Improving HIV and Alcohol-Related Outcomes Among HIV+ Persons in Clinic Settings: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Improving HIV and Alcohol-Related Outcomes Among HIV+ Persons in Clinic Settings (NCT02390908), a non-randomized na study. The primary outcome is Viral Load.

2. Clinical Question

3. Trial Design

This was a completed, non-randomized, parallel study with no masking. The registry lists enrollment as 174 and identifies HIV, alcohol use, and medication adherence as conditions. The study was conducted in clinical settings at multiple sites, with PLUS delivered immediately at three sites and a wait-list condition also listed. The wait-list condition was described as receiving PLUS at 12 months post baseline. These design features describe how conditions were organized, but they do not establish that participants in different conditions were comparable at the outset.

The intervention was behavioral: six sessions of Motivational Interviewing and Cognitive Behavioral Skills Training. The registry’s stated objective was to examine clinical and cost effectiveness for improving antiretroviral therapy adherence and alcohol-related outcomes. It names viral load as the primary outcome, while the reported timeframe is the most recent result within the past 90 days. The registry does not describe how sites or participants were assigned to conditions, or whether the timing of intervention delivery was chosen by participants, clinicians, or investigators. Those details are not reported in the registry record.

The study began on 2013-08-01, had a primary completion date of 2018-05-08, and is marked completed. The registry lists four outcome measures and 36 analyses, but does not report the specific analysis methods, outcome estimates, or results needed to interpret those analyses. In particular, it does not identify an adjusted primary comparison or provide estimates of uncertainty for the primary outcome. Non-random allocation is important: observed group differences, if reported, could reflect differences in participants or sites as well as differences associated with intervention timing. The record therefore supports a description of the design and stated objective, but not a causal conclusion from the primary endpoint.

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

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 Viral Load, measured over most recent result within past 90 days.

The primary endpoint is HIV viral load, expressed as log-transformed copies per milliliter. The registry says that standard-of-care HIV-1 viral load results were obtained from local laboratories and abstracted from electronic medical records. Each result represents the number of HIV copies in a milliliter, and the raw values were log-transformed to address a non-normal distribution. The reported timeframe is the participant’s most recent result within the past 90 days.

Viral load is relevant to this study because its stated aims concern HIV-related outcomes and antiretroviral therapy adherence. In general, laboratory measurements can provide a clinical outcome that is less dependent on a participant’s account of their own behavior. Log transformation is a statistical approach that can make a strongly skewed measurement more suitable for some analyses; it changes the scale on which differences are represented. The registry gives the minimum and maximum log-viral-load values in the sample as 1.28 and 6.07, respectively.

The endpoint description does not report group means, standard deviations, changes over time, or a comparison between conditions. It also does not specify how a participant without a recent laboratory result was handled. Accordingly, the endpoint’s definition is available, but its observed distribution and any treatment-related contrast are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a continuous measurement, and the registry describes it using a mean and standard deviation. In general, a mean represents the center of a set of measurements, while a standard deviation describes their spread around that center. For a comparison between conditions, group means can be used to express a difference on the measurement scale. That difference needs to be interpreted alongside its uncertainty and the analysis design. The registry record does not report the primary outcome means, standard deviations, or a comparative estimate.

The outcome uses a log-transformed scale. The registry says this transformation was used to deal with a non-normal distribution of raw viral-load values. In general, a transformation changes the scale and can reduce the influence of a long tail in a distribution, which may help an analysis meet assumptions about the residuals or outcome distribution. An estimate on a log scale is not a raw-scale difference in copies per milliliter. Interpretation must follow the scale on which the estimate is reported. The record does not give a comparative estimate or explain how any estimate was translated back to the original scale.

A confidence interval, in general, describes a range of values compatible with the observed data and the assumptions of a statistical model. Its width conveys precision: a wider interval indicates more uncertainty about the estimated quantity than a narrower interval. A confidence interval is not a guarantee that the true effect lies within its bounds. No confidence interval for the primary endpoint is reported in the registry record, so the precision of a primary comparison cannot be assessed.

Statistical significance and clinical importance are distinct ideas. A significance test, in general, assesses how compatible the data are with a specified null hypothesis under the model’s assumptions. A small p-value would not by itself establish a meaningful clinical difference, while a non-small p-value would not prove that there is no difference. Both the estimated magnitude and its uncertainty matter. The registry does not report a primary p-value or effect estimate, and no conclusion about statistical significance or clinical importance can be drawn.

Allocation was non-randomized and masking was absent. In general, non-random allocation can leave groups different in ways that affect the outcome. Differences in sites, participant characteristics, treatment access, or other factors could be associated with both condition and outcome. Statistical adjustment can sometimes account for measured differences, but it cannot automatically remove unmeasured confounding. The registry does not report how participants were assigned, the baseline balance of groups, or an adjustment model for the primary analysis. The degree to which confounding was addressed is therefore not reported in the registry record.

