Boston ARCH 4F Intervention to Reduce Fall Risk in People With HIV and Alcohol Use: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Boston ARCH 4F Intervention to Reduce Fall Risk in People With HIV and Alcohol Use (NCT04804579), a randomized na study. The primary outcome is Engagement With Intervention.

2. Clinical Question

3. Trial Design

This is a completed, randomized, parallel-group pilot study sponsored by Boston University. The registry record lists no masking and records enrollment of 23. The study dates run from 2021-10-20 to primary completion on 2022-08-01. Its stated aim is to assess the acceptability and feasibility of a randomized trial of a 10-week virtual intervention to reduce fall risk among people with HIV who consume alcohol. The stated hypothesis concerns feasibility and acceptability at the pilot stage, rather than a definitive test of clinical effectiveness.

The intervention combines customized at-home exercises, weekly virtual group sessions, and weekly phone check-ins. The group sessions are described as approximately 30 minutes per week for 10 weeks. The registry’s study summary says participants are asked to complete assigned exercises 3 times per week. The check-ins are intended to answer questions, give individual feedback on exercises, and arrange reminders tailored to participants’ needs. The control group receives an educational pamphlet about falls and alcohol use.

Standardized assessments are described as taking place in person at Boston University Medical Campus. They cover fall risk, fear of falling, physical performance measures such as grip strength, balance, and gait speed, substance use, and related measures. The registry lists self-reported falls, physical function on the Short Physical Performance Battery, Fried’s Frailty Phenotype, and days of alcohol and other drug use among the outcomes.

Although enrollment was 23, outcome-specific analysis populations vary. The primary engagement result includes 11 intervention participants who completed the 10-week assessment. The physical-function and substance-use summaries include 11 participants per group, while the frailty summaries include 11 intervention participants and 10 control participants. These endpoint-specific counts should be read as reported for each measure, not as interchangeable denominators. The registry record does not report the allocation sequence, a sample-size justification for the trial, or a detailed statistical analysis plan.

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

Analysis population for the primary outcome, as defined in the registry: Analysis population includes participants assigned to the Fall Prevention Intervention who completed the 10-week follow-up research assessment.

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 Engagement With Intervention, measured over 10 weeks (final assessment).

The primary endpoint is engagement with the intervention, measured as the mean number of weekly virtual group sessions attended out of 10 sessions at the 10-week final assessment. Attendance is a direct measure of participation in a central component of the program. In a pilot study whose stated aim is to assess feasibility and acceptability, it can describe whether participants attended scheduled group sessions during the intervention period.

The registry reports a mean of 3.27 sessions and a standard deviation of 3.44 among 11 intervention participants who completed the 10-week follow-up research assessment. The standard deviation describes variation in attendance across the participants included in that summary; it is not a measure of uncertainty around the mean. The registry does not report a control-group attendance result or a confidence interval for the intervention mean.

This endpoint has a specific scope. It captures attendance at virtual group sessions, not completion of at-home exercises, phone check-ins, overall adherence to every intervention component, or changes in fall risk. It also applies to the stated analysis population of intervention-assigned participants who completed follow-up. The registry record does not report attendance for intervention participants who did not complete that assessment or explain how their missing attendance information was handled. Accordingly, this result describes session attendance among the assessed participants; it does not establish acceptability for all enrolled participants or quantify an intervention effect relative to the pamphlet group.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is a continuous measure summarized by a mean and standard deviation. The mean gives the average number of virtual group sessions attended in the endpoint’s analysis population. The standard deviation describes how spread out individual attendance counts were around that mean. A larger standard deviation indicates greater variability among the observed participants; it does not indicate how accurately the mean estimates attendance in a broader population. The registry does not post a confidence interval for the primary mean.

What does a mean difference represent?

