Technology Intervention to Support Caregiving for Alzheimer's Disease (I-CARE): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Technology Intervention to Support Caregiving for Alzheimer's Disease (I-CARE) (NCT03119259), a randomized na study. The primary outcome is Feasibility: Data Completeness.

2. Clinical Question

The central question is about feasibility and estimation, rather than a definitive test of whether the app changes caregiver or patient outcomes. The registry identifies feasibility of BCN as a specific aim and names data completeness as the primary outcome. It also describes caregiver burden and behavioral and psychological symptoms of dementia (BPSD) as outcomes for which the study sought effect-size estimates. The registry record does not report those estimates or an analysis comparing the groups on these measures.

The clinical context is caregiving for Alzheimer’s disease, while the listed condition is Dementia, Alzheimer Type. The app is intended to support caregiving in the setting of existing clinical care; the intervention is not described as replacing that care. Because the comparison group also received clinical care, a group comparison would concern BCN in addition to the recorded care arrangement, not BCN versus no care. The record does not establish how app use was measured or how closely participants used it.

3. Trial Design

I-CARE is a randomized, parallel-group study marked completed in the registry. Its phase is listed as NA, and masking is listed as single. The record does not identify which people were masked or explain the masking procedures. Allocation is recorded as randomized, but details about sequence generation, allocation concealment, or implementation are not reported in the registry record.

The enrollment is 53. The registry lists Indiana University as the sponsor and identifies recruitment through the Aging Brain Care program of Eskenazi Health and Indiana University Health in central Indiana. The recorded study start is 2019-11-25, and primary completion is 2021-07-31. These dates describe the registry record; they do not by themselves establish when every participant entered or completed follow-up.

The comparison group received care through the ABC Clinical Program and IUHP. The intervention group continued those care arrangements and was assigned BCN software. The available intervention description says installation was on a caregiver’s personal mobile device if it met minimal technical requirements, or on a device from the study. The recorded arm description ends before the sentence is complete, so further details about the intervention should not be inferred. The registry also uses the labels “ABC Clinical Program Only and Usual Care” and “BCN Mobile App Plus ABC and BCN Mobile app only,” which do not fully clarify the intervention components for patients and caregivers individually.

Outcome listings include caregiver burden and BPSD at baseline, 3 months, and 6 months, along with acute care utilization at 3 months and 6 months. These listings describe intended measurements, not results. The registry reports no analyses, and it does not describe the statistical model or an analysis population. Consequently, randomization and the planned follow-up structure can be described, but the comparative evidence from the study cannot be assessed from the posted results.

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: Caregivers recruited to 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 Feasibility: Data Completeness, measured over 6 months after enrollment and randomization.

The primary outcome is titled “Feasibility: Data Completeness” and is measured as a count of participants. Its description defines the count as the number of enrolled and randomized individuals who contributed data at 3 and 6 months. Data completeness is a process measure: it indicates whether outcome information was obtained at follow-up, rather than how caregiver burden or patient symptoms changed.

Such a measure is useful in a feasibility study because collecting follow-up data is necessary for evaluating whether the study procedures can support later research. If follow-up information is absent, the intended clinical outcomes may be difficult to estimate. A count alone, however, does not explain why data are missing, whether missingness differs between groups, or how representative the people with follow-up data are of everyone randomized. Those details are not reported in the registry record.

There is a wording distinction in the primary outcome entry: its timeframe is described as 6 months after enrollment and randomization, while the outcome definition refers to data at both 3 and 6 months. The record does not explain this distinction. It also does not report the observed completeness counts, denominators by follow-up time, or a comparative analysis. Therefore, the endpoint’s purpose and definition can be described, but the achieved completeness cannot be assessed from the registry results.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The record identifies a randomized, parallel-group design, a single masking designation, an enrollment of 53, and a feasibility endpoint based on counts. These details establish the broad design, but do not provide the statistical methods used to analyze the outcomes. The distinction matters: knowing how participants were assigned is not the same as knowing how the resulting data were summarized or compared.

What does randomization contribute?

