This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Cannabidiol Treatment in Patients With Early Psychosis (NCT02504151), a randomized phase 2 study. The primary outcome is Positive and Negative Syndrome Scale (PANSS) Over Time.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT02504151 |
| Phase | Phase 2 |
| Condition | Schizophrenia, Schizoaffective Disorder |
| Allocation | Randomized |
| Design model | Crossover |
| Masking | Quadruple |
| Enrollment | 18 |
| Arms in the primary outcome | n/a |
| Lead sponsor | Yale University |
| Results posted | Yes |
2. Clinical Question
Population
Participants with early psychosis and diagnoses of schizophrenia or schizoaffective disorder. The registry record lists an enrollment of 18. Further eligibility details, including age range and criteria for defining early psychosis, are not reported in the registry record.
Intervention
Cannabidiol (CBD), at a total dose of 800mg/day, taken for four weeks during one study period.
Comparator
Placebo, taken for four weeks during the other study period. Participants were randomized to receive CBD followed by placebo or placebo followed by CBD, with a washout between periods.
Primary question
How did PANSS scores change over time during the crossover study, and how did those scores compare during CBD and placebo periods? Lower PANSS scores indicate improvement in symptoms, but the registry record does not report the outcome values or a treatment comparison.
The trial’s stated hypotheses included a reduction in PANSS scores and improvement on the Clinical Global Impression of Severity scale. The registry identifies PANSS over time as the primary outcome measure. It also lists functioning and quality of life as secondary areas of interest, assessed with the Patient Assessment of Own Functioning Inventory (PAOFI) and the Quality of Life Scale (QLS).
The registry record does not report participant-level data, treatment-period estimates, or the results of a statistical comparison. It therefore cannot establish from the posted results whether scores changed differently during CBD and placebo periods, or how precisely any difference was estimated. The clinical question is clear from the design, but the results needed to answer it are not reported in the registry record.
3. Trial Design
This was a phase 2, randomized, quadruple-masked, crossover trial sponsored by Yale University. The registry lists an enrollment of 18 and a completed study status. The listed start date is 2016-01-28, and the primary completion date is 2018-03-02.
Participants were randomized in a 1:1 ratio to one of two treatment orders: CBD followed by placebo, or placebo followed by CBD. Each treatment period lasted four weeks. The brief summary describes a washout of at least 2 weeks between periods; the arm descriptions specify a two week washout. The registry does not report whether the washout duration varied for any participant.
Because participants received both treatments in sequence, each participant could contribute observations during both the CBD and placebo periods. This is different from a parallel-group trial, in which participants receive only one assigned treatment. The crossover structure can support within-participant comparisons, but interpretation depends on the analysis accounting appropriately for treatment order, period, and the possibility that effects of the first period persist into the next. The statistical analysis plan for these issues is not reported in the registry record.
The study description refers to double-blind conditions, while the design metadata lists quadruple masking. The registry does not identify which people were masked or explain the terminology. It also does not report the actual number randomized to each order, withdrawals, or the number contributing data at every scheduled assessment. The enrollment figure should not be read as the number analyzed for every outcome time point.
4. Participant Flow and Accounting
The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.
First Intervention (4 Weeks)
| Stage | Cannabidiol, Then Placebo | Placebo, Then Cannabidiol |
|---|---|---|
| STARTED | 8 | 10 |
| COMPLETED | 7 | 9 |
| NOT COMPLETED | 1 | 1 |
Reasons for not completing this period:
| Reason | Cannabidiol, Then Placebo | Placebo, Then Cannabidiol |
|---|---|---|
| Physician Decision | 1 | 1 |
Washout (2 Weeks)
| Stage | Cannabidiol, Then Placebo | Placebo, Then Cannabidiol |
|---|---|---|
| STARTED | 7 | 9 |
| COMPLETED | 6 | 9 |
| NOT COMPLETED | 1 | 0 |
Reasons for not completing this period:
| Reason | Cannabidiol, Then Placebo | Placebo, Then Cannabidiol |
|---|---|---|
| Lost to Follow-up | 1 | 0 |
Second Intervention (4 Weeks)
| Stage | Cannabidiol, Then Placebo | Placebo, Then Cannabidiol |
|---|---|---|
| STARTED | 6 | 9 |
| COMPLETED | 4 | 7 |
| NOT COMPLETED | 2 | 2 |
Reasons for not completing this period:
| Reason | Cannabidiol, Then Placebo | Placebo, Then Cannabidiol |
|---|---|---|
| Withdrawal by Subject | 0 | 1 |
| Physician Decision | 2 | 1 |
Analysis population for the primary outcome, as defined in the registry: The number analyzed in one or more rows differ from the overall number analyzed to reflect the number of participants who completed the scale during each time point.
