Louisiana State University Health Care Sciences Division (LSU HSCD) Tele-Health Projects: Adult CHF Patient Population: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Louisiana State University Health Care Sciences Division (LSU HSCD) Tele-Health Projects: Adult CHF Patient Population (NCT01148563), a randomized na study. The primary outcome is Events Per Person-year.

2. Clinical Question

3. Trial Design

The registry describes a completed randomized trial sponsored by Pennington Biomedical Research Center. Its listed enrollment is 136, and the study dates are March 2009 to December 2010 for start and primary completion. The phase is listed as NA. The registry identifies two groups: Standard Care and Tele-health Monitoring Group. Standard care continued regular LSU HCSD disease-management care for heart failure; the monitoring group continued physician standard care and added tele-health monitoring.

There is an internal design-label inconsistency in the record: allocation is listed as randomized, while the model is listed as single-group despite the two arm descriptions and posted results for both groups. The registry record does not explain this inconsistency. Masking is listed as none, so the record does not describe masking of participants or study personnel. It does not report whether outcome assessors were masked.

The monitoring description names weight, blood pressure, pulse oximetry, pulse rate, and patient responses to disease-specific questions. Its text is incomplete, so details such as the monitoring schedule, how readings were handled, and what actions followed a reading are not reported in the registry record. The record describes the primary analysis population as intention-to-treat and says missing outcome data were imputed as necessary. The imputation method and the analysis population for the secondary outcome are not reported in the registry record.

The main posted outcome is events per person-year over 12 months, defined as emergency department visits plus inpatient stays divided by person-year. A cost outcome, U.S. dollars per person-year, is also posted for 12 months. These are distinct measures and should be interpreted separately. Enrollment and arm counts describe the recorded trial groups; they do not by themselves establish how missing observations were handled in each analysis.

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: Of the original 137 subjects recruited, one subject in the treatment group withdrew from the study prior to having home monitoring equipment installed, leaving a total of 136 subjects.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Used an Intention to Treat (ITT) Analysis. Missing outcome data were imputed as necessary.

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 Events Per Person-year, measured over 12 months.

The primary endpoint is the mean number of events per person-year during 12 months. An event is defined as an emergency department visit or an inpatient stay, with the two types of healthcare use combined in the numerator and person-time used as the denominator. The posted arm values are means with 95% confidence intervals.

A rate per person-time can describe recurrent events while accounting for the time people contribute to follow-up. That is useful when participants may contribute different amounts of observed time. Here, however, the combined endpoint does not distinguish emergency department visits from inpatient stays. A change in the combined rate could arise from either component or both; component-specific results are not reported in the registry record.

General: combining related forms of healthcare use can provide a single summary measure, but it can conceal differences in clinical meaning and frequency between its components. The endpoint is therefore best read as a measure of the combined event rate defined by the registry, not as a measure of every aspect of heart failure status, quality of life, or patient experience. The registry's stated study purpose includes several such broader outcomes, but results for those outcomes are not reported in the posted results.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

General: a mean summarizes the average value of a continuous measure, while a confidence interval describes the uncertainty around an estimated quantity under the statistical method used. The primary endpoint is expressed as events per person-year, and each arm has a posted mean and 95% confidence interval. The interval gives a range of values compatible with the data and method; it is not a range containing 95% of individual participants' outcomes.

Why use a rate per person-year?

General: a rate uses person-time in its denominator, which can make comparisons more meaningful when people contribute different amounts of time under observation. A rate is not the same as the proportion of people who experienced at least one event. Because this endpoint counts emergency department visits and inpatient stays together, a participant can contribute more than one event. The registry does not give the separate component counts or the distribution of repeat events.

What does the primary comparison mean?

The registry reports a risk ratio on a log scale, with the comparison estimated by bootstrapping. On a log scale, the null value for a ratio comparison is zero; on the original ratio scale, the null is one. The posted interval runs from -0.65 to 0.05 on the log scale and includes zero. That means the sponsor's analysis does not exclude the null comparison at the interval's stated confidence level. The reported p-value, 0.094, likewise does not cross the conventional 0.05 threshold. Neither fact establishes equivalence or proves that there is no difference.

General: bootstrapping estimates uncertainty by repeatedly resampling data according to a defined procedure. It can be useful when the sampling distribution is not well represented by a simple theoretical approximation. The registry names bootstrapping but does not report enough procedural detail to assess the exact implementation. The sponsor's reported analysis is the trial's own comparison and should be distinguished from unadjusted recalculations based on posted summaries.

What is the recalculated cost comparison?

The recalculated cost comparison is an unadjusted rate-ratio analysis based on the posted arm means and exposure information. It reports a rate ratio of 0.696, with a 95% confidence interval from 0.693 to 0.699 and p < 0.001. This calculation is not the sponsor's reported analysis. The sponsor reports a bootstrapped log risk-ratio estimate of .70, a 95% confidence interval from 0.45 to 1.08, and p = 0.105. The two analyses therefore yield different uncertainty and significance results. The registry does not explain the difference. The recalculation should be read as a simple unadjusted analysis of posted summary statistics, not as a replacement for the sponsor's analysis.

