Effect of Medical Marijuana on Neurocognition and Escalation of Use: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effect of Medical Marijuana on Neurocognition and Escalation of Use (NCT03224468), a randomized na study. The primary outcome is Mean Difference in Number of Cannabis Use Disorder Symptoms Averaged Over 2, 4, and 12 Weeks.

2. Clinical Question

The study asks whether people assigned to begin using medical marijuana immediately differ from people assigned to wait before beginning use, with particular attention to symptoms of cannabis use disorder. The registry record also describes questions about health outcomes, neurocognition, and brain-based changes, but it does not report results for every outcome named in the study description.

The population is not described in enough detail to identify eligibility characteristics beyond the conditions listed in the record: pain, insomnia, depression, and anxiety. The record does not report whether every participant had each listed condition, how diagnoses were established, or other eligibility criteria. Those details matter when considering which people the findings might describe.

The intervention group could begin using medical marijuana immediately. The comparator group agreed to wait 3 months before using it. The record does not report the product, dose, frequency, route of use, or whether use was standardized. Thus, the comparison concerns assignment to immediate access versus a waitlist, rather than a specified dose or formulation.

The primary question is whether the groups differed in their average number of cannabis use disorder symptoms across assessments at 2, 4, and 12 weeks. The registry reports a higher average symptom count in the immediate-access group than in the waitlist group. This is a group-level result for a symptom scale; it does not by itself describe every participant's experience or establish why the groups differed.

3. Trial Design

This was a randomized, double-masked, parallel-group study. Random allocation is intended to make the groups comparable, on average, at the point of assignment. A parallel design means that participants were assigned to one arm rather than crossing between the two assigned conditions. The registry lists an enrollment of 269 and marks the study as completed. Enrollment is not the same as the number contributing data to a particular outcome: the primary outcome results report 105 participants in the medical marijuana arm and 81 in the waitlist control arm.

The medical marijuana arm could begin use immediately. The waitlist control arm agreed to wait 3 months before using medical marijuana. Because the comparator is a waitlist, the contrast is between immediate access and delayed access, not between medical marijuana and a placebo treatment. The registry record does not describe how masking was maintained in practice, who was masked, or whether participants and staff could infer assignment. It also does not report adherence, crossover, or actual use during follow-up.

The study is listed with phase “NA.” The record gives a start date of 2017-07-01 and a primary completion date of 2021-02-05. These dates describe the registered study timeline; they do not provide information about the duration of each participant's follow-up beyond the stated outcome timeframes. The sponsor is Massachusetts General Hospital.

The primary measure was assessed at 2, 4, and 12 weeks and reported as an average across those time points. The record also lists physical and mental health scores and several cognitive task measures as secondary outcomes. Although the study description mentions brain-based changes, the registry record does not identify a posted result for a specific brain-based measure. Original design details, including the planned sample-size justification and the full analysis plan, are not reported in the registry record.

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.

Overall Study

Reasons for not completing this period:

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.

6. Primary Endpoint

The registry lists the primary outcome as Mean Difference in Number of Cannabis Use Disorder Symptoms Averaged Over 2, 4, and 12 Weeks, measured over 2 weeks, 4 weeks, and 12 weeks.

The primary endpoint is the mean number of cannabis use disorder symptoms, averaged over 2, 4, and 12 weeks. The registry describes the measure as a DSM-5 Cannabis Use Disorder checklist assessed through a structured interview by a trained study staff member. Each symptom item was scored as 1 when endorsed and 0 when not endorsed. The scale ranges from 0 to 11, with higher scores indicating more symptoms.

Using a symptom count makes the outcome directly interpretable as the number of endorsed symptoms on the checklist. Averaging measurements across the stated follow-up times gives a summary across repeated assessments, rather than a result for only one visit. The record says the average was used to align with the regression model estimates. It does not report how the averaging handled a participant who missed one or more assessments.

