Indiana University Dementia Screening Trial: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Indiana University Dementia Screening Trial (NCT01699503), a randomized na study. The primary outcome is Health Related Quality of Life (HRQOL).

2. Clinical Question

The trial was designed to assess the harms and benefits of dementia screening compared with no screening among older adults in primary care. Its primary outcome is health-related quality of life (HRQOL), recorded at baseline and at 1, 6, and 12 months. The registry identifies the primary measure as a mean on a scale and describes standard deviation as its dispersion measure.

The screening strategy includes more than administering a screening instrument: a score below 5 leads to referral for further assessment and care management. The comparator is usual care without study screening. The question therefore concerns the outcomes of assignment to these different care pathways, not simply the performance of the screening instrument as a diagnostic test. The registry record does not report whether participants or clinicians were masked, or who the single masking designation refers to.

The record does not report the primary outcome values or an estimated difference between groups. It therefore does not establish from the posted results whether HRQOL differed between the groups, or how precisely any difference was estimated.

3. Trial Design

This was a randomized, parallel-group clinical trial sponsored by Indiana University. The registry lists the allocation as randomized and the model as parallel. Its masking designation is single, but the registry record does not identify which people involved in the trial were masked. The study is listed as completed, with enrollment of 4005.

The two groups were a Screening Group and a No Screening group. Participants assigned to no screening received the usual standard of care. Participants assigned to screening received the MIS. Those with an MIS score below 5 were referred to the Collaborative Dementia Care Program for subsequent diagnostic assessment, counseling, and management. The listed intervention is behavioral. The record does not describe additional details of how the program was delivered or how usual care varied across practices.

The registry describes the trial’s purpose as assessing the harms and benefits of screening for dementia, compared with no screening, among older adults cared for in typical primary care practices. The conditions listed are dementia, Alzheimer’s disease, and mild cognitive impairment. The study’s primary outcome is HRQOL, assessed at baseline and at 1, 6, and 12 months. Other listed outcomes concern health care utilization and whether participants had an advance directive at 12 months.

Random assignment is intended to create groups that are comparable on average at the time of assignment. In general, that supports a comparison of outcomes by assigned group, although chance imbalances can remain. A parallel design means participants are assigned to one group rather than crossing between the two approaches. The registry record does not report details of the randomization procedure, allocation concealment, or masking beyond its single-masking designation.

The record does not report a statistical analysis plan, an analysis population definition beyond the primary-outcome note, or a sample-size justification. It lists 5 outcome measures and 0 analyses. The registry record also does not report the primary outcome findings.

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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The number analyzed for each row excludes those participants who died, withdrew, were lost to follow-up, or were excluded due to quality at each time point in the 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 Health Related Quality of Life (HRQOL), measured over 1 month, 6 months, 12 months.

The primary endpoint is health-related quality of life (HRQOL), measured with the 15-item Health Utility Index (HUI). The registry describes the HUI as a generic, utility-based instrument used in people with a wide range of medical conditions. It includes eight attributes: vision, hearing, speech, ambulation, dexterity, emotion, cognition, and pain.

Individual health-domain scores range from 0.00, described as maximum impairment, to 1.00, described as no impairment. The multi-attribute HUI index is a multiplicative function of individual attribute levels. Its stated range is 0.36 to 1.00, with anchors of 0.00 for dead and 1.00 for perfect health. The registry specifies a mean as the summary measure and standard deviation as the measure of dispersion.

Using a broad, generic measure allows the outcome to reflect several aspects of health rather than focusing on one symptom or condition. In general, a utility-based measure combines health attributes into a summary score; the meaning of a group average depends on the instrument’s scale and scoring. HRQOL was recorded at baseline and at 1, 6, and 12 months, allowing the planned assessment to cover multiple points after assignment. The registry record does not report the observed scores, group comparisons, or changes over time.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a mean and standard deviation describe?

