Propranolol Versus Prednisolone for Treatment of Symptomatic Hemangiomas: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Propranolol Versus Prednisolone for Treatment of Symptomatic Hemangiomas (NCT00967226), a randomized phase 2 study. The primary outcome is Decrease in Size of Hemangioma (Length x Width) in Square mm.

2. Clinical Question

The trial asked whether propranolol and prednisolone differed in their effects on symptomatic hemangiomas in infants. The registry’s primary outcome describes change in lesion surface area, measured using the lesion’s outer-margin length multiplied by width at baseline and again at 4 months. It says 5-month measurements were used as surrogate data when 4-month data were unavailable. The outcome title describes a decrease in size in square mm, while its description calls the primary outcome proportional change. That difference in wording makes the precise scale of the reported values less clear.

The registry also lists tolerability and several adverse-event categories as secondary outcomes, including serious adverse events and events involving growth and development, respiratory, allergy/immunology, dermatologic, endocrinologic, and gastrointestinal systems. The central statistical question is not whether either treatment should be used, but what the posted group estimates and comparison test can establish about the measured outcome. Interpretation depends on the limited outcome data available, the width of the group confidence intervals, and the absence of a confidence interval for the reported between-group difference. The trial record does not report a treatment-effect threshold for judging whether a difference would be important.

3. Trial Design

This was a phase 2, randomized, single-masked, parallel-group trial comparing propranolol with prednisolone. The registry lists enrollment as 19 and the trial status as terminated. The study start is recorded as 2009-07, and primary completion as 2013-01. The registry record does not state why the trial was terminated, who was masked, or what led to the recorded enrollment. Those details should not be inferred from the status or design labels.

The primary outcome was assessed at 4-5 months after therapy began. The record says the planned measure was proportional change in total surface area, based on lesion length multiplied by width at baseline and at 4 months, using 5-month data if 4-month data were unavailable. Data at 4 or 5 months were available for 9/11 propranolol participants and 6/8 prednisolone participants because of missed appointments. The record also reports that 90% (138/154) of study appointments were completed. Appointment completion is not the same as having the primary endpoint available for every participant.

The listed outcome summaries include 9 propranolol participants and 6 prednisolone participants. These are the group counts attached to the posted endpoint summaries, while the availability statement gives the numbers with data relative to each group’s larger count. The registry reports a two-sided t-test for the primary comparison. It does not report enough detail to establish whether the analysis followed an intention-to-treat approach, how missing measurements were handled in the statistical test, or whether the test adjusted for any other variables. The statistical analysis plan and a sample-size justification are not reported in the registry record.

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: Data available at 4 or 5 months for only 9/11 propranolol participants and for 6/8 prednisolone participants due to missed appointments. Overall, 90% (138/154) study appointments were completed.

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.

Population: Infants 2 weeks - 6 months of age with symptomatic, proliferating, infantile hemangiomas

6. Primary Endpoint

The registry lists the primary outcome as Decrease in Size of Hemangioma (Length x Width) in Square mm, measured over 4-5 months after initiating therapy.

The primary endpoint concerns change in the measured size of a hemangioma. The registry describes measuring the lesion’s outer-margin length and width at baseline and again at 4 months, with a 5-month measurement used when a 4-month value was unavailable. Multiplying length by width gives a simple surface-area proxy. Comparing measurements over time is intended to capture whether the lesion’s measured extent changed during the treatment period.

The endpoint is useful as a quantitative outcome because it turns a lesion’s measured dimensions into a continuous value that can be summarized by a mean and compared between groups. The registry describes the target as proportional change in total surface area, which can express change relative to the baseline measure. However, the endpoint title identifies a decrease in size in square mm. The record does not resolve this difference between the title and description, so the precise interpretation of the reported values’ scale is uncertain. The posted group means are 0.57 for propranolol and 0.63 for prednisolone, with 95% confidence intervals of 0.34 to 0.8 and 0.14 to 1.11, respectively.

The endpoint reflects the measurement specified in the registry; it does not by itself describe symptoms, function, appearance, or tolerability. Those are different aspects of a clinical outcome. The registry’s primary measurement also relies on follow-up observations, and data were unavailable at the designated 4- or 5-month time for some participants in each group. The record does not describe whether the length-by-width method was validated against another measure or how measurement variability was addressed.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry summarizes the primary endpoint as a continuous outcome using group means and 95% confidence intervals. A continuous outcome can take values across a scale rather than being restricted to categories. A mean describes the center of the observed values, while a confidence interval communicates uncertainty about the estimate. In general, a confidence interval is a range generated by a statistical procedure that, under its assumptions, has a stated long-run coverage property. It is not a guarantee that the true value lies in that particular interval, and it does not describe the spread of individual participant measurements.