For the primary outcome, the medical marijuana arm had a mean of 0.39 and the waitlist control arm had a mean of 0.12. The reported regression analysis gave a net mean difference of 0.28, with a 95% confidence interval from 0.15 to 0.40 and p < 0.001. The means and estimated difference describe groups, not a prediction for an individual. The scale indicates symptom count, but the registry record does not report a threshold for a clinically meaningful change or how a particular score should be interpreted for an individual participant.

General: A continuous outcome summarized by a mean can describe the average level across a group, while concealing variation between individuals. A symptom-count scale also treats each additional endorsed item as one unit in the score; it does not imply that every symptom has the same practical importance.

7. Results Posted in the Registry

• Medical Marijuana Arm: mean 0.39 versus 0.12 on control; difference 0.27 (95% CI 0.09 to 0.45), standardized effect d = 0.41; z-test p = 0.004.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Mean Difference in Physical Health Scores on the SF-12 Averaged Over 2, 4, and 12 Weeks

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Medical Marijuana Arm: mean 47.10 versus 48.20 on control; difference -1.10 (95% CI -4.42 to 2.22), standardized effect d = -0.10; z-test p = 0.516.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Medical Marijuana Arm and the control. The interval gives the range of effects that remain compatible with the data.

Mean Difference in Mental Health Scores on the SF-12 Averaged Over 2, 4, and 12 Weeks

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Medical Marijuana Arm: mean 44.90 versus 41.50 on control; difference 3.40 (95% CI -0.09 to 6.89), standardized effect d = 0.28; z-test p = 0.056.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Medical Marijuana Arm and the control. The interval gives the range of effects that remain compatible with the data.

Mean Difference in Attention Switching Task: Congruency Cost Scores on the CANTAB Averaged Over 4 and 12 Weeks

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Medical Marijuana Arm: mean 63.00 versus 77.00 on control; difference -14.00 (95% CI -28.77 to 0.77), standardized effect d = -0.30; z-test p = 0.063.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Medical Marijuana Arm and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes broaden the study's scope beyond cannabis use disorder symptoms. They include physical and mental health scores on the SF-12 and cognitive performance measures on CANTAB, including attention switching, visual information processing, paired associates learning, spatial working memory, and verbal recognition memory. The registry lists 9 secondary outcomes, but numerical results for several named measures are not reported in the registry record.

Across the numerical secondary comparisons available, estimates do not show a consistent, clearly separated pattern: some point estimates favor higher scores in the medical marijuana arm and others favor lower scores, while the confidence intervals include zero. These results are less conclusive than the primary comparison, whose reported interval is above zero. They neither confirm the primary finding across all domains nor demonstrate that the groups are equivalent on the secondary measures.

The secondary outcomes address related but distinct questions. A symptom checklist, a health-status score, and a cognitive task do not measure the same construct, so their estimates should not be combined into a single overall conclusion. Differences in their scales and assessment schedules also make direct comparisons of numerical effect sizes inappropriate.

General: when several outcomes are examined, chance alone can produce some apparently notable results. Unadjusted secondary p-values do not account for the full set of statistical tests. The registry record does not report a multiplicity adjustment or a hierarchy for interpreting the secondary measures. Consequently, these findings carry less confirmatory weight than the prespecified primary endpoint and should be read as exploratory signals with substantial uncertainty. The incomplete numerical reporting also prevents a comprehensive assessment of the secondary outcome pattern.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Mean Difference in Physical Health Scores on the SF-12 Averaged Over 2, 4, and 12 Weeks
• Mean Difference in Mental Health Scores on the SF-12 Averaged Over 2, 4, and 12 Weeks
• Mean Difference in Attention Switching Task: Congruency Cost Scores on the CANTAB Averaged Over 4 and 12 Weeks
• Mean Difference in Attention Switching Task: Switching Cost Scores on the CANTAB Averaged Over 4 and 12 Weeks
• Mean Difference in Rapid Visual Information Processing (RVP) Task: Discriminability Scores on the CANTAB Averaged Over 4 and 12 Weeks
• Mean Difference in Paired Associates Learning (PAL) Task: Total Errors Scores on the CANTAB Averaged Over 4 and 12 Weeks
• Mean Difference in Spatial Working Memory (SWM) Task: Repetition Errors Scores on the CANTAB Averaged Over 4 and 12 Weeks
• Mean Difference in Verbal Recognition Memory (VRM) Task: d' Scores on the CANTAB Averaged Over 4 and 12 Weeks

