Home Hazard Removal Program to Reduce Falls: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Home Hazard Removal Program to Reduce Falls (NCT02392013), a randomized na study. The primary outcome is Number of Falls.

2. Clinical Question

3. Trial Design

This was a randomized, parallel-group trial with no masking. Participants were assigned to a home-modification group or a usual-care group. The registry identifies Washington University School of Medicine as the sponsor and lists the condition as accidental falls. The study status is completed.

The registry records a total enrollment of 310. The brief description instead says investigators would conduct the trial in 300 older adults at risk for a fall. The record does not explain why these enrollment figures differ. The intervention was described as a hybrid effectiveness and implementation trial delivered through the aging services network. The intervention description specifies home visits by occupational therapists, three visits, and a booster session at six months. Usual care was described as care by an Area Agency on Aging.

The recorded study start was 2015-02-24, and primary completion was recorded as 2017-11. Participants were followed for 12 months according to the brief description. The primary outcome was number of falls; the posted secondary outcomes were daily activity performance, fear of falling, and health-related quality of life. The registry record does not report the randomization sequence method, allocation concealment, masking of outcome assessment, or further operational details of usual care.

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: No significant events occurred after participant enrollment but prior to randomization.

Allocation

Follow-up

Analysis population for the primary outcome, as defined in the registry: 135 intervention participants and 140 control participants were assessed for falls. This is different from the participant flow module, which shows that 127 intervention participants and 126 control participants completed follow-up, but the numbers in the participant flow module also took into consideration completion of secondary outcome measures.

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 Number of Falls, measured over 12 months.

The primary endpoint was the number of falls over 12 months. Falls were to be reported prospectively each day using a calendar. This endpoint represents a count of events, rather than a score or a time until a first event. Counting recurrent falls can retain information about repeated events that would be lost if an analysis considered only whether each participant fell at least once.

General statistical context: a count endpoint is often analyzed with methods designed for event counts. The choice of method depends on features such as variation in counts between participants and whether the data are more variable than a simple count model expects. Over-dispersion means that the observed variability exceeds the model’s expectation; accounting for it can affect uncertainty estimates and the comparison between groups. The registry does not report the distribution of individual fall counts, a dispersion estimate, or the analysis model used for the primary outcome.

The posted results give aggregate totals of 201 falls for 135 assessed participants in the home-modification group and 316 falls for 140 assessed participants in usual care. The registry does not provide the participant-level counts or an estimable two-arm comparison. These totals alone do not establish the precision of a group contrast.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a group mean difference represent?

For a continuous outcome, a mean difference subtracts the comparator group’s mean from the intervention group’s mean. A negative estimate means the intervention group’s recorded mean was lower; a positive estimate means it was higher. Whether either direction represents a favorable result depends on how the scale is defined. The registry does not provide enough information about the direction or clinical meaning of these secondary scales to assign benefit to a particular sign.

The secondary-outcome estimates are unadjusted recalculations from the posted group summaries. They do not incorporate covariates, account for any adjustment plan, or reproduce a sponsor analysis. The registry reports no analyses for these secondary outcomes. The comparisons are therefore descriptive statistical contrasts based on the posted means and standard deviations, not a record of the trial investigators’ inferential procedure.

How should confidence intervals be read?

General statistical explanation: a confidence interval describes the range of values compatible with the estimate under the assumptions of the method. A narrower interval indicates greater precision; a wider interval indicates more uncertainty. For a mean difference, an interval that includes zero is compatible with no mean difference as well as with differences in either direction. It does not prove the groups are equivalent. A confidence interval is not a range containing a fixed proportion of individual participants’ scores.

For daily activity performance, the recalculated mean difference is -0.92, and the 95% confidence interval runs from -1.973 to 0.133. For fear of falling, the estimate is 0.37, with an interval from -1.129 to 1.869. For health-related quality of life, the estimate is -1.48, with an interval from -4.205 to 1.245. All three intervals include zero. Their widths also indicate that the point estimates should not be read as precise measurements of a nonzero group difference.

What does a p-value add?

