Comparing Two Forms of Head Immobilization for Stereotactic Radiotherapy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Comparing Two Forms of Head Immobilization for Stereotactic Radiotherapy (NCT00328510), a randomized phase 3 study. The primary outcome is Distance From Ideal to Center of GTC Frame and BrainLab Thermoplastic Mask With Respect to Average and Variability.

2. Clinical Question

The clinical question is about positioning during SRT, not a comparison of tumor control, survival, symptoms, or other patient outcomes. The posted outcome is measured in millimeters and is described as a mean with standard deviation. A difference in this measure addresses the recorded positioning result under the two assigned immobilization approaches; it does not, on its own, establish whether any difference changes a clinical outcome.

3. Trial Design

This was a randomized, phase 3, parallel-group trial with no masking. The registry record lists 83 participants enrolled and records the trial as completed. Its start is listed as 2005-02 and primary completion as 2008-01. The sponsor was Sidney Kimmel Comprehensive Cancer Center at Thomas Jefferson University.

One group received stereotactic radiation therapy using a GTC frame to immobilize the head. The other received the same broad treatment setting using a BrainLab thermoplastic mask. Random allocation is intended to make the assigned groups comparable on average, although randomization does not guarantee that the groups are identical in every characteristic, especially in a trial of this size. The record does not report the allocation sequence method, whether allocation was concealed, or whether any stratification was used.

The trial was unmasked. The registry does not describe masking of the people taking or assessing the positioning measurements, and it does not report procedures intended to reduce measurement or analytic bias. The primary outcome was measured during SRT. The posted arm counts for that outcome are 32 for the GTC Frame and 37 for the BrainLab Thermoplastic Mask. The record does not reconcile those counts with the enrollment of 83 or explain why the outcome counts differ from enrollment.

The registry lists one outcome measure and no posted analyses beyond the displayed primary comparison. It does not describe the original sample-size rationale, planned analysis population, handling of missing outcome measurements, or a detailed statistical analysis plan. These omissions limit what can be inferred about whether the posted analysis followed a prespecified plan.

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:

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 Distance From Ideal to Center of GTC Frame and BrainLab Thermoplastic Mask With Respect to Average and Variability, measured over Measurements taken during SRT.

The primary measure concerns the distance from an ideal position to the center of the GTC frame or BrainLab thermoplastic mask, expressed in millimeters, and is summarized by a mean and standard deviation. Measurements were taken during SRT. The registry says ExacTRAC imaging was used to assess positioning and that images yield lateral, longitudinal, and vertical isocenter deviations, as well as head rotations about respective axes.

Positioning is relevant to fractionated stereotactic radiotherapy because the registry describes the need for precise and reproducible head immobilization. A measure of deviation from an ideal position can describe how closely the recorded position corresponds to that target, while variability describes how much measurements vary around their mean. These are related but distinct features: a mean describes the average recorded distance, and a standard deviation describes spread among measurements.

The record does not specify the precise formula for the distance, whether dimensions or rotations were combined, or how repeated measurements within a participant were handled to produce the posted arm-level values. Consequently, the posted result should be interpreted as the registry’s reported continuous positioning outcome rather than as a fully specified account of each directional deviation or rotation. The outcome is a technical positioning measure; it is not itself a measure of treatment effectiveness or patient benefit.

7. Results Posted in the Registry

• BrainLab Thermoplastic Mask: mean 3.17 versus 2.00 on control; difference 1.17 (95% CI 0.45 to 1.89), standardized effect d = 0.73; z-test p = 0.002.

8. Statistical Interpretation

9. Statistical Methods Explained

The registry presents a continuous outcome as a mean and standard deviation for each randomized group, along with a mean difference, confidence interval, and unadjusted p-value. These summaries address different questions. The group means describe the average recorded positioning value; the standard deviations describe the spread of participant-level or measurement-level values as defined by the study; and the comparison estimates the difference between the group means. The record does not clarify the level at which repeated measurements were summarized.

What does a mean difference represent?

A mean difference is the value for one group minus the value for the other group, using the comparison’s stated direction. Here, the reported comparison is BrainLab Thermoplastic Mask minus GTC Frame, and the estimate is 1.17 millimeters. A positive estimate therefore corresponds to a larger mean for the mask group on the reported scale. The meaning depends on how the outcome is defined: if larger values represent greater distance from the ideal position, the estimate describes a greater average measured distance. The registry does not fully specify how the distance was constructed from the listed directional deviations and rotations.

The mean difference is an absolute contrast in the outcome’s original units. It is not a ratio and does not describe each participant’s experience. Group summaries can conceal overlap between individual observations, and they do not show whether every participant in one group had a larger value than every participant in the other. Interpretation should therefore remain at the level of the reported group comparison.

Why report a standard deviation?

A standard deviation describes how dispersed observations are around their group mean. A relatively large value indicates more spread in the measured values; a relatively small value indicates less spread. It is not the standard error of the mean and should not be read as an uncertainty interval around the group mean. The reported standard deviations are 1.04 millimeters for the GTC Frame and 1.95 millimeters for the BrainLab Thermoplastic Mask. The registry does not report distribution plots or enough detail to determine whether the outcome was symmetric, contained outliers, or was summarized after accounting for repeated measurements.

