This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Health IT Generated PROs to Improve Outcomes in Cirrhosis (NCT03564626), a randomized na study. The primary outcome is Avoidable Readmissions.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT03564626 |
| Phase | NA |
| Condition | Cirrhosis |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None (open label) |
| Enrollment | 232 |
| Arms in the primary outcome | 2 |
| Lead sponsor | Virginia Commonwealth University |
| Results posted | Yes |
2. Clinical Question
Population
People with cirrhosis, with caregivers also described as part of the trial’s intended approach. The posted outcome groups are labeled “no Dyads”; how participants and caregivers were represented in those results is not reported in the registry record.
Intervention
Health IT using Patient Buddy and EncephalApp, with follow-up described as scheduled or as needed. Participants and caregivers were to use phones to enter medication information, issues and orientation questions daily.
Comparator
Standard-of-care counseling and discharge instructions under local hospital policy, with follow-up calls and visits at baseline and at 30 days.
Primary question
Whether the health-IT approach was associated with a different number of avoidable readmissions within 30 days compared with standard care.
The stated aim was to evaluate prevention of avoidable 30-day readmissions among patients with cirrhosis and their caregivers. The primary outcome is a count, not a direct measure of app use, engagement or patient experience. The registry lists 12 avoidable readmissions for the control group and 23 for the health-IT group. Those counts alone do not establish whether the groups differed beyond chance, or how precisely any difference is estimated: a two-arm comparison is not estimable from the posted results.
The treatment contrast also involves more than an app. The intervention description includes daily entries through Patient Buddy and EncephalApp, messages and calls as needed, and a visit scheduled at 15 days. The control description includes local-policy counseling and discharge instructions, with calls and visits at baseline and at 30 days. Accordingly, the clinical question concerns the described health-IT and follow-up approach as a package, not the isolated effect of either app.
3. Trial Design
The record describes a randomized, parallel-group trial with no masking. The listed enrollment is 232. The study condition is cirrhosis, and the registry status is completed. The recorded start date is 2018-06-22 and the primary completion date is 2024-04-22. The phase is listed as NA. The sponsor is Virginia Commonwealth University.
The control group received standard-of-care counseling and discharge instructions according to local hospital policy. Its description also includes standard-of-care follow-up phone calls and visits at baseline and at 30 days, with follow-up for 30 days. The health-IT group description specifies individual phones loaded with Patient Buddy and EncephalApp. Participants and caregivers were to enter medications, issues and orientation questions daily; messages and phone calls were to be provided as needed per the app. Visits were described at baseline and scheduled at 15 days.
These descriptions identify a contrast in both information technology and follow-up procedures. The record does not establish which elements were delivered to each participant, how often participants used the apps, or whether follow-up varied in practice. The posted outcome arm labels include “no Dyads,” while the intervention description refers to subjects and caregivers. The meaning of that distinction for the outcome population is not reported in the registry record.
Random allocation is intended, in general, to make comparison groups similar on average before treatment, although chance differences can remain. With no masking, participants and study staff were not masked according to the record. Whether outcome assessors were masked is not separately reported in the registry record. Details about the randomization procedure, allocation concealment, trial sites, eligibility criteria, and the statistical analysis plan are also not reported in the registry record. The recorded design therefore supplies the broad comparison and intervention descriptions, but not enough detail to reconstruct how allocation and all trial procedures were implemented.
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
| Stage | Control Group/Not Dyad | Health IT +/- Scheduled Follow Up/ no Dyads |
|---|---|---|
| STARTED | 116 | 116 |
| COMPLETED | 116 | 116 |
| NOT COMPLETED | 0 | 0 |
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.
| Characteristic | Unit | Control Group/ no Dyads | Health IT +/- Scheduled Follow Up/ no Dyads | Total |
|---|---|---|---|---|
| Participants analysed | 116 | 116 | 232 | |
| Age, Continuous | years | 58.42 (SD 8.89) | 59.1 (SD 12.38) | 58.8 (SD 10.6) |
| Sex: Female, Male: Female | Participants | 40 | 58 | 98 |
| Sex: Female, Male: Male | Participants | 76 | 58 | 134 |
| Ethnicity (NIH/OMB): Hispanic or Latino | Participants | 6 | 6 | 12 |
| Ethnicity (NIH/OMB): Not Hispanic or Latino | Participants | 110 | 110 | 220 |
| Ethnicity (NIH/OMB): Unknown or Not Reported | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): American Indian or Alaska Native | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): Asian | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): Native Hawaiian or Other Pacific Islander | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): Black or African American | Participants | 36 | 53 | 89 |
| Race (NIH/OMB): White | Participants | 80 | 63 | 143 |
| Race (NIH/OMB): More than one race | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): Unknown or Not Reported | Participants | 0 | 0 | 0 |
| Region of Enrollment: United States | participants | 116 | 116 | 232 |
6. Primary Endpoint
The registry lists the primary outcome as Avoidable Readmissions, measured over 30 days.