General statistical explanation: a p-value measures how unusual data at least as incompatible with a specified null model would be if that model and the test assumptions were correct. It is not the probability that the null hypothesis is true, nor a measure of the size or practical importance of an effect. The unadjusted p-values for the secondary outcomes are 0.087, 0.628, and 0.287. Each is above the two-sided significance level of 0.05 used in the planning illustration. Because no sponsor analysis is posted and no multiplicity procedure is described, these p-values are best treated as unadjusted exploratory summaries.

Why is the primary count result different?

The primary endpoint is a count of falls, while the posted secondary outcomes are means and standard deviations of scores. A count comparison needs an analysis that reflects event counts and their variation. General statistical context: over-dispersion matters when observed count variability is greater than a simple count model assumes; ignoring it can produce uncertainty estimates that are too small or otherwise poorly calibrated. Repeated events within one participant also mean that falls should not automatically be treated as independent observations. The registry does not report the model, its assumptions, the count distribution, or the handling of repeated events.

The record gives aggregate fall totals and assessed-participant counts but does not provide the information needed to reconstruct a defensible count-model comparison. It explicitly indicates that a two-arm comparison was not estimable for this outcome. No recalculated primary estimate, interval, or p-value is therefore reported in this statistical interpretation.

How should the planning illustrations be understood?

The sample-size planning illustrations are based on observed secondary-outcome effects, with a two-sided significance level of 0.05, equal allocation, and a normal approximation. They show participants per arm for 80% and 90% power at the stated fractions of the observed effect. For example, planning for 75% of the observed daily-activity effect gives 602 participants per arm for 80% power and 806 for 90% power. The corresponding figures for fear of falling are 7,542 and 10,097; for health-related quality of life they are 1,558 and 2,085.

General statistical explanation: such calculations are conditional planning illustrations. They depend on the estimated effect and variability used in the calculation, and smaller target effects generally require larger sample sizes when other assumptions remain fixed. They are not evidence about the probability that the completed trial detected an effect, and they are not the trial’s observed power. The registry does not report the original sample-size justification or assumptions used to design the trial.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome and three secondary outcomes. It does not report a multiplicity strategy, such as an adjustment to account for testing several outcomes or a hierarchy that would determine how results are interpreted. The three secondary p-values are therefore unadjusted recalculations, and their interpretation should account for the possibility of chance findings when several comparisons are considered. No sponsor-reported analyses are listed for these secondary outcomes.

The record does not report how missing outcome data were handled, whether imputation was used, or whether sensitivity analyses assessed the influence of missingness. Nor does it describe an intention-to-treat analysis, a per-protocol analysis, or another analysis-population definition. The participant counts differ across record components: the primary outcome description says 135 intervention and 140 control participants were assessed for falls, while participant flow reports 127 and 126 completed follow-up. The registry does not fully explain the relationship between these counts, and no assumptions about missing outcomes should be inferred from them.

The record also does not report a protocol-based statistical analysis plan, the intended model for the primary count outcome, interim analyses, or whether any analysis decisions changed after data collection. These details affect how confirmatory the results can be considered. In particular, the absence of an estimable primary comparison means there is no reported primary effect estimate against which the secondary outcomes can be formally contextualized.

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 zero serious adverse events among 155 participants at risk in each group. It also reports zero other adverse events among 155 participants at risk in each group. These are the posted event counts; they do not establish that no events occurred outside the registry’s reporting definitions or observation procedures.

The record separately lists 4 deaths in the home-modification group and 6 in usual care. It does not provide causes, timing, relationship to the intervention, or further details about how deaths were classified. The death counts should not be merged with the adverse-event counts or assumed to represent events caused by, or prevented by, either group. The registry does not list serious adverse-event terms in its posted summary. Safety figures are descriptive record entries and do not by themselves establish comparative safety.

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

The posted results establish that the registry records 201 falls in the home-modification group and 316 in usual care over the stated 12-month primary-outcome period. They also show group means and standard deviations for three secondary scores. These are observations from the trial record; the primary totals are not accompanied by an estimable comparison, and the secondary contrasts are unadjusted recalculations.

