Risperidone and Desipramine in Alcohol Use and Schizophrenia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Risperidone and Desipramine in Alcohol Use and Schizophrenia (NCT01411085), a na phase 2 study. The primary outcome is Timeline Followback Assessing Number of Drinks Per Week.

2. Clinical Question

3. Trial Design

This was a Phase 2, single-group study sponsored by Dartmouth-Hitchcock Medical Center. The registry lists an enrollment of 12, a start of 2011-12, a primary completion of 2014-08, and a completed status. Allocation is listed as NA and masking as none. The final design was open label and pre-post: participants receiving risperidone or a risperidone-like agent with desipramine were assessed during treatment and compared with their own pre-baseline period.

The registry’s summary explains that the design changed in June 2013. The original plan was a randomized controlled study comparing risperidone plus desipramine, risperidone, and placebo. Following the change, the aims were revised to assess alcohol use, symptoms of schizophrenia and depression, and side-effect burden during combination treatment compared with the period before baseline. These original and revised designs address different questions. The final record should therefore be read as a single-group pre-post study, not as a randomized treatment comparison.

The intervention description allows risperidone or a risperidone-like agent, including risperidone long-acting, paliperdione, and paliperidone palmitate. It also states that the target dose of oral risperidone was 4mg, while allowing variations. The description does not establish which formulation or dose each participant received. Nor does the record report protocol details such as the reasons for the design change or the operational rules for switching or adjusting treatment.

The primary analysis included all completers, N=7. The record does not explain how completion was defined, why the analysis population was smaller than planned enrollment, or how outcomes from people who did not complete were handled. The registry lists one outcome measure and one analysis, but does not report numerical primary outcome results or the statistical analysis used. The design supports a within-person description of change over time, but without a concurrent control it cannot separate treatment-period change from other changes that may occur over time.

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.

Before assignment: Two participants were enrolled in a risperidone + placebo arm, but in 6/2013 the design of the study was changed to include only the risperidone + desipramine arm.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The analysis was performed on all completers (N=7).

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.

Population: 7 completers in the risperidone + desipramine group were analyzed. 3 study subjects in the risperidone + desipramine group did not have sufficient days in study for analysis.

6. Primary Endpoint

The registry lists the primary outcome as Timeline Followback Assessing Number of Drinks Per Week, measured over Weekly for 14 weeks, using data from last 8 weeks.

The listed primary endpoint is number of drinks per week, expressed as a mean with standard deviation. It was assessed weekly for 14 weeks, using data from the last 8 weeks. The measurement method was the Timeline Followback (TLFB), a retrospective self-report assessment of alcohol and other substance use. The registry says the TLFB was enhanced by procedures intended to strengthen reliability and validity, but does not describe those procedures.

Weekly drinking quantity can describe alcohol use over time and is distinct from the number of days on which a person drinks. The revised study aims specifically mention fewer drinking days and fewer heavy drinking days during the final 8 weeks compared with pre-baseline. The listed primary measure, however, is drinks per week. The registry does not report numerical values for this measure or clarify how it was related analytically to the drinking-day aims.

In general, repeated weekly assessments can characterize patterns during treatment rather than relying on a single retrospective summary. A mean describes the average level in the analyzed group, while a standard deviation describes variation among observations or participants, depending on the analysis definition. The record does not specify the unit over which the standard deviation was calculated, nor does it report the endpoint’s observed mean or standard deviation. Consequently, the endpoint’s numerical change and its variability cannot be interpreted from the registry results.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes a continuous primary outcome summarized as a mean with standard deviation. It identifies the Timeline Followback method and a pre-baseline versus treatment-period question, but does not report the statistical procedure used to compare periods. The explanations below describe general statistical concepts; they are not a claim that a particular procedure was used in this study.

What does a pre-post comparison estimate?

In a single-group pre-post design, the central quantity is the difference in outcome between a participant’s measurement before treatment and measurement during treatment, or a group summary of those changes. Pairing measurements can account for the fact that observations from the same person are related. A paired analysis focuses on within-person change rather than treating the periods as unrelated groups. The registry does not state whether the analysis used paired differences, a model for repeated measurements, or another method.