The registry specifies the mean as the primary outcome summary and standard deviation as its dispersion measure. In general, a mean describes the average score among the observations included in an analysis. A standard deviation describes how spread out individual scores are around that average. They answer different questions: the mean concerns the group’s average level, while the standard deviation concerns variation among people in that group.

For an HUI score, interpretation also depends on the instrument’s scale and scoring. The registry describes the domain-score range as 0.00 to 1.00 and gives the HUI index range as 0.36 to 1.00, with its stated anchors. An average score is a summary across participants; it does not imply that every participant has that score or that changes in one attribute are the same as changes in another. The registry does not report the observed means or standard deviations.

What would a group comparison estimate?

In general, a comparison of the randomized groups could estimate the difference in average HRQOL between participants assigned to screening and participants assigned to no screening. The interpretation would depend on the specified time point or time points and on whether the estimate is based on assigned group or on the care actually received. The registry identifies a randomized trial, but the analysis approach and estimand are not reported in the registry record.

A confidence interval, when reported, describes a range of values compatible with the data and the statistical model under its assumptions. It conveys precision as well as the estimated direction and magnitude. A p-value addresses how compatible the observed data are with a specified null hypothesis under the analysis assumptions; it does not measure clinical importance or the probability that a hypothesis is true. No confidence interval or p-value for the primary outcome is reported in the registry record, so neither precision nor statistical evidence for a group difference can be assessed.

How do repeated measurements affect analysis?

The primary outcome is recorded at baseline and at 1, 6, and 12 months. In general, repeated scores from one person are related rather than independent. An analysis that uses several measurements from each participant needs to account for that within-person association. Analysts may also define whether the main estimate concerns a particular follow-up time, change from baseline, or an overall time pattern; each choice answers a distinct question. The registry record does not report which approach was used or how baseline measurements were incorporated.

Repeated measurement also makes missingness important. The record says that the number analyzed at each time point excludes those who died, withdrew, were lost to follow-up, or were excluded because of quality. It does not give counts by reason or explain whether the analysis used only observed scores, imputed values, or another method. In general, estimates based only on observed scores can be biased if the likelihood of a score being unavailable is related to the outcome, especially after accounting for measured information. The direction and size of any such bias cannot be determined from the information reported.

Why does dispersion matter?

For a continuous outcome, variability affects how precisely a group mean or a between-group estimate can be measured. In general, greater variability makes an estimate less precise when other design features are held constant. The standard deviation is therefore useful for describing score variation, while an uncertainty interval for a group contrast would describe the precision of that contrast. Because the registry does not report the outcome’s summary statistics or an interval estimate, the amount of variability and the precision of any comparison are unknown.

Over-dispersion is a general concern in analyses of count outcomes when the observed variability exceeds what an assumed count distribution allows. It can make standard errors too small if the model does not account for that extra variation. That issue is not a description of the continuous HUI endpoint; it is a general consideration for count outcomes such as utilization events, if those are analyzed as counts. The registry record does not report the statistical models used for its secondary outcomes.

What can be said about planning calculations?

The registry describes a planning illustration based on an observed effect, with a two-sided significance level of 0.05, equal allocation, and a normal approximation. It says that the effect to detect is a stated fraction of the observed effect. For time-to-event outcomes, it describes a total-events basis; for other outcomes, it describes participants per arm. These are planning concepts, not evidence about what the trial observed or achieved.

In general, a sample-size calculation depends on the target effect, outcome variability or event rate, design assumptions, and chosen error rate. A calculation based on an observed effect is illustrative and does not establish the trial’s original design rationale. It is not observed power, and it should not be interpreted as a post hoc test of whether the study succeeded. The registry record does not report a sample-size justification or the trial’s own design assumptions. No effect estimate or sample-size illustration value is reported in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The primary HRQOL measure is scheduled at baseline and at 1, 6, and 12 months. Considering several time points can create multiple opportunities to test for a difference. In general, when multiple hypotheses are tested, the chance of at least one apparently unusual result can increase unless the analysis plan accounts for multiplicity. A prespecified primary time point, a hierarchical testing strategy, or an adjustment can help define how multiple comparisons are interpreted. The registry record does not report a multiplicity strategy or identify a single primary follow-up time point for inference.