What do the group confidence intervals describe?

The propranolol mean is 0.57, with a 95% confidence interval from 0.34 to 0.8. The prednisolone mean is 0.63, with a 95% confidence interval from 0.14 to 1.11. These ranges describe uncertainty around each group’s mean, not the range of participant-level outcomes. Their widths reflect that the estimates are not precise, especially for prednisolone. Comparing the amount of overlap between two separate confidence intervals is not a substitute for a confidence interval or test for the difference itself. The registry does not provide a confidence interval for the reported mean difference.

What does the two-sided t-test assess?

The registry reports a two-sided t-test, a mean difference (net) of 0.06, and p = 0.77. In general, a t-test evaluates evidence against a null hypothesis about a difference in means, using a test statistic that relates the estimated difference to its uncertainty. A two-sided test allows departures in either direction. The p-value is calculated under the null hypothesis and the test’s assumptions; it does not quantify the probability that the null is correct, the probability that a result will replicate, or the size of an effect.

A p-value that is not small does not demonstrate equivalence or no effect. Equivalence requires a prespecified margin defining differences small enough to count as equivalent, along with an analysis designed to assess that margin. The margin and an equivalence analysis are not reported in the registry record. Nor does the p-value show whether a possible difference would matter clinically. That requires a justified threshold or context that the record does not provide.

Why does the contrast need its own uncertainty estimate?

The direct scientific comparison is the difference between the group means. A confidence interval for that contrast would show the range of differences compatible with the data under the analysis assumptions. The registry reports the difference and p-value but no contrast interval. A p-value provides a test result, but it does not provide the same information as an interval estimate: the interval would indicate both direction and precision across a range of effect sizes. The posted group intervals cannot be used as a replacement for that missing contrast interval without making additional assumptions or calculations, which are not reported in the registry record.

What does over-dispersion mean in general?

Over-dispersion is a general issue that arises when observed variability is greater than a statistical model expects. It can make standard errors too small if a model fails to account for the extra variability, leading to confidence intervals that are too narrow and tests that appear more conclusive than warranted. This concept is useful when outcomes are counts or otherwise have variance assumptions that may not fit the data. The primary outcome in this record is described as a continuous measure, and no over-dispersion model or diagnosis is reported. It would therefore be inappropriate to claim that over-dispersion affected this trial’s reported analysis.

How should the missing measurements affect interpretation?

The registry reports primary-outcome data for 9/11 propranolol participants and 6/8 prednisolone participants, attributing unavailable 4- or 5-month values to missed appointments. In general, an analysis based only on observed follow-up can be biased if the chance of having a measurement is related to the outcome or to other factors associated with it. Missingness can also reduce precision. The record does not state the assumptions made about missing data, whether any values were imputed, or whether sensitivity analyses were performed. Appointment completion of 90% (138/154) is a measure of visit completion, not proof that missing endpoint observations cannot affect the results.

The registry reports an enrollment of 19 and endpoint summaries for the participants with available measurements. It does not state whether the primary comparison included all randomized participants, how participants were assigned to the analysis population, or whether the analysis accounted for the difference between enrolled participants and those with observed endpoint data. Those analysis-population details matter because randomized comparisons are most readily interpreted when the handling of every randomized participant is clear.

What can planning illustrations tell readers?

In general, sample-size planning illustrations use assumptions about an effect, variability, significance level, allocation, and a target power to explore how much information a future study might need. The registry’s stated planning assumptions refer to a two-sided significance level of 0.05, equal allocation, a normal approximation, and a target effect expressed as a fraction of the observed effect. Such illustrations are planning exercises, not observed power and not evidence that this trial was designed around those assumptions. The trial’s own sample-size justification is not reported in the registry record. A planning result based on an observed effect is particularly sensitive to uncertainty in that effect, which is substantial when the available endpoint data are limited.

Overall, the statistical summaries support a cautious description: the registry reports group means, separate confidence intervals, and a two-sided test of the group comparison. They do not supply a confidence interval for the contrast, a full account of missing-data methods, or a stated threshold for clinical importance. These gaps limit how far the numerical comparison can be taken.

10. Analyses Reported by the Sponsor

The sponsor-reported analysis is a two-sided t-test, with a mean difference (net) of 0.06 and p = 0.77. The record does not provide a confidence interval for that contrast or a separately stated recalculated contrast estimate. The posted arm means and confidence intervals are descriptive summaries, but they do not by themselves supply the missing interval for the between-group difference.