9. Statistical Interpretation

10. Statistical Methods Explained

The main reported analysis is a linear regression estimate of the net mean difference between the study groups. The registry reports a difference of 0.28, a 95% confidence interval from 0.15 to 0.40, and p < 0.001. The full regression specification is not reported in the registry record, so it is not possible to identify from the record which covariates, repeated-measure procedures, or other model details were used.

What does a mean difference express?

A mean difference is the average score in one group compared with the average score in another group, expressed in the outcome's units. For the primary endpoint, the units are symptom-count points on the checklist. The reported positive estimate indicates a higher average symptom count in the medical marijuana arm than in the waitlist control arm. The estimate does not describe the distribution of individual responses or the probability that a particular participant will have a given score.

The arm summaries give means of 0.39 and 0.12, with standard deviations of 0.78 and 0.48. A standard deviation describes the spread of observed scores around a group mean; it is not the uncertainty in the estimated mean difference. The recalculated comparison uses these summary statistics to form a simple, unadjusted estimate. It does not account for any features of the sponsor's regression analysis that are not specified in the record.

What does the confidence interval say?

A confidence interval describes the statistical uncertainty around an estimated effect under the assumptions of the analysis. The sponsor-reported 95% interval ranges from 0.15 to 0.40. Since the whole interval is above zero, the values compatible with that reported analysis indicate a positive mean difference. The interval does not mean there is a 95% probability that the true effect lies within these particular limits; it is a procedure for estimating uncertainty that would have a stated coverage property over repeated use under its assumptions.

The reported p-value, p < 0.001, measures how unusual results at least as incompatible with a no-difference null hypothesis would be under the model assumptions. It is not the probability that the null hypothesis is true, and it does not measure clinical importance. Neither a p-value nor a confidence interval addresses all sources of bias, missing data, or measurement limitations.

Why compare the recalculation with the reported analysis?

The simple recalculation from group summaries is useful as a check on direction and approximate size. It gives a mean difference of 0.27, close to the reported regression estimate of 0.28. The recalculated interval lies above zero as well. This similarity supports a consistent broad reading of the primary result, but does not establish that the analyses used identical assumptions. The sponsor's reported regression result is the study's own analysis and should lead interpretation; the recalculation is an unadjusted aid for understanding the summaries.

How should secondary measures be handled?

The registry names outcomes involving SF-12 physical and mental health scores and CANTAB cognitive tasks, among others. The available recalculated results include both positive and negative estimated differences. For the three numerical secondary comparisons described in the record, the confidence intervals include zero. General: an interval that includes zero indicates that the data and analysis do not rule out no mean difference at the stated confidence level. It does not demonstrate equivalence or prove that the groups are identical.

Unadjusted p-values for secondary outcomes do not account for the fact that multiple outcomes are examined. General: when many statistical tests are conducted, the chance of at least one apparently unusual result can increase, even if there is no genuine difference for any outcome. A multiplicity procedure can address that issue, but a procedure and adjusted results are not reported in the registry record. Secondary estimates are therefore best treated as exploratory and interpreted alongside their confidence intervals, outcome definitions, and the primary result.

What is not established by the analysis description?

The record does not report the original sample-size calculation, a detailed statistical analysis plan, interim analyses, adjustment variables, missing-data procedures, or the analysis population definition. It also does not report whether the repeated measurements were modeled jointly or how incomplete follow-up was handled. These omissions limit independent assessment of the assumptions behind the reported regression estimate. They do not erase the registered result, but they mean that the reported estimate cannot be fully reconstructed from the information in the registry record.