The interpretation of a pre-post contrast is narrower than that of a randomized comparison. It describes a temporal change in the observed group, subject to the definitions and availability of the measurements. In the absence of a concurrent control, the contrast cannot distinguish a treatment-related effect from changes that would have occurred over time for other reasons. This limitation remains even if a statistical test finds a difference. Statistical association between treatment period and outcome is not, by itself, proof of causation.

What do a mean and standard deviation describe?

For a continuous measure, the mean is a summary of its average level. The standard deviation describes spread around that average. In general, a larger standard deviation indicates more variation in the observations, but interpretation depends on how observations are defined and summarized. With repeated weekly measures, for example, variability might refer to variation across participants, weeks, or participant-level summaries. The registry does not specify which definition was used for the listed standard deviation, and gives no numerical mean or standard deviation for the primary outcome.

Alcohol-use measures may be variable and may not follow a symmetric distribution. In general, a mean and standard deviation can still be useful summaries, but they may not fully characterize a skewed distribution or observations with unusually high values. A model or robust summary may be considered when the data structure and distribution warrant it. The registry record does not report an assessment of the outcome distribution, a diagnostic procedure, or an alternative summary.

Why does repeated measurement matter?

Weekly assessments can retain information about variation over the treatment period. Measurements from one participant are generally related to one another, so a method for repeated observations may need to account for that correlation. Ignoring the relationship can lead to misleading uncertainty estimates. In general, repeated-measures models can describe patterns over time while accounting for within-person dependence, but the chosen method must match the available observations and assumptions. The registry provides the weekly assessment schedule and says that data from the last 8 weeks were used, but does not say how those observations were combined or modeled.

Retrospective self-report has its own measurement considerations. The TLFB asks participants to estimate past alcohol and other substance use. Recall and reporting can vary, and methods to improve reliability do not eliminate all measurement error. The registry notes that procedures were used to strengthen reliability and validity, without describing them. No assessment of reporting accuracy or measurement error is reported in the registry record.

What does a confidence interval add?

In general, a confidence interval communicates uncertainty around an estimated effect. It helps readers distinguish an estimate that is relatively precise from one that is compatible with a broad range of values. Its width depends on features including the amount and variability of data and the analysis method. A confidence interval does not correct design limitations: an interval around an uncontrolled pre-post estimate would not remove possible time-related confounding. No interval or other precision measure for the primary outcome is reported.

A p-value, when reported, addresses how compatible the observed data are with a specified null model under the analysis assumptions. It does not measure the probability that a hypothesis is true, the size of an effect, or whether a change is important. Because the registry does not report a p-value or the test used, no significance conclusion can be drawn from its primary outcome entry.

How should the planning illustration be read?

The record describes a planning illustration based on the observed effect, using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It defines the effect to detect as a stated fraction of the observed effect. This is a hypothetical planning exercise, not a description of the final single-group pre-post design. The illustration’s assumptions do not establish that the trial itself was designed or powered that way, and the registry does not report a sample-size justification for the final design.

In general, sample-size calculations estimate how many observations a study might need under chosen assumptions about the effect, variability, significance threshold, and desired power. Such calculations are conditional on those assumptions; they do not demonstrate the actual power of a completed study. Any sample-size values based on the observed effect are planning illustrations rather than observed power. The record’s stated two-arm allocation assumption should not be confused with the final design, which was single group. The registry does not report an interim analysis plan, a detailed statistical analysis plan, or the method used for the primary outcome comparison.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Risperidone + Desipramine")
n   <- c(7)
m   <- c(35.9)
sp  <- c(20.4)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Risperidone + Desipramine  7  35.900  20.400  17.033  54.767

11. Multiplicity, Missing Data and Analysis Populations

The registry lists one outcome measure and one analysis, but the revised aims describe several domains: alcohol use, schizophrenia and depression symptoms, and side-effect burden. The primary measure listed is drinks per week. The aims also refer to drinking days and heavy drinking days. These outcomes are not interchangeable, and the record does not report numerical results for them or explain their status in a hierarchy of primary and secondary outcomes.