For the posted secondary outcomes, the registry reports a mean for each group and a mean difference comparing the intervention group with the control group. The difference is expressed in the outcome’s own units: score on a scale for physical function and frailty, and days for the substance-use measure. The direction of a difference depends on how the outcome is defined and coded. A positive or negative estimate should therefore be interpreted in the context of that measure, not as an automatic indication of benefit or harm.

The posted differences are unadjusted summaries based on the group means. The registry does not report a model-adjusted estimate for these comparisons. An unadjusted estimate does not account for possible baseline imbalances or other covariates through a statistical model. The record also does not report a prespecified adjustment strategy, so no adjusted result can be inferred from the posted summaries.

How should confidence intervals be read?

A confidence interval describes the statistical uncertainty around an estimated group difference under the assumptions of the method used. A narrow interval indicates greater precision than a wide interval, while an interval spanning zero indicates that the data are compatible with no group difference as well as differences in either direction. This does not mean that every value in the interval is equally plausible, or that the interval gives the probability that the true value lies within it. The registry’s intervals for the posted secondary comparisons span zero, and the width of the intervals shows that the estimates are imprecise.

A confidence interval is not a substitute for considering the study’s design, outcome definitions, analysis population, and missing data. In a small pilot, estimates can vary substantially from sample to sample. The interval for an estimate reflects sampling uncertainty under the analysis, but it does not address every potential source of bias or uncertainty, such as incomplete follow-up or limitations in measurement.

What does an unadjusted p-value say?

A p-value measures how incompatible the observed data are with a specified null hypothesis under the model and assumptions used for the test. It is not the probability that the null hypothesis is true, and it does not measure the size or importance of a difference. The posted p-values for secondary comparisons are unadjusted. Their interpretation should account for the exploratory setting and the presence of multiple outcomes; a p-value alone does not establish a finding.

The registry record does not report the exact testing procedure for these comparisons, whether assumptions were checked, or whether any analyses were prespecified. As a general statistical consideration, means and standard deviations can be affected by skewed distributions or influential observations. With limited data, it can be difficult to assess distributional assumptions or estimate variation precisely. The record does not provide participant-level data or distributional summaries that would allow those features to be evaluated.

What are the sample-size illustrations?

The planning illustrations are based on the observed effects, rather than on the trial’s own design justification. The record states that they assume a two-sided significance level of 0.05, equal allocation between groups, and a normal approximation; the target effect is a stated fraction of the observed effect. The resulting values are hypothetical participants per arm for specified power targets. They are not observed enrollment, not a report of the trial’s planned sample size, and not observed power. Because the calculations use effects estimated in a small pilot, they should be treated as planning illustrations rather than precise requirements for a future study. The registry record does not report a sample-size justification used to design this pilot.

For mean comparisons, general statistical methods often rely on assumptions about the distribution of the outcome or the sampling distribution of the mean. Over-dispersion, meaning more variability than a model assumes, can make standard errors too small if it is not accounted for. This is a general consideration, not a claim that over-dispersion occurred in this trial. The registry does not report a diagnostic assessment or a variance model for the posted summaries. The appropriate interpretation is therefore limited to the calculations and summaries recorded, without assuming unreported modeling choices.

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("Fall Prevention Intervention")
n   <- c(11)
m   <- c(3.27)
sp  <- c(3.44)
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  Fall Prevention Intervention  11  3.270  3.440  0.959  5.581

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists several outcomes, including intervention engagement and measures of falls, physical function, frailty, and substance use. It does not report whether a multiplicity adjustment was planned or applied, nor does it describe a hierarchy among the secondary outcomes. The posted secondary p-values are unadjusted. When many outcomes are examined, the chance of at least one apparently unusual result can rise; this general concern makes unadjusted secondary findings exploratory rather than definitive. No adjusted significance claims can be inferred from the record.

The primary endpoint’s analysis population is explicitly described as intervention-assigned participants who completed the 10-week follow-up research assessment. The registry reports a primary mean for 11 participants. It does not report how many intervention-assigned participants lacked that assessment, how missing attendance values were treated, or whether any imputation or sensitivity analysis was used. The control group has no posted primary attendance value, so a comparative analysis is not estimable from the posted primary endpoint results.

