Prevention of OUD: The HOME Project (Housing, Opportunities, Motivation and Engagement): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Prevention of OUD: The HOME Project (Housing, Opportunities, Motivation and Engagement) (NCT04135703), a randomized na study. The primary outcome is Frequency (% Days) of Opioid Use in Prior 3 Months.

2. Clinical Question

3. Trial Design

The HOME Project was a randomized, single-masked, parallel-group trial sponsored by Ohio State University. The registry lists its status as completed and enrollment as 240. It records a start date of 2020-06-23 and a primary completion date of 2024-05-14. The record does not identify who was masked, so the masking description cannot establish whether participants, service providers, outcome assessors, or another group were unaware of assignment.

The two arms were Risk Prevention Services +Housing and Risk Prevention Services Only. The registry lists 120 participants in each arm. Both arms received opioid and related risk-prevention services, while rental and utility assistance for 6 months was specified only in the housing arm. This design addresses the added contribution of the housing assistance package within the service approaches described; it does not compare receiving prevention services with receiving no services.

The primary outcome was a continuous measure reported as the mean percentage of days with opioid use in the prior 90 days. The registry describes assessment at baseline and at 3, 6, 9, and 12 months post-baseline, with the primary outcome timeframe listed as 12 months. The Form 90 interview was the stated assessment method.

The registry record identifies the trial as randomized but does not report the randomization procedure, allocation concealment, detailed masking procedures, or a statistical analysis plan. It also does not report an outcome comparison. Accordingly, the design features can be described, but the record does not support a conclusion about the estimated difference between assigned arms.

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: Failed: (n=154) * Not homeless (n=20) * Over the age of 24 (n=2) * Under the age of 18 (n=7) * Met DSM 5 criteria for OUD (n=27) * Declined Participation (n=39) * Failed to return (n=43) * Refused to wear a mask during intake (n=1) * Unable to focus to complete intake (n=1) * Difficulty/unable to answer questions (n=5) * Could not show proof of Guardianship (n=2) * Had difficulty understanding the project/process (n=7)

Overall Study

Analysis population for the primary outcome, as defined in the registry: Frequency of opioid use in prior 90 days (percentage of days use) was assessed at each follow-up point (3, 6, 9 and 12-months post-baseline) using the Form 90 interview.

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: Recruited as proposed.

6. Primary Endpoint

The registry lists the primary outcome as Frequency (% Days) of Opioid Use in Prior 3 Months, measured over 12 months.

The primary endpoint measures the frequency of opioid use as the percentage of days with use during the prior 90 days. It is a continuous measure: each participant’s reported frequency can take values across a range rather than being limited to a yes-or-no outcome. The registry states that the Form 90 interview assessed this outcome at baseline and at follow-up visits at 3, 6, 9, and 12 months post-baseline. The primary timeframe is listed as 12 months.

In general, a frequency measure can describe changes in use that a simple indicator of any use might not capture. It can distinguish, for example, between different reported frequencies among people who have some use. That makes it relevant to a prevention trial focused on opioid-use patterns. This general advantage does not guarantee that the measure is complete, unbiased, or accurately reported in any particular study.

The registry record does not report estimable arm means or a comparison for this endpoint. Although it displays 0.0 as the mean and 0.0 as the spread for each arm, it also indicates that values are missing or not estimable for a two-arm comparison. Those displayed zeros should therefore not be read as observed opioid-use frequencies or as evidence that use was absent. Without usable outcome estimates, the endpoint’s comparison between the assigned approaches cannot be assessed from the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the outcome measure? The primary endpoint is continuous and is reported as the mean percentage of days with opioid use in the prior 90 days. The registry says the Form 90 interview was used and lists assessments at baseline and at 3, 6, 9, and 12 months after baseline. The primary timeframe is 12 months. A continuous outcome is commonly summarized using a mean and a measure of spread, such as a standard deviation. The mean describes the average level in a group; the spread describes how much participant-level values vary around that average.

Why is an arm comparison needed? The study’s randomized parallel-group design assigns participants to one of two service approaches. The statistical question is not answered by describing either arm alone: it requires a comparison of the outcome between arms at the relevant follow-up, interpreted with appropriate uncertainty. In general, a confidence interval expresses the range of values compatible with the data and model assumptions, while a hypothesis test assesses compatibility with a specified null hypothesis. Neither can be interpreted for this primary comparison from the posted record because an estimable comparison is not reported.

What does “not estimable” mean in this record? The record flags the values as missing or not estimable for a two-arm comparison. That is different from a measured value of zero. The displayed 0.0 entries for both the mean and spread should not be treated as valid outcome estimates because the registry’s own status indicates that the comparison could not be estimated. The record does not explain why the values were unavailable or not estimable. Possible explanations should not be guessed from the record.