No discrepancy is flagged between the sponsor’s analysis and another analysis in the registry record. The sponsor’s reported result should therefore be read as the available inferential comparison, rather than replaced by an independently asserted result. In general, a recalculation from posted summary statistics would be an unadjusted calculation and could differ from a sponsor’s analysis if the sponsor used different participant-level data or methods. The registry does not report enough analysis detail to assess such possibilities, and no alternative test result is stated.

The p-value should be interpreted as a test result under the reported analysis, not as a treatment-effect estimate or a measure of clinical importance. The record’s lack of a contrast confidence interval means readers cannot directly assess the precision of the reported mean difference from the inferential summary. This limits what can be concluded even though the test and its p-value are reported.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome and several secondary outcomes, including tolerability, serious adverse events, and adverse-event categories involving growth and development, pulmonary/respiratory, allergy/immunology, dermatologic, endocrinologic, and gastrointestinal events. The record also reports 13 outcome measures and 12 secondary outcomes posted. The complete statistical plan, including whether outcomes or analyses were selected in advance and how multiple comparisons were handled, is not reported in the registry record.

Multiplicity is a general concern when many outcomes or tests are examined. The more comparisons made, the greater the chance that at least one result may appear notable by chance if each test is treated as a separate unadjusted test. Adjustments or a prespecified hierarchy can address this concern, but the record does not report such a method. This uncertainty does not establish that the analysis was inappropriate; it means readers cannot determine from the registry record how multiplicity was managed. The primary result should be distinguished from exploratory interpretation of secondary outcomes.

Missing primary endpoint measurements are also relevant. The record states that data at 4 or 5 months were available for 9/11 propranolol participants and 6/8 prednisolone participants because of missed appointments. It reports that 90% (138/154) of study appointments were completed. The latter statistic describes appointments rather than the availability of the primary outcome for every participant. The registry does not report a missing-data method, an imputation approach, or sensitivity analyses to examine how unavailable measurements might influence the comparison.

The analysis population is not specified in enough detail to establish whether all randomized participants were included in the primary comparison or how their data were handled. The endpoint summaries are based on the stated available-data counts, but the record does not describe whether any additional participants contributed to the test or how the test dealt with missing measurements. The protocol, detailed analysis plan, interim analyses, and sample-size justification are not reported in the registry record. Without those details, readers cannot verify the planned analysis choices or determine whether the posted analysis departs from them.

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

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: failure to thrive (62.5%), adrenal crisis (12.5%), dehydration (12.5%).

The registry lists serious adverse events for 1 of 11 propranolol participants and 5 of 8 prednisolone participants. These are descriptive counts of participants affected as reported in the record. The record also lists failure to thrive at 62.5%, adrenal crisis at 12.5%, and dehydration at 12.5% among the top serious-event terms. It does not provide enough information alongside those percentages to assign each term to a particular treatment group or to infer how many events occurred per participant.

These counts should not be treated as a stable estimate of comparative safety. The groups are small, and the registry record does not report confidence intervals or a statistical comparison for the serious-event counts. A difference in observed counts does not by itself establish a causal difference in risk. The record also lists other adverse-event categories, but their detailed results are not described in the registry entries. Death counts are not reported in the registry record.

Serious adverse events and ordinary tolerability are different safety summaries. The registry lists tolerability as a secondary outcome, but no tolerability result is stated. The serious-event counts therefore cannot substitute for a complete account of tolerability, event severity, timing, or treatment relationship. They should be read as limited descriptive safety information, not as an overall safety conclusion or a basis for treatment decisions.

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

The result means that the registry reports a mean of 0.57 for propranolol and 0.63 for prednisolone on the primary endpoint, with separate 95% confidence intervals, and a two-sided comparison test with mean difference (net) 0.06 and p = 0.77. The posted analysis does not show strong statistical evidence against the null hypothesis tested by that t-test. It does not establish that the treatments have equal effects, and it does not show that any possible difference is clinically unimportant.

The group intervals are wide relative to the estimates, especially for prednisolone, and no confidence interval is reported for the contrast. The available outcome data come from 9/11 propranolol participants and 6/8 prednisolone participants. These features limit precision and make the handling of missing measurements important. The record does not provide enough detail to assess missing-data assumptions, the analysis population, or the full statistical plan.

The measurement itself also has an interpretive limitation: the outcome title refers to decrease in size in square mm, while the description refers to proportional change in total surface area. The registry does not resolve this difference in wording. Readers should not assign a more specific scale interpretation than the record supports.