In a multisite study, site can matter statistically because outcomes and care processes may differ across locations. A model might account for site, or it might estimate effects separately by site, depending on the design and question. These are general possibilities, not claims about the analysis used. The registry provides site-specific descriptions of immediate PLUS delivery but does not state whether site was included in the primary statistical model, whether site-specific effects were assessed, or whether an overall estimate was calculated in a particular way.

Over-dispersion is a general concern when count outcomes vary more than a chosen statistical model expects. A model that underestimates variation can produce uncertainty estimates that are too narrow. This concept does not directly define the analysis of this continuous log-transformed endpoint, and the registry does not report an over-dispersion assessment. More generally, model checking is needed to evaluate whether the analysis assumptions fit the observed data. The record does not describe diagnostics, residual checks, or sensitivity analyses for the primary outcome.

The registry lists four outcome measures and 36 analyses, but does not identify the full set of tests, models, or contrasts. It does not report whether the primary analysis used a change from baseline, a follow-up value, repeated measurements, or a covariate-adjusted estimate. The endpoint description specifies the most recent result within the past 90 days, but it does not state how that measurement was linked to each analysis visit or how multiple measurements, if any, were handled. These details are not reported in the registry record.

The sample-size assumptions describe a planning illustration based on an observed effect, with a two-sided significance level of 0.05, equal allocation, and a normal approximation. The illustration is not the study’s own design justification and should not be treated as observed power. The registry does not report the trial’s sample-size justification or the information needed to assess its original planning assumptions. Planning calculations describe hypothetical requirements under specified assumptions; they do not show the probability that a completed study has demonstrated an effect. The registry record does not report an observed primary effect to assess in that way.

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("Site 1 Immediate PLUS Intervention", "Site 2 Immediate PLUS Intervention", "Site 3 Immediate PLUS Intervention", "Wait-list PLUS Condition", "No-Treatment Control EMR Group")
n   <- c(40, 22, 22, 90, 120)
m   <- c(3.1, 1.8, 1.8, 2.6, 2.5)
sp  <- c(1.7, 1, 0.6, 1.4, 1.4)
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  Site 1 Immediate PLUS Intervention  40  3.100  1.700  2.556  3.644
CARM  Site 2 Immediate PLUS Intervention  22  1.800  1.000  1.357  2.243
CARM  Site 3 Immediate PLUS Intervention  22  1.800  0.600  1.534  2.066
CARM  Wait-list PLUS Condition  90  2.600  1.400  2.307  2.893
CARM  No-Treatment Control EMR Group  120  2.500  1.400  2.247  2.753

11. Multiplicity, Missing Data and Analysis Populations

The registry lists four outcome measures and 36 analyses. When multiple outcomes, time points, or contrasts are examined, the chance of obtaining at least one apparently notable result can rise. In general, multiplicity procedures can adjust statistical testing or emphasize a prespecified primary comparison. The registry does not report whether any such procedure was planned or used, nor does it identify which of the listed analyses were primary, secondary, or exploratory. The multiplicity approach is not reported in the registry record.

The primary outcome is described as the most recent viral-load result within the past 90 days, but the record does not explain how missing laboratory measurements were handled. It does not state whether participants without an available result were excluded, whether another measurement was used, or whether any imputation method was applied. Missingness can affect interpretation when the likelihood of an observed measurement is related to participant characteristics or outcomes. Without details on the amount, pattern, and handling of missing data, the potential impact cannot be assessed.

The registry does not define the analysis population for the primary endpoint. It does not state whether analysis included everyone enrolled, only participants who received the intervention, or another group. It also does not describe adherence to sessions or how participants who did not complete follow-up were treated analytically. These distinctions can matter, particularly in a non-randomized design where the basis for inclusion and comparison may influence results. The analysis population, missing-data strategy, and multiplicity plan are not reported in the registry record.

12. 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.

The safety listing reports zero serious adverse events in each of the listed groups. It gives the numbers at risk for serious events as 40 for Site 1 Immediate PLUS Intervention, 22 for Site 2 Immediate PLUS Intervention, 22 for Site 3 Immediate PLUS Intervention, 90 for Wait-list PLUS Condition, and 120 for the No-Treatment EMR Control Group. These are group-specific registry counts; the record does not explain how the safety populations were defined or how follow-up time differed across groups.

Deaths are listed separately: zero for Site 1 Immediate PLUS Intervention, zero for Site 2 Immediate PLUS Intervention, one for Site 3 Immediate PLUS Intervention, three for Wait-list PLUS Condition, and zero for the No-Treatment EMR Control Group. The registry does not attribute these deaths to an intervention, and the counts alone do not establish causality. It also does not provide details about the deaths or their timing. No serious adverse event terms are listed. The reported counts should be read as the registry’s safety summary, not as a comparative risk analysis.