| Element | Registry entry |
|---|---|
| Result type | Number |
| Measure of spread | not applicable |
| Unit | Number of avoidable readmissions |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is the number of avoidable readmissions within 30 days. A readmission count describes how many qualifying events were recorded over the specified period; it is different from a binary outcome that classifies each participant only as having or not having a readmission. The registry gives the unit as “Number of avoidable readmissions” and describes the timeframe as 30 days.
Readmission is used in the stated trial aim because it captures a hospital-use outcome that the health-IT approach was intended to help prevent. The adjective “avoidable” matters: it implies that some criterion was used to distinguish these events, but the operational definition and adjudication process are not reported in the registry record. The record also does not clarify whether more than one readmission for an individual could contribute to the count.
The posted counts are 12 for the control group and 23 for the health-IT group. Counts alone do not show the distribution of events across participants, the number of participants with any readmission, or a rate adjusted for follow-up or other factors. Nor do they provide an uncertainty interval. The registry indicates that values were missing or not estimable for a two-arm comparison of this count outcome, so the posted counts should not be treated as a complete inferential comparison.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry reports 12 avoidable readmissions for the control group and 23 for the health-IT group, with 116 listed in each outcome group. These are arm-level event counts. They are not accompanied by a comparison estimate, confidence interval or inferential test in the posted results. The registry specifically indicates that values were missing or not estimable for a two-arm comparison of this count outcome.
That limitation governs interpretation. The observed counts describe what was recorded, but do not quantify the uncertainty around a treatment contrast. A confidence interval, in general, expresses the range of effect estimates compatible with the data and the chosen statistical procedure. Without a comparison estimate and its uncertainty, it is not possible to assess how precise a contrast would be or whether the data are compatible with a range of possible effects. A p-value is also not reported, and a p-value would not by itself describe the magnitude or practical importance of a difference.
The counts should not be read as rates or risks. Although the outcome groups each list 116, a count divided by a group size would be a separate calculation and would not necessarily match the intended analysis of a count endpoint. The registry does not provide participant-level event distributions, adjusted estimates, or enough detail to determine whether repeat events were counted. The recorded counts therefore cannot show whether one or more participants accounted for multiple events, or whether the groups differed in the number of participants experiencing any readmission.
The direction of the raw counts does not establish the direction of a reliable treatment effect. A count comparison needs an appropriate estimate and an assessment of uncertainty; where events are sparse, unevenly distributed, or more variable than a simple model assumes, that uncertainty can be substantial. The registry record does not report a model, adjustment variables, or an analysis population for the primary outcome. No recalculated comparison is available in the posted results, and no discrepancy is flagged.
The trial is recorded as completed, but completion does not make an outcome comparison estimable. The reported readmission counts remain descriptive in the absence of a posted analysis. They should not be interpreted as evidence that the health-IT approach did or did not change readmissions. The record also does not report secondary outcomes, so the primary count is the only posted outcome measure identified in the record.
9. Statistical Methods Explained
The primary outcome is a count. Statistical analysis of counts differs from analysis of a continuous measurement such as a score, because counts are nonnegative event totals and can be highly unevenly distributed. A comparison may be based on a model for counts, a rate, or another prespecified summary. The registry record does not report which method, if any, was used for the posted outcome.
Why does the outcome type matter?
A count can include repeated events for the same person, depending on the endpoint definition. The registry describes the outcome as the number of avoidable readmissions, but does not report the rules for counting repeat admissions or the process used to decide whether an admission was avoidable. Those details affect what the count represents and what statistical model is appropriate.
In general, a Poisson model is one option for count outcomes. Its basic form assumes that the event count’s variance is tied to its mean. When the data vary more than that assumption allows, the counts are over-dispersed. Ignoring over-dispersion can make uncertainty appear smaller than it is. Other approaches can allow extra variability, but the registry does not identify a model or indicate whether over-dispersion was assessed. It is therefore not possible to attribute a particular modeling choice or diagnostic to this trial.
What can an arm count establish?
An arm-level count is descriptive. To support a comparative conclusion, analysis generally needs a defined estimand, such as a difference or ratio in event counts or rates, together with an uncertainty measure. The choice depends on the question, follow-up, event distribution and assumptions. A confidence interval expresses uncertainty about an estimate under the selected method; its width reflects precision. The posted record does not give a comparative estimate or interval, and it states that values were missing or not estimable for a two-arm comparison. The counts of 12 and 23 therefore do not supply the inferential information that such an analysis would provide.