11. Confidence Intervals and Precision

The sponsor-reported primary estimate is 0.28 symptom-count points, with a 95% confidence interval from 0.15 to 0.40. The interval is on the same scale as the endpoint and stays above zero. Its lower and upper limits indicate a range of positive group differences compatible with the reported model, rather than a single certain value. The width of the interval conveys uncertainty: more precise estimates generally have narrower intervals, while less precise estimates have wider intervals.

The simple unadjusted recalculation gives an estimate of 0.27, and its interval also lies above zero. The recalculated and reported estimates are close, but the reported regression analysis remains the appropriate result for describing the study's own analysis. The record does not give sufficient model detail to attribute the small difference between estimates to any particular adjustment or modeling choice.

For the secondary results with numerical comparisons, uncertainty is greater relative to the estimates: the confidence intervals include zero. That means the registry results do not distinguish those group differences from no difference at the stated interval level. It is not evidence that the effects are exactly zero, and it does not establish that clinically important differences are absent.

General: confidence intervals capture sampling uncertainty under the analysis assumptions. They do not automatically account for bias from incomplete follow-up, measurement error, unmasking, or selective reporting. Precision is also distinct from clinical importance: a narrowly estimated difference could be small in practical terms, while a wider interval may leave a broad range of possible effects.

The record does not report a minimum clinically important difference for the symptom-count scale. Consequently, statistical precision can be described, but the registry alone does not establish whether the size of the estimated difference is important to participants or clinicians.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning estimates are expressed as participants per arm, not as a total across both arms and not as total events. They are illustrations based on the observed effect, not the trial's original sample-size calculation. The original design assumptions and justification are not reported in the registry record.

The listed assumptions for these planning calculations are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each planning scenario uses a specified fraction of the observed effect as the effect to detect. The figures are based on a simple planning model; they do not account for unreported design features such as attrition or other analysis choices.

The 80% power column gives the estimated number of participants per arm that the planning model would require for an 80% chance of detecting the specified effect under its assumptions. The 90% power column gives the corresponding planning estimate for a 90% chance. These are prospective design quantities, not measures of how likely the completed study was to detect its result. This is not observed power.

As the target effect becomes smaller, more participants are needed to distinguish it from random variation with the same planned power and significance level. The scenarios for 100%, 75%, and 50% of the observed effect illustrate that relationship. They should not be read as recommendations for a future study without considering the research question, expected outcomes, design, and appropriate assumptions.

The curve represents the same planning relationship across effect assumptions: detecting a smaller fraction of the observed difference requires a larger planned sample under the stated model. The values are statistical planning illustrations rather than observed trial outcomes. They cannot establish whether the completed enrollment was adequate under the study's original design, because that design calculation is not reported in the registry record.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 0.270; pooled standard deviation σ = 0.666.
2. n = 2 × (1.96 + 0.842)² × 0.666² / 0.270² = 95.6, rounded up to 96 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- 0.27                      # observed difference in means
sigma <- 0.666372398822943                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(3224468)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 96 (80% power), 129 (90% power)
 75% of observed effect: n per arm = 170 (80% power), 228 (90% power)
 50% of observed effect: n per arm = 383 (80% power), 513 (90% power)
power.t.test: 96.6 
Simulated power at n = 96 per arm: 0.799 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The sponsor's reported primary analysis is a linear regression with a net mean difference of 0.28, a 95% confidence interval from 0.15 to 0.40, and p < 0.001. This is the analysis to prioritize when describing the trial's own reported result. The unadjusted recalculation from the arm means and standard deviations estimates a difference of 0.27, with an interval that also lies above zero.

The two estimates point in the same direction and are close in size. The registry record does not flag a discrepancy between the reported and recalculated primary result. Still, the calculations are not interchangeable: the recalculation uses posted summary statistics and is unadjusted, while the reported result comes from the sponsor's regression analysis. The record does not include enough detail to reproduce that regression model or explain any difference in its estimate and interval.