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

The record establishes that a completed, non-randomized study evaluated a behavioral intervention called PLUS in real-world clinical settings. Its stated objective concerned antiretroviral therapy adherence and alcohol-related outcomes among HIV-positive individuals who drink at harmful or hazardous levels. The primary outcome named in the registry was log-transformed viral load, measured using a recent standard-of-care result abstracted from electronic medical records. These details describe the research question and measurement, not the study’s findings.

The registry does not report the primary outcome results. It therefore does not show whether viral load differed between the immediate-intervention and wait-list conditions, or whether any difference was precise or compatible with no difference. It also does not report comparative results for adherence or alcohol-related outcomes. A completed status does not mean that the registry contains a complete results record.

Even if a difference were reported, non-random allocation would require careful interpretation. Differences among participants or sites could contribute to an observed contrast, and the registry does not provide the baseline information or analysis details needed to assess that possibility. The absence of masking is also part of the recorded design, but its implications depend on outcome measurement and other study procedures that are not fully described.

The safety counts are descriptive. They list no serious adverse events in the reported groups and give separate death counts, but do not establish whether any death was related to an intervention. They cannot replace a comparative analysis of benefits or harms.

In short, the record supports a description of the study design, intervention, and endpoint, but it does not support a conclusion about whether PLUS changed viral load, adherence, or alcohol-related outcomes. The size and direction of any effect, and the uncertainty around it, are not reported in the registry record.

14. Why This Trial Matters Statistically

This study raises a statistical question that is common in clinical research: how to interpret outcome comparisons when an intervention is implemented in real-world settings without random assignment. A multisite setting can make evidence more reflective of varied clinical contexts, but it can also make site differences important to the analysis. The registry records immediate PLUS delivery at three sites and a wait-list condition, while leaving assignment procedures and the primary modeling strategy undescribed.

The endpoint also illustrates why the measurement scale matters. Viral-load values were log-transformed because the raw values were described as non-normal. Statistical estimates on a transformed scale need to be interpreted in terms of that scale, and a clear report should state what contrast was estimated and how it relates to the underlying measurement. The registry defines the endpoint and gives its observed range, but does not provide the comparative estimates needed for that interpretation.

Finally, a complete statistical account requires more than a study’s enrollment and completion status. Readers need outcome summaries, an estimate of the comparison, measures of uncertainty, and information about missing data and analysis populations. Those items are not reported in the registry record. This makes the study’s question and design identifiable, while leaving its primary statistical findings unavailable for evaluation.

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 trial-specific limitations of interpretation follow from the registry entries. The allocation is explicitly non-randomized, and the record does not describe how participants or sites entered the different conditions. Consequently, the registry does not establish that the groups were comparable before intervention. It also does not report baseline viral-load summaries or other baseline characteristics that would help readers assess differences among conditions.

The intervention was delivered immediately at three sites, while the wait-list condition was described as receiving it at 12 months post baseline. The record does not explain whether outcome timing was aligned across these conditions, how the wait-list period was incorporated into analyses, or how site differences were addressed. The No-Treatment EMR Control Group appears in the safety listing, but its role in the primary outcome analysis is not reported.

The primary endpoint description specifies the most recent result within the past 90 days and notes that local laboratories posted the standard-of-care results. The record does not report the availability of measurements by group, whether laboratory methods differed across sites, or how missing results were handled. It also gives no outcome estimates, statistical uncertainty, model description, multiplicity procedure, or analysis-population definition. These omissions prevent assessment of the magnitude, precision, and robustness of any primary comparison. The registry does not report a sample-size justification for the trial; the stated planning illustration is not that justification.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Improving HIV and Alcohol-Related Outcomes Among HIV+ Persons in Clinic Settings — NCT02390908.
• Registry record last updated 2021-04-12; status last verified 2021-04.
• Results first posted 2021-04-12.
• 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

The registry describes a completed, non-randomized, parallel study of PLUS, a behavioral intervention consisting of six sessions of Motivational Interviewing and Cognitive Behavioral Skills Training. Immediate delivery was listed at three sites, alongside a wait-list condition described as receiving PLUS at 12 months post baseline. The study’s stated objective concerned antiretroviral therapy adherence and alcohol-related outcomes among HIV-positive individuals who drink at harmful or hazardous levels.

Enrollment is listed as 174. The primary outcome was log-transformed HIV viral load, based on the most recent result within the past 90 days. The registry reports a sample range of 1.28 to 6.07 for log-viral-load values, but does not report comparative primary outcome results. It lists four outcome measures and 36 analyses without describing the analysis plan in detail. Serious adverse event counts are reported as zero in the listed groups; death counts are reported separately. The record does not establish causality for deaths or report a primary treatment-effect estimate.