5. Baseline Characteristics
The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.
| Characteristic | Unit | Cannabidiol, Then Placebo | Placebo, Then Cannabidiol | Total |
|---|---|---|---|---|
| Participants analysed | 8 | 10 | 18 | |
| Age, Continuous | years | 22.875 (19 to 26) | 26.1 (22 to 29) | 25 (19 to 29) |
| Sex: Female, Male: Female | Participants | 2 | 2 | 4 |
| Sex: Female, Male: Male | Participants | 6 | 8 | 14 |
| Ethnicity (NIH/OMB): Hispanic or Latino | Participants | 3 | 1 | 4 |
| Ethnicity (NIH/OMB): Not Hispanic or Latino | Participants | 5 | 9 | 14 |
| Ethnicity (NIH/OMB): Unknown or Not Reported | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): American Indian or Alaska Native | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): Asian | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): Native Hawaiian or Other Pacific Islander | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): Black or African American | Participants | 4 | 5 | 9 |
| Race (NIH/OMB): White | Participants | 1 | 3 | 4 |
| Race (NIH/OMB): More than one race | Participants | 1 | 2 | 3 |
| Race (NIH/OMB): Unknown or Not Reported | Participants | 2 | 0 | 2 |
| Region of Enrollment: United States | participants | 8 | 10 | 18 |
6. Primary Endpoint
The registry lists the primary outcome as Positive and Negative Syndrome Scale (PANSS) Over Time, measured over Completed each time subject is seen over 10 weeks (screening, period 1 baseline, period 1 week 1, period 1 week 2, period 1 week 3, period 1 week 4, period 2 baseline, period 2 week 1, period 2 week 2, period 2 week 3, period 2 week 4).
| Element | Registry entry |
|---|---|
| Result type | Least Squares Mean |
| Measure of spread | Standard Error |
| Unit | score on a scale |
| Endpoint classification | Continuous Endpoint |
The primary outcome was the Positive and Negative Syndrome Scale (PANSS) over time. The registry describes PANSS as a 30-item scale assessing positive and negative symptoms of schizophrenia, with a total score range of 30-210. A lower score reflects an improvement in symptoms. The measure is reported as a score on a scale, with least-squares means and standard errors specified as the summary format.
The registry lists assessments at screening, period 1 baseline, period 1 weeks 1 through 4, period 2 baseline, and period 2 weeks 1 through 4. This repeated-measure approach describes the course of scores across the study rather than reducing the outcome to a single final visit. It could help characterize whether scores changed during each treatment period, but interpretation requires clarity about which contrasts were planned and how repeated observations were handled. Those analysis details are not reported in the registry record.
PANSS is used in this trial because the stated research question concerns changes in psychosis symptoms. The direction of the scale is important: lower scores indicate improvement, whereas a higher score does not. The registry record does not report the actual least-squares means, standard errors, or a treatment contrast, so the magnitude and precision of any difference cannot be described.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry record identifies the primary outcome and the intended comparison, but it does not report PANSS results. No treatment-period means, change estimates, confidence intervals, or inferential test results are available in the registry record for this analysis. As a result, there is no posted estimate from which to describe the direction, magnitude, or precision of the CBD-versus-placebo comparison. The presence of an outcome definition is not itself evidence of a treatment effect.
The crossover design means that the intended comparison involves two treatment periods experienced by participants in different orders. A sound interpretation would need to distinguish treatment differences from changes associated with time or period, and to consider whether the sequence of treatments matters. The registry does not report how the analysis handled those features, whether it included period or sequence terms, or whether carryover was assessed. These are relevant design questions, not evidence that a particular bias occurred.
The enrollment of 18 describes the study’s registered enrollment, not the number with complete PANSS observations at every visit. The outcome description explicitly notes that the number analyzed can differ across time points, reflecting completion of the scale at particular visits. The visit-specific numbers are not reported in the registry record. Without them, the extent and pattern of missing outcome data cannot be characterized.