In general, random allocation is intended to reduce systematic differences in prognosis between treatment groups at the start of a study. It supports interpretation of a group comparison when the assigned groups are analyzed appropriately and follow-up is sufficiently complete. It does not guarantee exact balance in a small study, and it cannot resolve bias introduced after allocation, such as differential loss to follow-up or differences in outcome measurement. Baseline characteristics, allocation procedures, and the method for handling participants who did not follow their assignment are not reported in the registry record.

What does a count endpoint measure?

For data completeness, the endpoint is a count of enrolled and randomized individuals who contributed data at the stated follow-up times. Counts can describe whether information was collected, but they do not directly quantify the clinical outcome. A count is most interpretable alongside its denominator and the definition of what qualifies as complete data for each visit. The record does not give observed counts, time-specific denominators, or rules for partial questionnaires or other incomplete measurements. Without those details, it is not possible to assess the achieved follow-up completeness or the reasons for missing information.

General statistical practice distinguishes missing outcome data from an observed outcome value. Missing data do not, by themselves, reveal whether a participant improved, worsened, or remained unchanged. If missingness differs by group or is related to participant experience, a comparison among those with observed data may not represent the randomized groups. Methods for addressing missing data, including any assumptions or sensitivity analyses, are not reported in the registry record.

How should planned estimates be understood?

The study aims describe estimating the effect size of BCN on caregiver burden and BPSD at 6 months. An effect estimate expresses the size and direction of a contrast under a defined outcome scale and analysis. Its interpretation depends on the outcome definition, comparison, and method used. Precision is commonly described with a confidence interval, which gives a range of values compatible with the data under the assumptions of the analysis. The registry reports no effect estimates or confidence intervals, so neither magnitude nor precision can be assessed.

For this trial, repeated outcome listings at baseline, 3 months, and 6 months create a need to clarify how observations over time were modeled or summarized. General approaches may compare a prespecified follow-up measurement, model changes over time, or account for correlation among repeated observations from the same participant. These approaches answer related but not identical questions. The registry does not report which approach was used, whether baseline values were incorporated into the comparison, or how within-participant correlation was handled. No particular approach should be assumed.

The same issue applies to the listed acute care utilization outcomes at 3 months and 6 months. The record does not state the form of those data, the statistical model, or whether utilization was analyzed as a count, a binary indicator, or another measure. A method appropriate for one outcome structure may not be suitable for another. The available information is not sufficient to describe the actual analyses.

Why does over-dispersion matter in general?

For count outcomes, a statistical model may rely on assumptions about how much variation is expected relative to the average count. Over-dispersion means that the observed variation is greater than a model’s basic assumptions allow. If it is ignored, uncertainty estimates may be too small and tests may appear more conclusive than warranted. This is a general consideration, not a claim about the data in I-CARE. The registry does not report a count-outcome model, a dispersion assessment, or a method for accommodating over-dispersion.

Similarly, binary outcomes generally require a clearly defined event and comparison, while time-to-event analyses require event timing and follow-up information. The registry’s broad outcome listing and the adverse-event counts do not establish that these methods were used. The statistical analysis plan, model specifications, covariate adjustments, and assumptions are not reported in the registry record.

What do the planning assumptions mean?

The registry describes a planning illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. It says that the effect to detect is a stated fraction of the observed effect. For time-to-event tables, the stated quantity is total events; for other designs, it is participants per arm. These are planning assumptions, not observed trial results or evidence of achieved power. The registry does not identify a sample-size justification from the trial’s own design. Because it reports no observed effect estimate, no numerical planning illustration can be interpreted from the available result entries.

More generally, a planning sample size describes what a future study might require under chosen assumptions. It does not establish the probability that a completed study has detected an effect, and it cannot replace estimates and uncertainty from the completed study. Any such planning value should be treated as conditional on the assumed effect and model, rather than as a finding about participants in I-CARE.

The registry also labels the study single-masked, but does not identify the masked party or explain whether outcome assessors were masked. Masking can reduce some forms of measurement or expectation bias, depending on who is masked and what outcomes are assessed. Without the operational details, the likely protection offered by masking cannot be evaluated. In sum, design labels are present, but the specific statistical methods and analysis results needed for inference are not reported.