Can the posted values be recalculated? Recalculation of a two-arm continuous-outcome comparison would ordinarily require usable group summaries and the relevant sample information. Here, the record does not provide an estimable effect or a valid spread for the outcome. It therefore does not support a meaningful unadjusted recalculation of the primary contrast. No recalculated result should be inferred from the displayed zeros. The record does not report a sponsor analysis for this outcome that could be compared with a recalculation.

What do the planning assumptions mean? The stated sample-size assumptions describe planning illustrations based on an observed effect, a two-sided significance level of 0.05, equal allocation, and a normal approximation. The effect to detect is described as a stated fraction of the observed effect. Such calculations ask how many participants per arm might be needed under specified assumptions; they are not results from the trial and are not observed power. A planning illustration depends on the effect used as its starting point, as well as assumptions about variability, analysis, and the desired operating characteristics. The registry record does not report an estimable primary effect to interpret alongside those assumptions. No sample-size illustration should be read as evidence that the trial had a particular probability of detecting an effect.

Why does variability matter? In general, greater variability makes a group mean less precise when other design features are held constant. If the outcome distribution is highly variable or has features not well represented by a normal approximation, the chosen analysis may need to address that. This is a general statistical consideration, not a finding about the HOME trial: the registry does not report the observed outcome distribution, a model assessing it, or an analysis plan. The record also does not say whether an adjusted model, repeated-measures model, or another approach was used.

How should repeated assessments be understood? The registry describes repeated assessment at baseline and several post-baseline visits. In general, measurements from the same participant are related, so a longitudinal analysis may account for within-person correlation rather than treating each visit as an unrelated observation. The record does not report how repeated measurements were handled, whether the primary contrast used the 12-month assessment alone, or whether earlier follow-up measurements contributed to the primary analysis. Those details are not reported in the registry record.

What can be concluded statistically? The study’s assignment structure and intended outcome are documented, but the primary two-arm estimate is not. Statistical concepts can clarify what an estimate, spread, confidence interval, or planning calculation would mean, but they cannot recover missing outcome information. The registry record does not provide a primary effect estimate, uncertainty interval, or test result. Accordingly, the magnitude, direction, and precision of the primary comparison remain unavailable in the posted results.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome measure and no secondary outcomes posted. It also lists no analyses. This means the registry record does not provide a set of reported outcome tests from which to assess how multiple statistical comparisons were handled. It does not establish that the trial had no planned secondary outcomes, subgroup analyses, or other analyses; those details are not reported in the registry record.

In general, when many outcomes or comparisons are tested, the chance of finding an apparently notable result by chance can increase. Statistical plans may address this through prespecified priorities or multiplicity adjustments. The registry does not report whether such methods were planned or used. With no posted primary comparison, there is also no reported test result whose interpretation could be assessed for multiplicity.

The record describes outcome assessments at baseline and at 3, 6, 9, and 12 months, but does not report the amount or pattern of missing outcome data at these assessments. It does not describe follow-up completion, reasons for missing measurements, imputation, sensitivity analyses, or methods for handling participants with incomplete repeated measures. These matters are not reported in the registry record and should not be inferred from the arm enrollment entries.

The analysis population is also not reported. The registry record does not specify whether the primary analysis followed assignment regardless of intervention received, included only participants with observed outcome data, or used another population definition. Without that information and an estimable outcome comparison, the impact of missing data and analysis-population choices cannot be evaluated.

11. 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 registry reports serious adverse events affecting 0 of 120 participants at risk in the Risk Prevention Services +Housing arm and 0 of 120 at risk in the Risk Prevention Services Only arm. It also reports other adverse events affecting 0 of 120 at risk in each arm. Separately, the death counts are 4 in the housing arm and 1 in the services-only arm.

These are counts recorded in the registry, not estimates of comparative risk. The record does not describe the causes or timing of the deaths, whether they were judged related to study participation or an intervention, or how the adverse-event categories were defined and collected. The reported death counts therefore cannot establish that either service approach caused or prevented deaths. The serious and other adverse-event counts should not be taken to explain or reclassify the deaths.

In general, safety counts require context about surveillance, follow-up, event definitions, and exposure time. Those details are not reported in the registry record. Nor does the record provide a formal statistical comparison of the safety outcomes. The counts should be read as descriptive registry entries, with their interpretation limited by the absence of further context.