The registry lists PAOFI and QLS over time as secondary outcomes and describes hypotheses involving the Clinical Global Impression of Severity scale. Their numerical results are not reported in the registry record. The number of outcomes or analyses listed does not reveal whether or how statistical multiplicity was controlled, and the analysis plan is not reported.
Accordingly, the posted information supports a description of the trial’s question, design, and measurement schedule, but not a conclusion about the observed primary outcome. It is not possible to judge whether an estimated effect was clinically meaningful, statistically distinguishable from no difference, or estimated with useful precision. Those judgments require results and analysis details that are not reported in the registry record. Any sample-size planning illustration based on an observed effect is a planning exercise, not a measure of the trial’s observed power or proof that an effect occurred.
9. Statistical Methods Explained
The registry specifies a continuous outcome summarized as a least-squares mean with a standard error. In general, a mean describes the central level of a numeric measure, while a least-squares mean is a model-based estimate that can account for factors included in the statistical model. The registry does not state which model terms were used to obtain the PANSS estimates. Without those details, the adjustment represented by the least-squares means cannot be described.
A standard error describes the estimated uncertainty in a statistic, such as a mean. It is not the same as the spread of individual participants’ scores: a standard deviation describes participant-level variation, whereas a standard error concerns uncertainty in an estimate. The registry identifies standard error as the dispersion measure, but does not report its numerical values for PANSS. It is therefore not possible to assess the precision of the outcome estimates from the record.
In general, a confidence interval expresses a range of values compatible with the data and the method used to construct the interval. Its width reflects precision: a wider interval indicates greater uncertainty about the estimated quantity. A confidence interval is not reported for the primary comparison in the registry record. Nor are a test statistic or p-value for that comparison reported. Without an estimate and uncertainty interval, the evidence cannot be characterized in terms of either magnitude or precision.
Why the crossover structure matters
In a crossover study, treatment order is part of the design. Each participant receives both interventions, so observations from the same participant are related rather than independent. A suitable analysis generally needs to recognize this within-participant structure. It may also need to account for period and sequence, depending on the prespecified model and scientific question. The registry record does not report the model, the planned contrast, or how within-participant correlation was handled.
A crossover comparison can be affected if the outcome changes over time for reasons unrelated to treatment, or if the effect of one period persists into the next. A washout is intended to separate periods, but its presence alone does not establish that carryover is absent. Here, the registry describes a washout of at least 2 weeks and arm descriptions refer to a two week washout. Whether carryover was evaluated and what assumptions were used are not reported in the registry record.
Repeated measurements and interpretation
PANSS was scheduled at multiple points across screening and the two treatment periods. Repeated measurements can describe a trajectory, but they also raise questions about how visits were modeled, how baseline values were used, and which time points or contrasts were designated as primary. The registry names PANSS over time as the primary outcome and lists the assessment schedule. It does not report a specific model for repeated measures, a primary time-point contrast, or a rule for handling incomplete follow-up.
In general, over-dispersion matters when observed variation is greater than a statistical model expects. A model that does not accommodate the data’s variability can give uncertainty estimates that are too small or otherwise poorly calibrated. The registry does not report an over-dispersion assessment or indicate that an over-dispersion model was used; this is a general statistical consideration, not a claim about this trial’s data.
Planning illustrations are not observed power
In general, sample-size planning asks how many participants would be needed to detect a specified effect under stated assumptions about variability, allocation, and the significance threshold. The planning assumptions described for this trial use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the target effect expressed as a fraction of the observed effect. Such illustrations are not the trial’s original sample-size justification unless the registry says they were used in planning; that justification is not reported in the registry record.
A planning calculation based on an observed effect is not observed power. It does not establish that the study had a particular probability of detecting its effect, and it does not substitute for the observed estimate and its uncertainty. The reported enrollment of 18 is not, by itself, enough to determine power because the calculation also depends on the outcome variability, analysis, and effect being targeted. Those inputs and any numerical planning results are not reported in the registry record.
10. Multiplicity, Missing Data and Analysis Populations
The primary outcome is PANSS over time, while the registry also lists PAOFI and QLS over time as secondary outcomes and describes a hypothesis involving the Clinical Global Impression of Severity scale. Testing several outcomes, visits, or contrasts can increase the chance of a nominally notable finding arising by chance. In general, multiplicity procedures address this issue by defining a primary comparison or adjusting how evidence is assessed across multiple tests.