The data do not provide a confidence interval or p-value for the primary fall-count comparison. Without a reported analysis that accounts for the event-count structure and its uncertainty, the totals cannot quantify how precisely the groups differed. The different participant counts reported for fall assessment and follow-up further limit interpretation of denominators. Details on missing-data handling and analysis populations are not reported in the registry record.

The secondary estimates do not show a statistically distinguishable mean difference at the 0.05 level, but that does not prove equivalence or rule out differences that the estimates’ intervals still permit. The scales’ clinical interpretation and thresholds for meaningful change are not reported. Multiple outcomes were posted, while a multiplicity approach is not reported, which makes isolated unadjusted results less conclusive.

In general, a randomized design supports comparisons between assigned groups, but the strength of an inference also depends on complete outcome data, suitable analysis, and transparent reporting. The registry record leaves important analytic details unresolved. Accordingly, the posted results describe group observations but do not establish a definitive effect on falls or the secondary outcomes.

13. Why This Trial Matters Statistically

This trial illustrates a central issue in clinical biostatistics: the headline endpoint and the most easily compared results are not always the same kind of data. The primary outcome is a recurrent event count, whereas the secondary outcomes are continuous scores. Each requires an analysis suited to its structure. Aggregate event totals alone do not supply the variance or dependence information needed to quantify uncertainty in a count comparison.

The record also shows why denominators must be examined alongside outcomes. The fall-assessment counts differ from the follow-up-completion counts, and the registry does not fully explain the relationship. A reader cannot safely assume that all reported outcomes use an identical analysis population or that missing observations were handled in a particular way.

Finally, the secondary results demonstrate why estimates, intervals, and analysis plans matter together. A point estimate without precision can overstate certainty; a p-value without attention to multiplicity can overstate the strength of an isolated pattern. The record’s missing primary comparison, absent analysis details, and unadjusted secondary recalculations are therefore statistically important features of interpretation, not minor technicalities. They define what can be concluded from the posted results and what remains unresolved.

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.

Several limitations are specific to the information recorded for this trial. The primary outcome is presented as aggregate fall counts, without an estimable comparison, confidence interval, or p-value. The registry does not provide participant-level event distributions, a count model, or a dispersion measure. It is therefore not possible to assess from the posted record how repeated falls and between-participant variability were handled.

There is an unresolved difference between the primary-outcome assessment counts and the follow-up-completion counts. The record does not explain the relationship between 135 and 140 assessed for falls and 127 and 126 who completed follow-up. Details on missingness, exclusions, and analysis populations are not reported. The registry’s short description also refers to 300 participants, while enrollment is listed as 310; no explanation is given.

For the secondary outcomes, analyses are not reported in the registry, and the recalculated comparisons are unadjusted. A multiplicity plan, scale scoring details, thresholds for meaningful differences, and missing-data procedures are not reported. The registry also lacks details on allocation concealment, outcome-assessor masking, and the statistical analysis plan. These gaps limit assessment of bias, clinical interpretation, and the extent to which the posted comparisons match the trial’s intended analyses.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Home Hazard Removal Program to Reduce Falls — NCT02392013.
• Registry record last updated 2026-11-04; status last verified 2026-11.
• Results first posted 2024-08-30.
• 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

NCT02392013, titled “Home Hazard Removal Program to Reduce Falls,” is a completed randomized, parallel-group trial with no masking. It enrolled 310 participants and compared a home-modification program with usual care by an Area Agency on Aging. The intervention was delivered by occupational therapists over three home visits, with a booster session at six months.

The primary outcome was number of falls over 12 months, recorded prospectively using a daily calendar. The registry posts 201 falls for 135 assessed intervention participants and 316 for 140 assessed control participants, but no estimable two-arm comparison. Three 12-month secondary outcomes are posted: daily activity performance, fear of falling, and health-related quality of life. Recalculated unadjusted mean-difference intervals for these outcomes include zero. The registry does not report sponsor analyses, missing-data procedures, or a multiplicity plan.