10. Multiplicity, Missing Data and Analysis Populations

The record lists 14 outcome measures and 0 analyses. The planned outcomes include caregiver burden and BPSD at baseline, 3 months, and 6 months, and acute care utilization at 3 months and 6 months. Multiple outcomes and follow-up times can create multiple opportunities to assess group differences. General statistical practice treats this as a multiplicity issue because examining many comparisons can increase the chance of highlighting a result that appears notable by chance. Whether I-CARE used a prespecified primary comparison, a multiplicity adjustment, or another strategy is not reported in the registry record.

The primary outcome is feasibility: data completeness, defined by counts of enrolled and randomized individuals contributing data at 3 and 6 months. The record does not report the observed counts or explain whether completeness was assessed separately at each follow-up. The primary outcome timeframe is described as 6 months after enrollment and randomization, while its definition names both 3 and 6 months. The registry does not clarify the relationship between those descriptions.

Missing-data methods are not reported. There is no information about reasons for missing observations, whether missingness differed between randomized groups, or whether analyses included all randomized participants or only those with observed outcomes. The analysis populations, including any intention-to-treat or per-protocol population, are not reported in the registry record. No sensitivity analyses are described. These omissions prevent assessment of how incomplete follow-up might affect the planned clinical outcome estimates.

These limits do not establish that multiplicity or missing-data methods were absent from the trial; they mean that such methods cannot be characterized from the registry record. Without analysis details and outcome results, readers cannot evaluate how robust any group comparisons are to these issues.

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 lists adverse-event counts separately for caregivers and patients in the usual-care and BCN groups. Serious adverse events are recorded as zero in each group. Other adverse events are also recorded as zero in each group. The listed numbers at risk are 27 for usual-care caregivers, 26 for BCN caregivers, 27 for usual-care patients, and 26 for BCN patients. The death count is one for usual-care patients and zero for each of the other listed groups.

These entries are descriptive counts, not estimates of comparative risk. The registry does not report event timing, definitions beyond serious and other adverse events, attribution to the intervention, or methods for collecting and reviewing events. It also does not state whether the risk periods were identical across groups. Therefore, the counts cannot establish whether an event was caused by BCN, by usual care, or by another factor. Zero recorded events should not be read as proof that no events occurred outside the registry’s recorded categories or observation period.

The death entry is a count in the usual-care patient group; the registry does not report a cause or a relationship to study participation. No safety comparison or inferential analysis is reported. These counts provide a limited description of the safety entries in the registry, rather than a complete account of clinical safety.

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

The study record describes a randomized feasibility trial of BCN added to existing clinical care for Alzheimer’s disease caregiving. The primary endpoint concerns whether data were collected at follow-up, and the study also sought to estimate effects on caregiver burden and patient BPSD at 6 months. Those aims define what the study was designed to examine; they do not establish what the results were.

The registry marks the trial completed and reports an enrollment of 53, but does not report outcome analyses. As a result, the record does not establish the amount of data completeness, the direction or size of any difference in caregiver burden or BPSD, or how precise an estimate might be. It also does not provide a basis for judging whether the app was used as intended or whether outcomes changed over time. Those questions require results and analysis details that are not reported in the registry record.

Randomized allocation is relevant to the intended comparison, but the presence of randomization alone is not evidence of benefit or harm. The intervention group continued clinical care and was assigned the BCN software, while the comparison group received the recorded usual-care arrangement. Any causal interpretation would depend on the actual outcome data, follow-up, adherence, analysis population, and methods. These are not fully described in the record.

The adverse-event entries report zero serious and zero other adverse events across the listed caregiver and patient groups, alongside one death in the usual-care patient group. They do not establish cause, comparative safety, or the absence of unrecorded events. Safety interpretation also depends on definitions, monitoring, and follow-up, which are not described.

Thus, the record supports a limited conclusion about trial structure and intended measurement. It does not support a claim that BCN does or does not improve caregiver or patient outcomes, nor a conclusion about clinical effectiveness or comparative safety. The missing analyses are especially important because the listed outcomes are estimates the study aimed to explore, rather than results presented in the registry.