The registry also lists secondary outcomes for health care utilization and advance directives at 12 months, alongside 5 outcome measures overall. It does not report which outcomes were included in formal analyses, how outcomes were prioritized, or whether any adjustment across outcomes was planned. The record lists 0 analyses, so there are no reported analysis results to evaluate for multiplicity.

For HRQOL, the registry says that the number analyzed at each time point excludes participants who died, withdrew, were lost to follow-up, or were excluded because of quality. The record does not report how many participants fell into each category, whether exclusions overlapped, or how many contributed data at each time point. It also does not describe imputation, sensitivity analyses, or assumptions about missingness.

These omissions limit interpretation because the analyzed population and missing-data approach can affect both the estimate and its uncertainty. In general, excluding participants with unavailable measurements can change the composition of the group being summarized. Without the counts and methods, the potential effect cannot be assessed. Further details about analysis populations, missing observations, and multiplicity are not reported in the registry record.

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 5% reporting threshold.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: incident severe depression ( (2.0%), incident severe anxiety (1.1%).

The registry records serious adverse events in 52 of 1723 participants at risk in the Screening Group and 47 of 1693 participants at risk in the No Screening group. It records 0 other adverse events in each group, with 1723 and 1693 participants at risk, respectively. The registry also lists 21 deaths in the Screening Group and 25 in the No Screening group. These figures are descriptive counts as recorded, not adjusted comparisons.

The serious-event terms listed include Incident Severe Depression ( at 2.0% and Incident Severe Anxiety at 1.1%. The registry record does not further explain the denominator or calculation for these percentages in the information available for this summary. The figures should therefore be read as the registry’s listed entries rather than as independently recalculated rates.

Counts alone do not show whether an event was caused by the assigned strategy, whether groups differed beyond what could arise by chance, or how precise a comparison would be. The record does not report a statistical comparison, confidence interval, or event-attribution details for the safety figures. Deaths should not be assumed to be caused by screening based on these counts. The posted information describes the recorded safety figures but does not establish a causal interpretation.

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

The registry record describes a randomized comparison of dementia screening with no screening in older adults receiving primary care. The screening pathway includes the MIS and referral to the Collaborative Dementia Care Program for participants whose score is below 5. The primary outcome is HRQOL measured with the HUI at multiple follow-up times. These details define the clinical question and the planned outcome, but they are not themselves evidence of an outcome difference.

The primary HRQOL results are not reported in the registry record. There is consequently no reported estimate of the difference between groups, no confidence interval describing precision, and no p-value for that outcome. The available information cannot establish whether the groups had different average HRQOL, the direction or magnitude of any difference, or whether any difference was precise. Lack of reported results is not evidence that the two strategies had the same result.

The safety figures provide counts of serious adverse events, other adverse events, and deaths by group. Without a reported comparative analysis, attribution information, or uncertainty estimates, these counts cannot establish a causal effect or quantify how much the groups differed. They should be understood as descriptive registry entries.

In general, a sound interpretation of a randomized trial combines the outcome estimate with its uncertainty, the analysis population, missing-data methods, and the prespecified analysis plan. The registry record does not report these details for the primary outcome. It also lists 0 analyses. As a result, the record supports a description of the trial design and the measures it intended to collect, but not a conclusion about the comparative HRQOL results.

The trial’s scope is also specific: a screening strategy in typical primary care practices, with referral to a named program for participants meeting the stated MIS threshold. Findings from a different setting or screening pathway would address a different question. The registry record does not report the primary outcome findings needed to characterize the results of this trial’s comparison.