Multiplicity arises in general when researchers examine more than one outcome, time point, subgroup, or statistical contrast. The more analyses considered, the greater the opportunity for apparently notable results to arise by chance if each is interpreted independently. Statistical plans can address this by identifying a primary analysis in advance or by adjusting interpretation for multiple testing. The registry does not report a multiplicity adjustment, a testing hierarchy, or enough analysis details to determine how the distinct aims were treated. This absence does not show that an adjustment was or was not used outside the registry record.

Missing data are also important in a longitudinal study. The primary outcome population is described as all completers, N=7, whereas enrollment was 12. The record does not report why participants did not complete, whether all weekly measurements were available among completers, or how missing weekly reports were handled. It does not describe imputation, a missing-data model, or sensitivity analyses. The impact of missing outcomes and the assumptions behind any handling method are therefore not assessable.

Analysis populations should be interpreted according to their definition. The registry identifies completers for the primary outcome but does not state whether an intention-to-treat population or another population was analyzed. It also does not explain the inclusion rules for each assessment. Without participant-level flow and analysis definitions, readers cannot determine how exclusions or incomplete follow-up may have influenced the reported analysis—nor can they evaluate that influence when the numerical outcome result itself is not reported.

12. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: suicidal ideation (10.0%), suicide atttempt (10.0%), alcohol intoxication (10.0%).

The registry lists 10 participants at risk for serious adverse events in the risperidone plus desipramine group, with 3 participants affected. It also lists 10 participants at risk for other adverse events and 10 affected. The record identifies suicidal ideation, suicide attempt, and alcohol intoxication among the serious-event terms, each reported as 10.0%. These are safety counts and percentages in the registry; they do not establish that the medication combination caused an event.

The record does not explain whether participants could experience more than one event, how events were grouped, or how the listed percentages were calculated. The serious-event terms should not be added or treated as mutually exclusive categories without that information. The registry lists deaths as null; a death count is therefore not reported in the registry record, rather than establishing that no deaths occurred.

In general, adverse-event counts describe events observed and reported under a study’s collection procedures. They do not by themselves establish a causal relationship to treatment. Interpretation also depends on the number of participants observed, duration of observation, event definitions, and completeness of reporting. The registry record does not provide enough detail to assess those features beyond the risk counts and terms listed. The safety figures should be read as descriptive reports from this study group, not as a controlled comparison or a risk estimate applicable to other populations.

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

The record establishes that the final study design was an open-label, single-group pre-post assessment of risperidone or a risperidone-like agent with desipramine. Its revised aims concerned alcohol use, symptoms, and side-effect burden. The primary outcome entry specifies drinks per week assessed with the TLFB, using data from the last 8 weeks, and identifies an analysis of all completers, N=7.

The numerical primary outcome result is not reported in the registry record. As a result, the posted information does not show whether drinks per week changed during treatment, how large any change was, or how precisely it was estimated. The revised hypothesis about fewer drinking days and fewer heavy drinking days is a research question, not a reported finding. Nor are symptom outcomes or numerical findings for side-effect burden included in the record details.

Even if a pre-post change were available, the design would describe change in participants over time rather than provide a randomized estimate against a concurrent control group. In general, such a change can coincide with treatment without being attributable to treatment. Other changes over time, selection of completers, measurement variation, and regression to the mean are possible considerations for an uncontrolled comparison. The registry does not report analyses that would resolve these concerns.

The listed adverse-event counts should likewise be interpreted descriptively. The record reports serious and other adverse events among the treated group, but does not provide a control comparison or establish event causation. It also does not report a death count in interpretable form. These data do not support conclusions about comparative safety.

In short, the registry describes the study’s revised question, design, outcome measure, analysis population, and certain adverse-event counts. It does not provide the primary outcome estimate or its uncertainty. Conclusions about the direction, magnitude, precision, or causal meaning of alcohol-use changes are therefore not supported by the posted results. This is a statement about what the registry record reports, not a claim about results that may exist elsewhere.