Secondary endpoint counts differ by measure. Physical function and days of alcohol use and other drug use have posted group summaries for 11 participants in each group. The frailty summary has 11 intervention participants and 10 control participants. These differences show that the analysis populations are not identical across outcomes, but the registry does not explain the reasons for the differing counts or the handling of missing measurements. It also does not report whether analyses followed intention-to-treat principles or another population definition for the secondary endpoints.

These are important limits because missingness can affect both the precision and interpretation of an estimate, particularly if the availability of measurements is related to participants’ outcomes or engagement. That is a general statistical concern; the record does not establish that missingness was related to any outcome in this study. Without a reported missing-data strategy, the posted summaries should be understood as summaries of the participants included in each endpoint analysis, rather than as complete follow-up results for everyone enrolled.

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 0% reporting threshold.

The safety record reports zero serious adverse events and zero other adverse events affecting participants in the intervention group, with 12 participants at risk for each category. For the control group, it reports zero serious adverse events and zero other adverse events affecting participants, with 11 at risk for each category. Deaths are reported as 0 in both groups. No serious adverse event terms are listed.

These figures describe the events recorded in the registry for the stated at-risk groups. They do not establish that no event could have occurred outside the reporting period or that the intervention has no safety risks. The registry record does not provide event-level descriptions, a definition of the reporting window, or additional detail about how adverse events were collected and assessed. The at-risk counts are safety denominators and should not be assumed to match the analysis population for a particular efficacy or engagement outcome. In a pilot study, the absence of reported events is a descriptive observation, not a precise estimate of the frequency of uncommon harms.

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

The clearest primary result is that the intervention participants included in the follow-up analysis attended a mean of 3.27 virtual group sessions, with substantial variation in attendance as reflected by the standard deviation of 3.44. This provides a descriptive measure of engagement with the group-session component among the assessed participants. It does not summarize every component of the intervention, and it cannot be compared with control attendance because no control result is posted for this endpoint.

The secondary results offer preliminary estimates for changes in physical function, frailty, and substance-use days. Their confidence intervals span zero, and the estimates are unadjusted. They therefore do not precisely determine whether there was a group difference or how large such a difference might be. These findings are compatible with a range of possibilities and should not be treated as definitive evidence for or against an effect. The registry’s stated focus on feasibility and acceptability is consistent with interpreting this as a pilot record, not a confirmatory efficacy evaluation.

The record also does not provide a numerical result for change in self-reported falls, despite listing that measure among the outcomes. It does not report detailed missing-data handling, an adjusted secondary analysis, or a multiplicity plan. Those omissions restrict how confidently the endpoint summaries can be generalized or combined into an overall conclusion.

In particular, session attendance is not itself a measure of fall reduction. A participant may attend sessions without a corresponding change in falls, and the attendance mean cannot be used to infer an outcome that was not posted. Likewise, the secondary point estimates do not establish that the intervention works or does not work. They describe the available data and their uncertainty in this pilot. The safety counts are also limited to reported events and do not establish absence of risk.

The most defensible interpretation is therefore narrow: the registry records attendance among intervention participants with completed follow-up, exploratory group summaries for several secondary measures, and no reported adverse events in the stated at-risk groups. A larger, fully reported study with defined analysis populations, a prespecified analysis strategy, and complete outcome reporting would be needed to provide more precise comparative estimates.

14. Why This Trial Matters Statistically

This study illustrates why a pilot’s primary endpoint needs to be read in light of its stated purpose. The primary measure is session attendance, which is directly relevant to engagement and feasibility. It is not a surrogate calculation of a clinical effect on falls. Keeping those questions separate prevents a descriptive participation result from being overinterpreted as evidence about health outcomes.