For secondary outcomes with both reported sponsor analyses and recalculated comparisons, estimates are not identical. For physical health, the sponsor reports a mean difference of 0.8 with a 95% confidence interval from -1.0 to 2.5, while the unadjusted recalculation estimates -1.1. For attention switching, the sponsor reports -7.7 with a 95% confidence interval from -19.0 to 3.5, while the unadjusted recalculation estimates -14. The record does not provide the methods needed to reconcile these differences. Both reported intervals include zero. The sponsor's own analysis takes precedence as the trial's analysis; recalculations are supplemental checks rather than corrections.

15. Multiplicity, Missing Data and Analysis Populations

The registry record identifies a primary outcome and a broad set of secondary outcomes, but it does not report a statistical plan describing how multiplicity was handled. It does not state whether any secondary outcomes were designated as confirmatory, whether testing followed a hierarchy, or whether p-values or confidence intervals were adjusted for multiple comparisons. Without that information, the secondary results should not be interpreted as a collection of independent confirmations.

General: testing multiple hypotheses increases the chance that at least one result will appear statistically unusual by chance. Adjustment methods can control particular error rates, but they often change the thresholds or intervals used for interpretation. Because no method is reported, the unadjusted secondary p-values should be regarded as descriptive rather than definitive evidence.

Missing-data handling is also not reported in the registry record. The primary results include 105 participants in the medical marijuana arm and 81 in the waitlist control arm; enrollment was 269. These figures do not establish why fewer participants contributed to the primary outcome summaries. The record does not describe loss to follow-up, reasons for missing assessments, imputation, or whether analyses included all randomized participants.

Repeated outcome measurements were averaged across specified follow-up times. The record does not explain how that average was calculated for participants lacking one or more measurements, or how the regression analysis handled incomplete repeated observations. These details matter because missingness can affect estimates if the likelihood of missing data is related to outcomes or group assignment.

The analysis populations and participant flow are not reported in enough detail to determine whether the results use an intention-to-treat population, a per-protocol population, or another set of participants. The registry also does not report interim analyses or the original statistical analysis plan. These omissions limit evaluation of selective reporting and robustness; they do not by themselves show that the reported analyses are biased.

16. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: cardiac ablation (1.0%).

The registry records 1 serious adverse event in the medical marijuana arm and 0 in the waitlist control arm. It records 90 other adverse events in the medical marijuana arm and 60 in the waitlist control arm. Deaths are listed as 0 in both arms. The serious adverse event term listed among the leading terms is Cardiac Ablation, reported as 1.0%.

These are reported counts, not estimates of comparative risk or proof that an event was caused by the assigned condition. The registry record does not describe the event's clinical context, timing, relatedness assessment, or the criteria used to classify events. It also does not provide enough detail to interpret the other adverse events by type or severity.

The groups have different numbers at risk in the adverse-event record: 105 in the medical marijuana arm and 81 in the waitlist control arm. Counts alone do not account for differences in exposure time or follow-up completeness. The recorded safety figures therefore offer a descriptive account of reported events, not a complete comparative safety analysis. Absence of a recorded death or serious event in an arm should not be taken as proof that no such risk exists.

17. What the Result Does, and Does Not, Mean

The registry results indicate that the average number of cannabis use disorder symptoms during the measured period was higher in the medical marijuana arm than in the waitlist control arm. The reported regression estimate is positive, and its 95% confidence interval excludes zero. This is evidence of a group difference on the registered symptom-count outcome under the reported analysis.

The result does not establish that every participant experienced an increase, that the same difference would occur in other populations, or that a particular pattern of marijuana use caused the observed scores. The intervention details and actual use are not reported in enough detail to isolate dose, formulation, or adherence. The comparison is immediate access versus a waitlist, so it should not be recast as a comparison against a placebo or against every possible alternative.

The result also does not establish the clinical importance of a difference of this size. The registry record does not report a threshold for a meaningful change on the checklist. Nor does it show whether the difference persisted after the follow-up period, whether symptoms changed in a particular way at each visit, or how participants valued the changes.