The registry record does not report whether a multiplicity adjustment was planned or used, which PANSS time point or contrast was designated as the main test, or how secondary outcomes were treated in the inferential analysis. It also does not report a statistical analysis plan or the test results. The count of listed outcomes and analyses does not by itself establish how many hypotheses were tested or whether adjustment was appropriate.
Missingness is relevant because the registry notes that the number analyzed can differ across PANSS time points, reflecting the participants who completed the scale at each visit. Visit-specific analysis counts, reasons for missing measurements, and the pattern of missingness are not reported in the registry record. The record also does not describe imputation, a repeated-measures approach for incomplete data, or sensitivity analyses assessing alternative assumptions about missing values.
The registry does not state whether all randomized participants were included in the primary analysis, whether an intention-to-treat population was defined, or whether any per-protocol or safety population was used. These analysis-population details are not reported in the registry record. Consequently, the reported enrollment cannot be treated as the denominator for every outcome analysis, and the effects of missing data or population selection cannot be evaluated from the record.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Cannabidiol, Then Placebo | 2 | 8 | 25.0% |
| Placebo, Then Cannabidiol | 0 | 10 | 0.0% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Cannabidiol, Then Placebo | 0 of 8 |
| Placebo, Then Cannabidiol | 0 of 10 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 5% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Cannabidiol, Then Placebo | 5 of 8 |
| Placebo, Then Cannabidiol | 6 of 10 |
Most frequent serious adverse event terms
| Event term | Organ system | Cannabidiol, Then Placebo | Placebo, Then Cannabidiol |
|---|---|---|---|
| Hospitalization | Psychiatric disorders | 2 of 8 | 0 of 10 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Cannabidiol, Then Placebo | Placebo, Then Cannabidiol |
|---|---|---|---|
| Dry mouth | General disorders | 1 of 8 | 2 of 10 |
| Diarrhea | Gastrointestinal disorders | 1 of 8 | 1 of 10 |
| Depressed mood, racing thoughts | Psychiatric disorders | 1 of 8 | 0 of 10 |
| Nausea | Gastrointestinal disorders | 1 of 8 | 0 of 10 |
| ER Visit | Reproductive system and breast disorders | 1 of 8 | 0 of 10 |
| Tension headache | General disorders | 1 of 8 | 0 of 10 |
| Migraine | General disorders | 1 of 8 | 0 of 10 |
| Sedation | General disorders | 1 of 8 | 0 of 10 |
| Scratchy throat and cold symptoms | General disorders | 0 of 8 | 1 of 10 |
| Stye in left eye | Eye disorders | 0 of 8 | 1 of 10 |
Serious events with the highest arm-level rate: hospitalization (25.0%).
The registry reports serious adverse events by treatment order: “Cannabidiol, Then Placebo” and “Placebo, Then Cannabidiol.” In the first group, 8 were at risk for serious adverse events and 2 were affected; in the second, 10 were at risk and 0 were affected. The reported death count is 0 in each group. These are group-level figures attached to treatment sequences, not a separate count for each treatment period.
Because this was a crossover trial, participants in each sequence received both CBD and placebo at different times. The sequence-level counts therefore cannot be interpreted as a direct CBD-versus-placebo comparison of adverse events. The record does not specify the timing of each event relative to treatment, whether an event occurred during one or both periods, or how events were attributed to a treatment.
Other adverse events were reported for 5 of 8 at risk in the CBD-then-placebo group and 6 of 10 at risk in the placebo-then-CBD group. The most frequently listed serious-event term is hospitalization, reported as 25.0%. The registry does not explain the denominator underlying that percentage or provide further detail about event timing. These counts describe the posted safety record and do not, on their own, establish a comparative risk.
12. What the Result Does, and Does Not, Mean
The registry record makes the trial’s question identifiable: a randomized crossover comparison of CBD and placebo, with PANSS measured repeatedly as the primary outcome. It also describes the treatment orders, dose, period duration, washout, masking designation, and outcome scale. Those details help readers understand what the study set out to assess.
The record does not report the numerical primary outcome results. It therefore does not show whether PANSS scores differed between treatment periods, how large any difference was, or how uncertain an estimate might be. No claim that CBD improved symptoms, failed to improve symptoms, or had a particular effect can be supported from the posted primary outcome information. The same limitation applies to the listed secondary outcomes: their numerical results are not reported in the registry record.