The record also illustrates the importance of endpoint-specific denominators. The primary attendance mean applies to intervention participants who completed the follow-up assessment, while the posted secondary summaries have their own group counts. Differences in those counts affect the populations represented by each estimate and make it important not to treat all outcome summaries as if they came from the same complete set of participants.

Statistically, the secondary confidence intervals communicate more than point estimates alone. Intervals that span zero show that the data do not sharply resolve the direction of a group difference, while their breadth conveys limited precision. The unadjusted p-values should be interpreted alongside that uncertainty, the exploratory setting, and the presence of multiple outcomes. A small pilot can generate useful estimates for planning, but its observed effects can be unstable. Translating them into hypothetical sample-size illustrations does not make the original estimates precise or convert planning scenarios into evidence of observed power.

Finally, incomplete reporting is itself relevant to interpretation. The absence of a posted control attendance result prevents a randomized comparison on the primary endpoint. Missing-data handling, a multiplicity approach, and a detailed analysis plan are not reported. Recognizing these limits is not a judgment about the study’s value; it distinguishes what the registry can support from questions that remain unresolved. That distinction is central to sound statistical reading of pilot trials.

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 registry record constrain interpretation. The primary endpoint is reported only for the intervention group, despite the randomized control group. Without a control-group attendance summary, the endpoint cannot support a between-group comparison. The result also represents only intervention-assigned participants who completed the 10-week follow-up assessment, and the record does not explain attendance among those not included in that analysis population.

The study is small, and the endpoint-specific counts are limited. The physical-function and substance-use summaries include 11 participants per group, while the frailty summaries include 11 intervention participants and 10 control participants. The reasons for the different counts are not reported in the registry record. The handling of missing observations and the analysis population definitions for the secondary outcomes are also not reported.

Secondary comparisons are unadjusted, and a multiplicity strategy is not reported. The record does not describe the statistical tests, distributional checks, covariate adjustment, or prespecified analysis hierarchy. These omissions make it difficult to assess whether the posted summaries fully reflect a planned analysis or how model assumptions were evaluated. The reported intervals and p-values should therefore be interpreted as the registry’s posted comparisons, not as a complete account of every possible analysis.

Outcome reporting is incomplete in the available record: change in self-reported falls is listed, but a numerical result is not posted among the endpoint summaries. The primary engagement measure captures attendance at virtual groups rather than adherence to all intervention components. The record does not provide a numerical assessment of home-exercise completion or phone-check-in participation. Finally, the safety figures are event counts without detailed event descriptions or reporting-window information. Together, these limitations favor cautious, descriptive interpretation and do not support broad conclusions about effectiveness, acceptability across all enrolled participants, or uncommon harms.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Boston ARCH 4F Intervention to Reduce Fall Risk in People With HIV and Alcohol Use — NCT04804579.
• Registry record last updated 2023-06-29; status last verified 2023-06.
• Results first posted 2023-06-29.
• 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 record identifies NCT04804579 as a completed, randomized, parallel-group pilot sponsored by Boston University, with no masking and enrollment of 23. The study evaluates a 10-week fall-prevention intervention for people with HIV who consume alcohol. Its components are at-home exercises, virtual group sessions, and weekly phone check-ins; the control group receives an educational pamphlet about falls and alcohol use.

The primary endpoint is mean attendance at virtual group sessions among intervention participants who completed the 10-week follow-up assessment. The posted mean is 3.27 sessions, with a standard deviation of 3.44, among 11 participants. No control attendance result is reported. Secondary summaries cover physical function, frailty, and days of alcohol and other drug use; the posted unadjusted comparisons are imprecise, with confidence intervals spanning zero. A numerical result for change in self-reported falls is not reported in the registry record.

The registry reports zero serious and zero other adverse events affecting participants in both groups, and zero deaths. Missing-data handling, a multiplicity strategy, and a detailed statistical analysis plan are not reported. The record is best read as a pilot’s descriptive engagement and exploratory outcome evidence, rather than as a definitive comparative assessment.