The listed group sizes are 116 in each arm, but the record does not report the participant-level event data or an exposure-adjusted rate. Equal listed group sizes do not, by themselves, establish equal observation time for every participant, complete outcome ascertainment, or an appropriate denominator for a rate analysis. The primary endpoint specifies 30 days, but details about how that window was applied to participants who may have had incomplete follow-up are not reported in the registry record.
What does the record say about the analysis plan?
The record lists no posted analyses for the outcome and does not give a statistical analysis plan. The model, covariates, handling of repeat events, missing outcome data, and analysis population are not reported in the registry record. It is not possible to determine whether the primary analysis was adjusted or unadjusted, whether it followed randomized assignment, or whether any alternative analyses were planned.
The record’s planning statement describes illustrations based on the observed effect, with a two-sided significance level of 0.05, equal allocation between the two arms and a normal approximation; the effect to detect is described as a stated fraction of the observed effect. Such sample-size calculations are planning illustrations, not observed power and not evidence about the trial’s ability to detect an effect. No particular planning result should be interpreted as a result of the trial. The registry does not report the trial’s own sample-size justification, so the planning statement should not be mistaken for that justification.
In general, a significance level sets a decision threshold for a specified testing procedure; it does not express the probability that a treatment effect is true. A sample-size calculation also depends on assumptions about the outcome and the effect to detect. For a count endpoint, assumptions about event frequency and variability can be consequential. These general considerations explain why a planning illustration cannot replace a prespecified analysis of the observed data.
Why are the posted numbers not enough for recalculation?
A simple recalculation from summary statistics requires enough information to reproduce a defined comparison and estimate its uncertainty. The posted results provide arm counts and group sizes, but the registry marks a two-arm comparison as missing or not estimable. The record does not provide event-level data, a variance estimate, or a model-based result. Reconstructing an inferential comparison from these arm totals would require assumptions that are not documented, so no such estimate should be inferred from the counts alone.
10. Multiplicity, Missing Data and Analysis Populations
The registry identifies one primary outcome measure and reports no secondary outcomes. It lists no analyses. The record does not describe a multiplicity strategy, such as how testing would be handled across multiple outcomes or repeated analyses. Since no secondary outcomes are posted, the record does not support an account of how multiple outcome tests were controlled or interpreted.
Missing-data methods and completeness of outcome ascertainment are not reported in the registry record. The primary endpoint has a 30-day timeframe, but the record does not state how incomplete follow-up, unavailable readmission information, or other missing outcome data were handled. It also does not report whether participants were analyzed according to randomized assignment, whether exclusions occurred, or how the “no Dyads” outcome labels relate to the population included in the analysis. The analysis population is therefore not defined in the posted results.
These omissions matter because missing outcomes can affect a count comparison, especially if missingness differs between groups or is related to readmission. General methods for addressing missing data require assumptions about why values are missing; those assumptions cannot be assessed from the posted record. No imputation method, sensitivity analysis, interim analysis, or other planned analysis is reported in the registry record. Without these details, the reported arm counts cannot be used to determine whether missingness or analysis choices affected the result.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Control Group/ no Dyads | 0 | 116 | 0.0% |
| Health IT +/- Scheduled Follow Up/ no Dyads | 0 | 116 | 0.0% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Control Group/ no Dyads | 0 of 116 |
| Health IT +/- Scheduled Follow Up/ no Dyads | 0 of 116 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Control Group/ no Dyads | 0 of 116 |
| Health IT +/- Scheduled Follow Up/ no Dyads | 0 of 116 |
The posted safety groups list 116 participants at risk in the control group and 116 in the health-IT group. Serious adverse events affected 0 in the control group and 0 in the health-IT group. Other adverse events affected 0 in each group, and deaths are listed as 0 in each group. No serious adverse-event terms are listed.
These figures describe the safety entries in the registry. They should be read as reported event counts, not as proof that no adverse events occurred outside the record’s collection or reporting process. The safety observation period, event definitions, collection procedures and any thresholds for reporting are not reported in the registry record. The primary endpoint’s 30-day timeframe should not automatically be assumed to define the safety follow-up period.
Zero reported events also do not quantify the uncertainty around the possibility of uncommon events. The registry’s safety figures do not provide comparative estimates or uncertainty intervals. They therefore offer a limited descriptive account of recorded serious events, other events and deaths, rather than a complete characterization of safety.
12. What the Result Does, and Does Not, Mean
The posted results establish that the registry records 12 avoidable readmissions in the control group and 23 in the health-IT group for the primary endpoint. They also show listed outcome-group sizes of 116 in each arm. The record does not provide an estimable two-arm comparison, an uncertainty interval, or a reported statistical test. As a result, the counts are descriptive rather than a basis for deciding whether the intervention changed readmissions.