General: over-dispersion occurs when observed event counts vary more than a statistical model assumes. If a count-rate analysis assumes less variability than is present, its uncertainty can be understated and its confidence interval too narrow. The recalculated cost comparison uses a Poisson approach, which relies on model assumptions about event variation. The registry record does not report a dispersion assessment for this recalculation, nor does it report enough detail to determine whether the sponsor's analysis addressed over-dispersion in the same way.

How should planning illustrations be read?

The posted planning illustration uses a two-sided significance level of 0.05, equal allocation, and a normal approximation, with effects set to stated fractions of the observed effect. It reports participants per arm for 80% power and participants per arm for 90% power; each displayed value is 1 for each listed effect fraction. These are planning illustrations computed from the observed effect, not the trial's original design calculations, and not observed power. The registry record does not report the trial's sample-size justification or design assumptions. The unusually small displayed planning values should not be treated as a reliable recruitment target without examination of the underlying model and inputs. They are not evidence that the trial had adequate power.

General: observed power is not a useful substitute for reading an estimate and its confidence interval; it is largely determined by the observed effect and its uncertainty. Planning calculations can help illustrate how assumptions affect a future study, but a calculation based on an observed effect can be unstable, especially when the estimate is uncertain. No original sample-size calculation, interim analysis plan, or detailed statistical analysis plan is reported in the registry record.

10. Analyses Reported by the Sponsor

The sponsor's primary analysis is a bootstrapped comparison reported as a log risk ratio of -0.30, with a 95% confidence interval from -0.65 to 0.05 and p = 0.094. The interval includes the null value on the log scale, so the result does not provide conventional statistical evidence of a difference at the 0.05 level. The posted arm means are lower for the monitoring group, but the comparative analysis carries the inferential result.

For the cost endpoint, the sponsor reports a bootstrapped log risk-ratio estimate of .70, with a 95% confidence interval from 0.45 to 1.08 and p = 0.105. The unadjusted recalculation based on posted summary statistics instead reports a rate ratio of 0.696, a 95% confidence interval from 0.693 to 0.699, and p < 0.001. The contrast indicates that the analyses do not produce the same uncertainty assessment. The registry does not explain the discrepancy in the cost comparison. The recalculation is not a substitute for the sponsor's analysis and does not establish which modeling choices are appropriate; it is a limited, unadjusted re-analysis of the posted summaries. The registry record does not report enough information to reconcile the methods fully.

11. Multiplicity, Missing Data and Analysis Populations

The primary outcome population is described as intention-to-treat, with missing outcome data imputed as necessary. This indicates that the record intended to include participants according to their randomized group in the primary analysis, while addressing missing outcome values through imputation. The registry record does not report how much outcome data were missing, which imputation method was used, what assumptions were made about missingness, or whether sensitivity analyses tested those assumptions.

General: imputation replaces missing values with estimates under a model and assumptions. Different assumptions can produce different results, especially if missingness is related to outcomes or differs between groups. An intention-to-treat label does not by itself resolve these issues; the handling of missing data and the analysis population still need to be understood. The registry record does not report whether the secondary cost analysis used the same population or missing-data approach.

The registry posts a primary endpoint and a secondary cost endpoint, alongside a broader study purpose that names additional areas. It does not report a multiplicity strategy, such as adjustment of significance levels or a formal testing hierarchy. General: when multiple outcomes or comparisons are tested, the chance of at least one nominally significant result can increase. Unadjusted secondary p-values should therefore be treated as exploratory unless the analysis plan establishes otherwise. The record does not report an interim analysis plan or a detailed statistical analysis plan.

The registry labels the study randomized but also lists a single-group model while describing two arms. It does not explain that inconsistency. Masking is listed as none, and the record does not describe masked outcome assessment. These design and reporting details limit how fully the reported analyses can be evaluated. The sample-size justification and its assumptions are not reported in the registry record; the posted planning illustration is not evidence of the trial's original design rationale.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: any er visit or inpatient stay (58.0%).

The registry lists serious adverse events affecting 40 participants in Standard Care and 31 in Tele-health Monitoring Group. The most frequent serious-event term listed is “Any ER visit or Inpatient stay,” reported as 58.0%. The event term overlaps conceptually with the primary endpoint, which also combines emergency department visits and inpatient stays, but the registry does not provide enough detail to determine how the serious-event coding relates to the endpoint counts.

The record lists zero other adverse events in each group. Death counts are not reported in the registry record. These figures describe the posted safety categories; they do not establish that the groups differ in safety. The record does not give a comparative safety analysis, confidence intervals, or details about event severity beyond the serious-event classification. The serious-event figures also should not be interpreted as the number of unique people with any event beyond the group counts as recorded.