The safety figures are separate descriptive observations. The serious adverse-event counts are small, and no comparative uncertainty estimate is reported. They should not be read as proof that one treatment is safer than the other. Likewise, a nonsignificant test does not establish that either treatment is ineffective, effective, equivalent, or preferable. Those claims would require evidence and analyses beyond what is reported in the registry record.

The result is best understood as an imprecise comparison of the observed primary endpoint among participants with available follow-up, accompanied by a reported test that does not reject its null hypothesis. It does not settle the treatment comparison, quantify the full range of plausible differences, or provide a definitive account of safety.

14. Why This Trial Matters Statistically

This trial is statistically instructive because it illustrates the distinction between a test result and an estimate of effect. The reported p-value is 0.77, but the p-value alone does not state how large a difference might be compatible with the data. The registry gives group-level confidence intervals but no confidence interval for the group contrast. A contrast interval would be more directly useful for understanding the precision of the comparative estimate.

The available follow-up also matters. The primary outcome was available for 9/11 propranolol participants and 6/8 prednisolone participants, while 90% (138/154) of appointments were completed. Those figures describe different things: endpoint availability and appointment completion. Recognizing that distinction prevents a high appointment-completion percentage from being mistaken for complete primary-outcome data.

The endpoint description demonstrates why clear definitions matter. The title refers to decrease in size in square mm, while the outcome description calls the measure proportional change. Even when an outcome is quantitative, ambiguity about its scale can affect how a mean is understood and whether comparisons with other results are meaningful. The record does not resolve that ambiguity.

The trial also provides a reminder that safety counts need context. Serious adverse events are reported in both groups, but the small counts and absence of comparative uncertainty estimates limit statistical interpretation. Event counts are not automatically evidence of a reliable treatment difference.

Finally, a planning illustration based on an observed effect is not observed power and does not show what the trial was designed to detect. The registry does not report the trial’s own sample-size justification. Separating planning assumptions, observed estimates, and inferential conclusions is essential when reading a small trial record.

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 trial-specific limitations constrain interpretation. The status is terminated, but the reason for termination is not reported in the registry record. Enrollment is listed as 19, while the primary endpoint summaries include 9 propranolol participants and 6 prednisolone participants. The record attributes missing 4- or 5-month outcome data to missed appointments, but it does not explain the missingness for individual participants or describe any method for handling unavailable measurements.

The registry reports that 90% (138/154) of study appointments were completed. This gives a broad measure of visit completion, but it does not eliminate the missing primary outcome data or establish that missingness was unrelated to outcome. The analysis population and whether the primary test included only observed measurements are not described in sufficient detail. No sensitivity analysis is reported in the registry record.

The outcome has a wording ambiguity: the title describes decrease in size in square mm, while the description identifies proportional change in total surface area. The record does not clarify which scale the posted values represent. It also does not report a confidence interval for the between-group difference, a threshold for clinical importance, or an equivalence margin. The statistical analysis plan, sample-size justification, and multiplicity strategy are not reported.

The trial is identified as randomized, single-masked, and parallel, but the registry does not state who was masked or provide further information about allocation procedures. The reason for termination and the detailed analysis methods are not reported. These omissions do not show that the conduct was flawed; they limit the ability to evaluate design and analysis choices from the registry record alone.

Safety interpretation has its own limits. Serious adverse-event counts are reported, but no comparative uncertainty estimate is provided, and the record does not give detailed group-specific results for each listed serious-event term. Death counts are not reported in the registry record. The available safety information is therefore descriptive rather than a comprehensive comparative assessment.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Propranolol Versus Prednisolone for Treatment of Symptomatic Hemangiomas — NCT00967226.
• Registry record last updated 2016-02-24; status last verified 2016-01.
• Results first posted 2016-02-24.
• 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 record describes a terminated, randomized, single-masked, parallel phase 2 trial comparing propranolol with prednisolone for symptomatic hemangiomas in infants. Enrollment is listed as 19. The primary endpoint concerns change in hemangioma size at 4-5 months after therapy began, measured from lesion dimensions, although the title and description differ on whether the scale is a decrease in square mm or proportional change.

Primary-outcome data were available for 9/11 propranolol participants and 6/8 prednisolone participants. The reported group means are 0.57 and 0.63, with 95% confidence intervals of 0.34 to 0.8 and 0.14 to 1.11. The reported two-sided t-test gives a mean difference (net) of 0.06 and p = 0.77. No confidence interval for the difference is reported.

The registry lists serious adverse events in both groups, alongside additional adverse-event categories. Missing-data methods, the analysis population, the sample-size justification, and the reason for termination are not reported in the registry record. The efficacy and safety summaries should therefore be interpreted cautiously.