Even if an estimate were available, a difference in a scale score would need to be interpreted in context, including its precision, the analysis population, missing observations, and the handling of repeated measurements and treatment order. The record does not provide the analysis details needed to assess these issues. A lack of reported results is not evidence of no effect; it means the treatment contrast cannot be evaluated from this record.
The safety figures are also limited. They are presented by randomized sequence, although participants received both study treatments, and the event timing and treatment attribution are not reported. They should not be read as treatment-specific comparative rates.
Thus, the most defensible conclusion is about reporting, not efficacy: the registry describes a completed phase 2 crossover trial but does not report the results necessary to estimate the primary treatment contrast. The study’s enrollment and design alone cannot resolve that uncertainty.
13. Why This Trial Matters Statistically
This trial is statistically notable for its crossover structure. When participants receive both treatments in sequence, the design can use within-participant information, but the analysis must respect the relationship between repeated observations. Treatment order, period effects, possible carryover, and incomplete follow-up all become important to interpretation. These considerations make the analysis plan particularly relevant, yet that plan is not reported in the registry record.
The primary measure was collected at multiple assessments across two treatment periods. This creates a distinction between describing a trajectory and estimating a prespecified treatment contrast. A long sequence of measurements can support different summaries, and the intended primary contrast is needed to understand what inference was planned. The registry names PANSS over time and lists the visit schedule, but does not report which contrast or statistical model was used.
The trial also illustrates why design information and outcome evidence are different. Randomization and masking describe safeguards in how treatment was assigned and delivered; they do not supply the missing outcome estimates. Enrollment describes the registered study size, but it does not establish the number with usable observations at each visit or the precision of the treatment comparison.
Finally, registry safety figures grouped by treatment sequence are not equivalent to period-specific comparisons in a crossover study. Without event timing and attribution, the sequence-level counts have limited value for separating treatment-specific event patterns. The study therefore raises broadly useful questions about analysis, reporting, and interpretation, while its posted outcome information remains insufficient to answer the primary clinical question.
14. Limitations
- The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
- Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.
The central limitation for interpreting this trial is that the registry record does not report numerical results for PANSS, the primary outcome. It also does not report estimates or inferential results for the listed secondary outcomes. This prevents assessment of effect direction, magnitude, precision, or statistical evidence.
The registry does not report the statistical model, the prespecified treatment contrast, or whether period, sequence, baseline score, or within-participant correlation were accounted for. It does not say whether carryover was assessed. Although the washout is described as at least 2 weeks in the brief summary and as two weeks in the arm descriptions, the record does not say whether the planned interval was followed for each participant or whether it was considered sufficient for the analysis.
The registry notes that PANSS analysis counts can vary by time point, but does not report the visit-specific counts, reasons for missingness, or methods used to handle missing observations. The analysis populations and any sensitivity analyses are not reported. These omissions limit assessment of potential selection or missing-data concerns.
The masking metadata says quadruple, while the brief summary describes double-blind conditions; the registry does not explain who was masked. Eligibility details beyond the listed conditions and early-psychosis context are not reported. The safety counts are grouped by treatment order and lack event timing or treatment attribution. Together, these reporting gaps constrain interpretation; they do not demonstrate that a particular methodological problem occurred.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Cannabidiol Treatment in Patients With Early Psychosis — NCT02504151.
- Registry record last updated 2022-09-06; status last verified 2022-08.
- Results first posted 2022-09-06.
- Data retrieved from ClinicalTrials.gov on 2026-11-09.
- Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
NCT02504151 is a completed phase 2 randomized crossover trial sponsored by Yale University, with a registered enrollment of 18. It compared cannabidiol at 800mg/day with placebo, using two four-week treatment periods and a washout described as at least 2 weeks. Participants were assigned to CBD followed by placebo or placebo followed by CBD.
The primary outcome was PANSS over time, a 30-item symptom scale for which lower scores indicate improvement. PAOFI and QLS were listed as secondary outcomes. The registry describes serious and other adverse events by treatment sequence.
Numerical results for the primary and secondary outcomes, the statistical analysis plan, visit-specific analysis counts, and details of missing-data handling are not reported in the registry record. The posted information therefore describes the design and planned measurement, but does not permit a conclusion about the observed treatment effect.