13. Why This Trial Matters Statistically

This trial raises a statistical question that is broader than whether a screening instrument identifies people for further assessment. The assigned strategy combines screening with a referral pathway, while the comparator is usual care without study screening. The main outcome is a generic, utility-based HRQOL measure, intended to reflect several dimensions of health. This makes the estimand and the timing of measurement important: a comparison at one follow-up time may answer a different question from an overall comparison across follow-up.

Random assignment offers a basis for comparing groups, but interpretation still depends on how outcomes were analyzed and who contributed data. The registry specifically notes exclusions at each time point for death, withdrawal, loss to follow-up, or quality-related reasons. If the set of participants contributing scores changes over time, the means at different visits may summarize different groups. Clear reporting of denominators, missingness, and analysis populations is therefore central to understanding a longitudinal endpoint.

The record also illustrates why outcome definitions and results must be distinguished. It gives a detailed account of the HUI scale and states that the primary summary is a mean with standard deviation, but does not report observed scores or a group contrast. A well-described measure does not substitute for an estimate and its precision. Similarly, safety counts are useful descriptive information but do not by themselves quantify causal differences.

Finally, the registry lists multiple outcomes and follow-up times, yet reports no analyses. Without an analysis plan or results, readers cannot determine how multiplicity was handled, which comparison was prioritized, or how missing data influenced inference. These are not peripheral details: they shape what a reported estimate would mean and how confidently it could be interpreted. For this record, the statistically appropriate emphasis is on the documented design and the limits created by absent results.

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 most consequential limitation for interpreting the registry record is that it does not report the primary HRQOL results. There are no group means, standard deviations, estimates of between-group differences, or measures of uncertainty. The registry lists 0 analyses, leaving no reported inferential findings to assess. The clinical question is described, but its answer cannot be established from the posted results.

The primary outcome is measured at several time points, but the registry record does not specify which time point was intended as the principal inferential comparison, whether change from baseline was analyzed, or how repeated observations were modeled. It also does not report how baseline scores were incorporated. These omissions prevent readers from identifying the exact estimand or evaluating the approach to within-participant correlation.

The record notes that participants who died, withdrew, were lost to follow-up, or were excluded due to quality are not included in the number analyzed at each time point. It does not provide the corresponding counts, explain the quality exclusions, or describe how missing observations were handled. Consequently, the extent and possible implications of missing data cannot be assessed.

There is no reported statistical analysis plan, multiplicity approach, or sample-size justification. The registry includes a planning-illustration description, but does not report specific illustration values or the trial’s own design rationale. The single-masking designation is not accompanied by an explanation of who was masked. Details of randomization procedures and allocation concealment are also not reported in the registry record.

Safety information is limited to recorded counts and listed event terms. Comparative uncertainty and attribution are not reported. These gaps mean that the record supports a description of the trial and its intended measures, but not a full assessment of efficacy, precision, safety differences, or potential biases.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Indiana University Dementia Screening Trial — NCT01699503.
• Registry record last updated 2019-01-23; status last verified 2019-01.
• Results first posted 2019-01-23.
• 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.

18. Record Summary

The Indiana University Dementia Screening Trial, also identified as the IU CHOICE Study, is a completed randomized, parallel-group trial sponsored by Indiana University. The registry records enrollment of 4005 and describes a study of dementia screening compared with no screening among older adults in primary care.

The screening group received the MIS; participants with a score below 5 were referred to the Collaborative Dementia Care Program for diagnostic assessment, counseling, and management. The no-screening group received usual standard of care. The primary outcome is mean HRQOL measured with the 15-item HUI at baseline and at 1, 6, and 12 months.

The registry lists 5 outcome measures and 0 analyses. It describes exclusions from the analyzed number at each time point, but does not report the primary outcome results, analysis methods, or missing-data procedures. Serious adverse-event and death counts are listed without a comparative analysis or attribution details. The record therefore documents the design and planned measures, but does not report findings that answer the primary comparison.