The result does not isolate the effect of an app. The intervention description combines daily phone-based entries using Patient Buddy and EncephalApp with messages, calls as needed, and a scheduled visit. The control description includes local-policy counseling and discharge instructions, as well as follow-up calls and visits. Any interpretation of the planned contrast must therefore concern those described care approaches as a whole. The record does not report implementation or adherence details that would allow separation of the components.
The outcome is an event count, not a measure of symptoms, app engagement, caregiver experience or quality of life. Those outcomes are not posted as secondary measures in the record. Nor do the counts reveal how events were distributed across individuals or whether any individual had repeated readmissions. The operational definition of “avoidable” is not reported in the registry record.
The trial is marked completed, but the posted results do not establish a treatment effect. They do not establish that health IT reduced readmissions, increased readmissions, or had no effect. A non-estimable comparison is not evidence of equivalence or evidence that the groups were the same. Further interpretation would require a clearly defined analysis, suitable estimates with uncertainty, and information about outcome ascertainment and missing data that are not reported in the registry record.
13. Why This Trial Matters Statistically
This trial raises an important statistical distinction between reporting an outcome and estimating a treatment contrast. The registry gives arm-level counts for avoidable readmissions, but marks values for a two-arm comparison as missing or not estimable. The counts are informative as a description of the posted record; they do not, by themselves, answer the comparative question stated in the trial aim.
That distinction is particularly relevant for count outcomes. A count can reflect both how many participants experience an event and how events are distributed among them. If repeat events are possible, the same total count may arise from different participant-level patterns. Statistical methods must match the outcome definition and its variability. The record does not report the event-count model or details needed to evaluate its assumptions, so readers cannot assess whether a particular count analysis would be appropriate.
The intervention also illustrates the challenge of interpreting multi-component approaches. The record describes two apps alongside messaging, calls and differing follow-up arrangements. A comparison of these groups would address the combined approach, not identify which element accounts for an outcome pattern. Implementation and use information would help describe how the intervention was delivered, but such information is not reported in the registry record.
Finally, transparent reporting of an estimable effect, its precision, the analysis population and missing-data handling is central to evaluating clinical trial results. Those details are not available for this primary outcome. The record’s completed status and posted counts cannot substitute for a comparative analysis. This makes the distinction between descriptive data and inferential evidence especially important when readers assess what the trial record can support.
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 primary limitation for interpreting the result is that the registry reports arm counts but no estimable two-arm comparison. No effect estimate, confidence interval, or statistical test is posted. Without those elements, the precision and uncertainty of a comparative result cannot be evaluated.
The endpoint definition is also incomplete for independent interpretation. The record gives the phrase “avoidable readmissions within 30 days,” but does not report how avoidability was judged, whether reviewers used a standardized process, or how repeat readmissions were counted. The posted group sizes and totals do not reveal the participant-level distribution of events. Outcome ascertainment procedures and completeness are not reported.
The analysis details are limited. The statistical method, analysis population, adjustment strategy, handling of missing outcome data, and any sensitivity analyses are not reported in the registry record. The posted outcome groups carry “no Dyads” labels, while the intervention description refers to subjects and caregivers; the impact of that distinction on the analyzed population is unclear. The record also does not provide adherence or implementation details for the apps and follow-up components.
Finally, the trial compares a package that includes health IT and follow-up features with standard care that has its own counseling and follow-up procedures. The record does not enable attribution of any outcome pattern to a particular component. These limitations do not alter the counts as posted, but they restrict what can be concluded from them.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Health IT Generated PROs to Improve Outcomes in Cirrhosis — NCT03564626.
- Registry record last updated 2025-09-17; status last verified 2025-08.
- Results first posted 2025-09-17.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
The registry record for NCT03564626, titled “Health IT Generated PROs to Improve Outcomes in Cirrhosis,” describes a randomized, parallel-group trial with no masking. It lists cirrhosis as the condition, enrollment of 232, and Virginia Commonwealth University as sponsor. The recorded study status is completed; the start date is 2018-06-22 and the primary completion date is 2024-04-22.
The primary outcome is the number of avoidable readmissions within 30 days. The posted counts are 12 for the control group and 23 for the health-IT group, with 116 listed in each outcome group. The registry indicates that values are missing or not estimable for a two-arm comparison. No secondary outcomes or analyses are listed. Serious adverse events, other adverse events and deaths are each recorded as 0 in both groups. The record does not report the statistical analysis plan, missing-data methods, analysis population, or detailed endpoint adjudication procedures.