13. Why This Trial Matters Statistically

I-CARE is statistically interesting because it combines a feasibility question with planned estimation of caregiver and patient outcomes. Feasibility and clinical effect are related but distinct aims. Feasibility asks whether trial procedures, including follow-up data collection, can be carried out. Estimation asks how large an outcome contrast might be and how uncertain that estimate is. A study can contribute to planning future research through feasibility information even when its clinical effect estimates are imprecise, but the registry does not report the observed feasibility counts or outcome estimates needed to assess that contribution.

The study also places the app within an existing care setting. The comparison is not described as app versus no care: both groups had a clinical care arrangement, and the intervention group was assigned BCN in addition to continuing care. This makes clarity about the intervention components, group assignment, adherence, and outcome collection important for interpreting any later comparison. The available record gives some care descriptions but leaves the intervention arm description incomplete.

Repeated measures of caregiver burden and BPSD at baseline, 3 months, and 6 months raise familiar statistical questions about time, correlation, and the choice of comparison. Multiple outcome measures also make prespecification and multiplicity management relevant. The registry lists the measures but does not report the analysis plan or results. This makes the study a useful example of why design information alone cannot substitute for transparent reporting of outcome definitions, analysis populations, missing-data handling, effect estimates, and uncertainty.

The posted record documents a completed randomized study with an enrollment of 53, while leaving its main numerical findings unavailable. That distinction is important for biostatisticians and clinical researchers: completion status and a defined endpoint indicate that a study was conducted, but do not show whether its feasibility aims were met or what its clinical estimates were.

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 is the lack of posted analyses. The registry reports 0 analyses and does not give the observed data-completeness count, outcome values, group comparisons, effect estimates, confidence intervals, or p-values. The clinical aims can therefore be described, but their results cannot be evaluated.

The primary outcome entry has a potential ambiguity: its timeframe refers to 6 months after enrollment and randomization, while its definition refers to data at 3 and 6 months. No explanation is reported. It is also unclear whether the count was intended to be assessed separately at each visit or under another summary definition. The record gives no observed denominators or account of missingness.

The intervention description is incomplete at the end of the recorded arm text. Although it describes BCN software installation and continued clinical care, it does not fully specify the intervention or clarify every component implied by the arm label. Information about app uptake, use, fidelity, and technical support is not reported in the registry record. This limits assessment of what exposure a group comparison would represent.

Details needed to evaluate design and analysis are also unavailable. The registry does not identify who was masked, explain allocation concealment, report baseline group characteristics, or state the analysis population. It does not describe adjustment for baseline values, repeated-measures methods, multiplicity control, missing-data methods, or sensitivity analyses. The listed outcome measures alone cannot fill these gaps.

The safety entries are limited to recorded counts, without event definitions, timing, attribution, or a stated observation period. The cause of the death entry is not reported. These limitations mean that safety counts should not be treated as a comprehensive or comparative safety assessment. Together, the missing outcome results and methodological details sharply limit what can be inferred from the registry record.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Technology Intervention to Support Caregiving for Alzheimer's Disease (I-CARE) — NCT03119259.
• Registry record last updated 2024-09-19; status last verified 2024-08.
• Results first posted 2024-09-19.
• 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

NCT03119259, titled “Technology Intervention to Support Caregiving for Alzheimer's Disease (I-CARE),” is recorded as a completed randomized, parallel-group study sponsored by Indiana University. The registry lists an enrollment of 53, a single-masking designation, and phase NA. The recorded study start is 2019-11-25, and primary completion is 2021-07-31.

The study evaluated feasibility of the Brain Care Notes app alongside existing care and aimed to estimate effects on caregiver burden and patient BPSD at 6 months. The primary outcome is data completeness, defined as a count of enrolled and randomized individuals contributing data at 3 and 6 months. The registry lists 14 outcome measures and 0 analyses; observed outcome results are not reported.

Safety entries record zero serious adverse events and zero other adverse events in each listed group. One death is recorded in the usual-care patient group, and zero in the other groups. The registry does not report causal assessments or comparative safety analyses.