores (Blewett and Chwelos) yielded a statistically significant chi‑square test when cases were split at a reference score of 16.5 (p<.05), whereas Scale 2 was not significant. Nonschizophrenic patients showed a slightly higher proportion of "good" experiences and a slightly lower proportion of "poor" experiences than schizophrenic patients; spiritual content distributions were reported as nearly identical between groups. Two lysergide‑treated patients (both classified as nonschizophrenic) exhibited marked loss of contact with reality during the acute experience and had incomplete recall afterwards. Drinking outcomes: When drinking control was dichotomised into complete abstinence versus non‑abstinence, nonschizophrenic lysergide‑treated patients had the highest percentages abstinent at 3 months and maintained a relative advantage through 12 months. The nonschizophrenic lysergide group was superior to other groups in varying degrees; the difference between nonschizophrenic lysergide patients and nonschizophrenic control group I reached statistical significance beyond the .01 level. However, the authors caution that control group I showed unexpectedly poor outcomes relative to control group II, raising concerns about a deprivation or expectation effect among volunteers who did not receive lysergide. Overall, across the full set of comparisons the investigators judged that differences did not support a clear conclusion that lysergide was beneficial as an add‑on to the regular programme. Predictors and recrudescence: Attempts to predict benefit from measures of the acute experience found no relationship between any of the investigated measures (psychometric or subjective classification) and drinking control at 3 months or 12 months. Regarding recrudescence, 61 patients provided information and 19 (31%) reported at least one vivid recurrence during the year; an additional 7 patients (11%) reported altered dreaming that might be related. Recrudescence tended to be reported more frequently among patients who drank post‑discharge, roughly twice the rate observed in abstainers. Timing of first recrudescence was concentrated at the 3‑month point (65% of first occurrences), with a minority occurring earlier or later, and several patients experienced multiple recurrences (distribution: 22% one, 32% two, 36% three, 10% four or more). The investigators report no instances of the severe, prolonged complications (for example sustained psychosis or suicidal behaviour) described in other literature; the recrudescences they recorded were relatively brief and varied from uncomfortable to pleasant. Sensitivity to missing data: The authors note that removal or loss of cases likely biases observed abstinence upward; if all lost cases were assumed non‑abstinent, abstinence percentages in the "completely abstinent" category would fall by approximately 4–17% depending on group.

Tomsovic and colleagues interpret their findings cautiously: although nonschizophrenic lysergide‑treated patients had higher abstinence rates at measured intervals, the overall pattern did not allow the investigators to conclude that lysergide provided a clear therapeutic benefit when added to the regular rehabilitation programme. They highlight a statistically significant poorer outcome among volunteers who expected lysergide but did not receive it, which the authors interpret as a possible "deprivation effect" driven by disappointment and expectation; an alternative explanation suggested by staff observations is that volunteers were a self‑selected group less engaged in ordinary programme work because they awaited a drug experience, and therefore had a poorer prognosis irrespective of treatment allocation. The study also reproduces earlier observations of adverse outcomes among schizophrenic alcoholics: the investigators report a marked negative effect of lysergide in this subgroup and a statistically significant difference on Scale 1 reflecting a distinct acute reaction profile for schizophrenic patients. Given these findings the authors emphasise the need for caution with schizophrenic alcoholics in future research and clinical application. Limitations acknowledged by the investigators include progressive loss to follow‑up and case removals that likely biased outcomes in favour of better results, imperfect control conditions (notably the problematic volunteer control), and possible insensitivity of their psychometric and subjective measures to the aspects of the lysergide experience that might predict durable change. They conclude that recrudescence is a real, not uncommon post‑treatment phenomenon—occurring in about a third of respondents in this sample—with most episodes brief and none manifesting the severe complications that appear in case reports elsewhere. Finally, the authors recommend future studies incorporate a credible control drug experience to address expectation effects and to further investigate differential effects in psychiatric subgroups, while recognising the methodological challenges inherent in studying a dramatic, mind‑altering agent.