The secondary outcomes do not provide a uniform pattern of clear differences. Their uncertainty and the absence of a reported multiplicity plan mean that they should not be used to make broad claims about physical health, mental health, or neurocognition. Missing-data procedures, analysis populations, and several other design details are also not reported in the registry record.

Accordingly, the most direct conclusion is limited to the registered primary measure and study comparison: the reported analysis found a positive mean difference in cannabis use disorder symptoms between the randomized arms over the stated assessments. That finding is informative about this study, but it is not a complete evaluation of all health effects, harms, or outcomes relevant to participants.

18. Why This Trial Matters Statistically

This trial illustrates why a registered primary outcome and a defined comparison are important for statistical interpretation. The central estimate concerns a symptom-count scale averaged across repeated assessments, and the reported confidence interval gives a range of uncertainty for the group difference. Readers can distinguish that primary result from a number of secondary questions that involve different measures and different domains.

It also illustrates the value and limits of recalculating a simple comparison from group summaries. The unadjusted primary estimate is close to the sponsor's regression estimate and has the same direction. That agreement offers a useful check on the broad reading, while the distinction between the simple recalculation and the reported model prevents false precision about how the sponsor's analysis was constructed.

The record highlights that statistical significance is not the same as clinical importance. The primary confidence interval lies above zero, but the record does not identify a threshold for a meaningful symptom-count difference. Secondary results with intervals including zero likewise do not prove no difference. Both interpretations require attention to uncertainty and to what each measure represents.

Finally, this study shows why design and reporting details matter. Enrollment, outcome-specific analysis counts, event counts, and repeated-measure summaries address different questions. The registry does not report key information about missing data, multiplicity, or the original sample-size rationale. These gaps constrain reproducibility and the strength of conclusions, even when the primary result is clearly reported.

19. 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 registry record leaves important limits on interpretation. Eligibility details beyond the listed conditions are not reported, so the characteristics of the enrolled population and the reach of the result are difficult to assess. The medical marijuana product, dose, route, frequency, and actual use are not described. The waitlist comparison therefore cannot identify the effect of a standardized exposure or separate access from other behavioral differences that may accompany assignment.

Although the study is listed as double-masked, the record does not identify who was masked or how masking was maintained. It does not report adherence, crossover, or whether participants in the waitlist arm began use before the planned wait ended. Those omissions make it difficult to evaluate the practical contrast between randomized assignment and exposure actually received.

The primary outcome is an average over repeated assessments, but the record does not report the handling of missing visits or the analysis population. The difference between enrollment and the outcome-specific counts cannot be explained from the record. Reasons for missingness, participant flow, imputation, and sensitivity analyses are not reported.

The record does not provide the full statistical analysis plan, original sample-size justification, adjustment variables, or interim-analysis information. It also does not report a multiplicity strategy for the numerous secondary outcomes. Numerical results are not available for every named secondary measure in the registry record, limiting assessment of the full neurocognitive and health-outcome pattern.

Finally, the registry does not state a clinically meaningful threshold for the primary scale. The statistical difference can therefore be described, but its practical significance cannot be settled from this record alone. The adverse-event counts also lack context about event relatedness, timing, and exposure duration.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Effect of Medical Marijuana on Neurocognition and Escalation of Use — NCT03224468.
• Registry record last updated 2023-11-15; status last verified 2023-02.
• Results first posted 2023-11-15.
• 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.

23. Record Summary

The registry describes a completed randomized, double-masked, parallel-group study sponsored by Massachusetts General Hospital, with enrollment of 269. Participants were assigned to immediate medical marijuana access or a waitlist control that agreed to wait 3 months. The record lists pain, insomnia, depression, and anxiety as conditions.

The primary outcome was the average number of cannabis use disorder symptoms at 2, 4, and 12 weeks. The sponsor-reported regression estimate was a mean difference of 0.28, with a 95% confidence interval from 0.15 to 0.40 and p < 0.001, indicating a higher average symptom count in the immediate-access arm. Available secondary comparisons are less conclusive, and several design and analysis details are not reported in the registry record.