General: adverse-event counts depend on definitions, collection procedures, follow-up, and whether the unit is events or participants affected. The record labels the figures as participants affected for serious events, but does not report a detailed event-level breakdown. These limitations make simple between-group comparisons inappropriate from the posted figures alone.

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

The primary results describe the combined rate of emergency department visits and inpatient stays per person-year during 12 months. The monitoring arm has a lower posted mean than the standard-care arm, while the sponsor's comparative estimate has a 95% confidence interval that includes the null and a p-value of 0.094. The data therefore do not establish a statistically significant difference under the sponsor's reported primary analysis. They also do not establish that the groups are equivalent or that no meaningful difference exists.

General: a non-significant result can occur when an effect is small, when data are variable, or when the estimate is imprecise. A confidence interval is more informative than a threshold alone because it indicates the range of effects compatible with the analysis. The registry does not report a threshold for a clinically important difference, so statistical uncertainty cannot be translated into a judgment about clinical importance.

The secondary cost outcome has a lower posted point estimate for the monitoring group, but the sponsor's analysis and the unadjusted recalculation differ substantially in their uncertainty assessments. That secondary result should not be taken as decisive evidence about cost benefits. It is also not a direct measure of the other aims named in the study purpose, such as quality of life or patient satisfaction.

These results apply to the population, intervention, comparator, endpoint definitions, follow-up period, and analysis methods recorded for this trial. Incomplete intervention details, unspecified missing-data procedures, absent multiplicity information, and the design-label inconsistency constrain interpretation. The results do not by themselves establish what would happen in other settings or under a different monitoring program.

14. Why This Trial Matters Statistically

This trial illustrates why a reported arm-level pattern and a formal comparison answer different questions. The monitoring arm's posted primary mean is lower, but the sponsor's comparative interval includes the null. Reading only the direction of the means could overstate what the trial establishes; reading only a thresholded p-value could hide the range of effects still compatible with the data. Both the estimate and its uncertainty matter.

The record also provides an example of how model choice can affect a secondary analysis. The sponsor's cost comparison and the unadjusted recalculation lead to notably different uncertainty assessments despite point estimates in a similar direction. This is a reminder that a recalculation from summary statistics is not automatically a better analysis. Assumptions about the outcome distribution, exposure, variability, and analysis population matter, and the registry does not give enough detail to reconcile the difference fully.

The trial's rate endpoint combines two forms of healthcare use and expresses them per person-year. That approach highlights recurrent events and exposure time, while leaving the component-specific pattern unclear. The registry also shows why missing-data methods, multiplicity plans, complete intervention descriptions, and consistency in design labels are important for interpreting clinical trial records. Their absence does not invalidate the posted numbers, but it limits the strength and scope of conclusions that can be drawn from them.

15. 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 specific to this record constrain interpretation. The registry lists randomized allocation but labels the model single-group while describing and analyzing two groups. It does not explain the inconsistency. Masking is listed as none, and whether outcome assessors were masked is not reported in the registry record.

The tele-health intervention description is incomplete. It lists measures collected by the monitor but does not provide the full protocol, monitoring frequency, response procedures, or other implementation details. As a result, the intervention cannot be characterized fully from the registry description. The primary endpoint combines emergency department visits and inpatient stays; results for those components separately are not reported in the registry record.

The record states that missing primary outcome data were imputed as necessary, but does not report the extent of missingness, imputation method, assumptions, or sensitivity analyses. It does not report a multiplicity plan, an interim analysis plan, a detailed statistical analysis plan, or the trial's sample-size justification. These omissions limit assessment of whether the analyses followed a prespecified strategy and how robust the findings are to analytic choices.

The cost comparison differs between the sponsor's reported analysis and the unadjusted recalculation. The registry does not explain the difference, and the posted details are insufficient to assess all modeling assumptions. The serious adverse event record does not provide a comparative analysis or detailed event-level information, and deaths are not reported. Finally, results for several outcomes named in the study purpose are not reported in the registry record. These limitations mean the posted results address only part of the intended research question.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Louisiana State University Health Care Sciences Division (LSU HSCD) Tele-Health Projects: Adult CHF Patient Population — NCT01148563.
• Registry record last updated 2024-01-25; status last verified 2024-01.
• Results first posted 2017-12-19.
• 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.

19. Record Summary

The registry describes a completed randomized tele-health study in an adult congestive heart failure population, with standard care compared with standard care plus tele-health monitoring. The primary outcome is the 12-month rate of emergency department visits plus inpatient stays per person-year. The sponsor's primary comparison is a bootstrapped log risk ratio of -0.30, with a 95% confidence interval from -0.65 to 0.05 and p = 0.094. The posted arm means are lower in the monitoring group, but the sponsor's comparative interval includes the null.

A secondary cost outcome is posted, and the sponsor analysis differs from an unadjusted recalculation in its uncertainty assessment. Interpretation is limited by incomplete intervention and missing-data details, absent multiplicity information, and an unexplained inconsistency between the randomized allocation label and the single-group model label. Results for several broader outcomes named in the study purpose are not reported in the registry record.