14. Why This Trial Matters Statistically

This study illustrates why the distinction between a planned randomized comparison and a final single-group pre-post design matters. The registry says the design changed in June 2013, from a randomized controlled study to an open-label pre-post study. Those designs answer different statistical questions. A randomized comparison can estimate differences between assigned groups under its design assumptions; a pre-post study describes change within the observed group. The record’s final allocation and masking fields are consistent with reading the completed study as an uncontrolled, open-label assessment.

The change also affects how a reader should interpret the analysis population. The primary outcome analysis was conducted on completers, N=7, while the planned enrollment was 12. Without participant flow, reasons for non-completion, and missing-data methods, it is difficult to assess how representative the completers are of those enrolled. This is a general concern in studies where outcome analysis depends on completion, not evidence that any particular bias occurred.

The endpoint description also illustrates the value of distinguishing related measures. Drinks per week is not the same quantity as drinking days or heavy drinking days. The revised aims discuss these measures, while the listed primary outcome names weekly drink quantity. Clear reporting of which outcome was analyzed, how repeated assessments were summarized, and what estimate and uncertainty resulted is necessary for statistical interpretation.

Finally, a planning calculation is not evidence of observed power. The record describes a hypothetical two-arm planning illustration, even though the final study was single group. Treating such an illustration as the design’s achieved power would conflate assumptions with study results. The registry’s missing numerical primary outcome results make this distinction especially important: a study’s intended question, a planning exercise, and an observed estimate are different forms of information.

15. 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.

Several limitations specific to this record constrain interpretation. First, it describes a design change from randomized controlled comparisons to an open-label pre-post study. The final study does not include a concurrent control group, so changes during treatment cannot be separated from other changes over time using the design described. The registry does not explain the reasons for the change or report whether any analyses from the original design were conducted.

Second, the primary outcome analysis included all completers, N=7, while enrollment was 12. The record does not report reasons for non-completion, participant flow, or how missing observations were handled. It is therefore not possible to assess whether the analyzed participants differed from those not included in the primary analysis, or whether incomplete weekly reports affected the estimate.

Third, the outcome descriptions do not align perfectly in wording. The revised aims emphasize drinking days and heavy drinking days, while the listed primary outcome is number of drinks per week. The record does not report numerical results for the listed measure or clarify the analytical relationship among these measures. It also does not provide results for the symptom and side-effect aims.

Fourth, the intervention description permits risperidone-like agents and dose variations. The record does not identify the exact medication formulation and dose received by each participant. The open-label design also means masking was none. These features limit the detail with which treatment exposure and assessment conditions can be characterized from the registry.

Finally, no primary outcome estimate, confidence interval, p-value, or analysis method is reported. The statistical plan, interim analyses, sample-size justification for the final design, and missing-data approach are not reported in the registry record. These are reporting limitations; they do not establish that such procedures were absent from study conduct or from documents outside the record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Risperidone and Desipramine in Alcohol Use and Schizophrenia — NCT01411085.
• Registry record last updated 2018-03-23; status last verified 2018-03.
• Results first posted 2018-03-23.
• 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.

19. Record Summary

NCT01411085, “Risperidone and Desipramine in Alcohol Use and Schizophrenia,” is a completed Phase 2, single-group study sponsored by Dartmouth-Hitchcock Medical Center. The registry describes an open-label pre-post design following a change in June 2013 from an originally planned randomized comparison. Participants received risperidone or a risperidone-like agent with desipramine.

Planned enrollment was 12. The primary outcome was number of drinks per week, assessed with the Timeline Followback method; the analysis population was all completers, N=7. The registry does not report the numerical primary outcome result or its uncertainty. It lists serious and other adverse-event counts for the treatment group, including serious-event terms, but does not establish causation or report an interpretable death count. The available record supports a description of the design and measurement plan, but not a conclusion about the direction, size, or precision of alcohol-use change.