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

The record documents a randomized comparison of risk-prevention services plus housing assistance against risk-prevention services alone among a trial population described as homeless youth. It specifies the intervention components, the primary opioid-use frequency outcome, and repeated assessment through 12 months. Those details clarify the intended question: whether the added housing assistance, as part of the described service package, was associated with a different frequency of opioid use.

However, the posted primary outcome values are marked missing or not estimable for a two-arm comparison. The displayed 0.0 means and 0.0 spreads do not provide valid evidence that the observed outcome was zero or identical between arms. The record contains no estimable primary effect, confidence interval, or p-value. It therefore does not show the direction, size, or precision of a primary-outcome difference.

This absence of an interpretable comparison does not establish that the approaches are equivalent, that one is better, or that neither has an effect. It means the registry record does not make the primary outcome comparison assessable. The randomized design can strengthen a properly reported comparison, but it cannot fill in missing outcome estimates.

The record’s safety counts are separate descriptive information. Without causes, timing, event definitions, or a formal safety analysis, the listed deaths and adverse-event counts cannot be attributed to the interventions or used to draw a comparative safety conclusion. They also do not resolve the missing primary-outcome result.

Accordingly, the result’s meaning is limited: the trial’s question and design are recorded, but the posted results do not answer the primary statistical question. Further outcome and analysis information is not reported in the registry record.

13. Why This Trial Matters Statistically

This trial addresses a statistical and design challenge: evaluating a service package in a population that the registry describes as facing substantial barriers and health challenges. The randomized parallel-group design compares two active approaches, with housing assistance specified in one arm alongside prevention services. That comparison is relevant to the incremental question of what the housing component contributes within the described program, rather than to the effect of prevention services versus no services.

The primary outcome also illustrates why choosing a measure matters. Frequency of opioid use, expressed as the percentage of days with use in the prior 90 days, captures more detail than a simple indicator of any use. Repeated assessment through 12 months can describe outcomes over time, although the registry record does not report how those measurements were incorporated into an analysis.

Statistically, the case underscores that a randomized design is only one part of the evidence. A usable outcome estimate, its uncertainty, and clear information about the analysis are needed to interpret a treatment contrast. Here, the primary values are not estimable for a two-arm comparison, and no analyses are posted. This prevents readers from determining the magnitude or precision of the trial’s primary result from the registry record.

The listed adverse-event and death counts also demonstrate why descriptive counts require context. Without information on timing, causes, follow-up, and attribution, they cannot establish a causal comparison. The record is therefore important for identifying the intended question and design, while also showing the limits of interpretation when outcome analyses are not available.

14. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The main limitation for interpreting this trial’s primary result is that the registry record marks the continuous endpoint values as missing or not estimable for a two-arm comparison. Although the arm records display 0.0 for the mean and spread, those entries cannot be treated as observed outcome summaries in light of that status. No effect estimate or uncertainty measure for the primary comparison is reported.

The registry does not report detailed eligibility criteria or participant characteristics, so the population represented by the enrolled participants cannot be described beyond the record’s focus on homeless youth. It also does not report the randomization procedure, allocation concealment, who was masked, or the statistical analysis plan. These omissions limit assessment of design implementation and analytic choices.

Repeated assessments are described, but the record does not explain how the visits were analyzed, how missing outcomes were handled, or which analysis population was used. It does not report follow-up completeness, reasons for missing data, or sensitivity analyses. The absence of posted secondary outcomes and analyses does not establish that none were planned or conducted.

The safety information is limited to adverse-event and death counts. Causes, timing, event definitions, intervention attribution, and comparative analyses are not reported in the registry record. These limitations prevent causal interpretation of the safety counts as well as a detailed assessment of the primary outcome.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Prevention of OUD: The HOME Project (Housing, Opportunities, Motivation and Engagement) — NCT04135703.
• Registry record last updated 2025-09-15; status last verified 2025-01.
• Results first posted 2025-09-15.
• 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.

18. Record Summary

The HOME Project, NCT04135703, is listed as a completed randomized, single-masked, parallel-group trial sponsored by Ohio State University. Enrollment was 240, with 120 participants listed in each arm. The comparison was risk-prevention services plus housing assistance versus risk-prevention services alone.

The primary outcome was mean frequency of opioid use, expressed as percentage of days with use in the prior 90 days, assessed by Form 90 through 12 months after baseline. The registry lists assessment at baseline and at 3, 6, 9, and 12 months. It marks the outcome values as missing or not estimable for a two-arm comparison; the displayed zeros are not interpretable outcome results. No analyses are posted.

The registry reports serious and other adverse-event counts of 0 in each arm, and death counts of 4 in the housing arm and 1 in the services-only arm. Context needed to attribute or compare these events is not reported in the registry record.