When observations are skewed or include influential outliers, a mean and standard deviation may not fully characterize their distribution. General statistical practice may use additional summaries or models suited to such features, but the registry record does not report whether any such checks or alternatives were used. It also does not report whether participants contributed equal numbers of measurements, a detail that could matter if the posted summaries were based on repeated imaging observations.

What do the p-value and confidence interval add?

A p-value measures how incompatible the observed data are with a specified null model, assuming that model and the analysis assumptions are correct. The unadjusted p-value is 0.002. A small p-value can indicate that the observed difference would be unusual under a no-difference model, but it does not measure the size or clinical importance of the difference. Nor does it establish that the analysis model was appropriate. The specific test or model used for the posted p-value is not reported in the registry record.

A confidence interval expresses uncertainty about the estimated difference under the method used to construct it. The reported interval lies above zero, which is consistent with a positive mean difference for mask minus frame. A confidence interval is not a guarantee that the true value lies within its endpoints, and it does not include every possible source of uncertainty. In particular, it cannot remedy incomplete reporting about measurement construction, missing outcomes, or analytic choices.

Why does over-dispersion matter in general?

In general, over-dispersion means that observed variability is greater than a model assumes. If a model treats data as less variable than they are, standard errors can be too small and confidence intervals too narrow, making evidence appear more precise than it is. Appropriate modeling can account for extra variability, clustering, or repeated observations. This is a general statistical consideration, not a claim that over-dispersion occurred in this trial. The registry does not describe a dispersion assessment, a repeated-measures model, or another adjustment for within-participant correlation.

What can be said about the analysis specification?

The comparison is unadjusted, and no discrepancy is recorded between the posted result and the simple unadjusted recalculation. The registry does not report covariate adjustment, the chosen inferential test, assumptions checks, or an analysis plan. It also does not give enough detail to determine whether the unit of analysis was a participant, an image, a treatment fraction, or a summary for each participant. These omissions matter because measurements taken repeatedly from one participant are generally not independent observations. If treated as independent when they are correlated, the apparent amount of information can be overstated.

The trial’s randomized parallel design supports a comparison by assigned group, but the registry does not report whether the analysis followed assignment regardless of treatment received, whether any participants crossed over, or how deviations from assigned immobilization were addressed. These methods cannot be inferred from the arm labels or the outcome summaries. General statistical principles can explain why such decisions matter, but they do not establish which approach the investigators used.

10. Confidence Intervals and Precision

The estimated mean difference is 1.17 millimeters, with a reported confidence interval that extends from 0.446 to 1.894 in the registry record. The interval lies above zero, so the reported range is consistent with a positive mean difference for BrainLab Thermoplastic Mask minus GTC Frame. The interval communicates uncertainty around the estimated contrast; it does not make the estimate exact or establish a clinically important threshold.

The point estimate and interval should be read together. The estimate gives the central value of the comparison, while the interval shows that values other than the point estimate remain compatible with the analysis. The reported p-value is 0.002, but the p-value and interval do not answer the same question: the p-value summarizes compatibility with a null model, while the interval displays a range of estimates under its construction method.

Precision depends on the number and quality of observations, the variability in the outcome, and the analytic method. The posted outcome counts are 32 and 37, and the standard deviations are 1.04 and 1.95 millimeters in the respective groups. The registry does not report why the outcome counts differ from enrollment, how repeated measurements were handled, or the method used to calculate the confidence interval. Those reporting gaps limit assessment of precision beyond the interval itself.

The interval reflects only the uncertainty captured by its statistical procedure. It does not account for unreported measurement choices, possible bias in an unmasked trial, missing outcome data, or limitations in applying the findings to other populations and settings. Thus, the numerical interval is informative but not a complete measure of uncertainty about the broader clinical meaning of the result.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm, not total events. Each entry illustrates the number of participants per arm for a specified fraction of the observed effect and a target power. The displayed examples include planning for 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect. For 100% of the observed effect, the figures are 30 participants per arm for 80% power and 40 participants per arm for 90% power. For 75% of the observed effect, they are 52 and 70 participants per arm, respectively. For 50% of the observed effect, they are 117 and 157 participants per arm, respectively.

The 80% power column gives the planning count associated with an 80% chance of detecting the specified effect under the stated assumptions, if that effect and those assumptions were correct. The 90% power column gives the corresponding count for a 90% chance. These are prospective planning illustrations based on the observed effect, not observed power calculated from the completed trial. Smaller effects need more participants because they are harder to distinguish from random variation.

The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is the stated fraction of the observed effect. The trial’s own original design assumptions and sample-size justification are not reported in the registry record, so these illustrations should not be interpreted as the trial’s original target or as evidence that the trial was designed around its eventual result.

12. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 1.170; pooled standard deviation σ = 1.595.
2. n = 2 × (1.96 + 0.842)² × 1.595² / 1.170² = 29.2, rounded up to 30 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- 1.17                      # observed difference in means
sigma <- 1.59485927887165                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(328510)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 30 (80% power), 40 (90% power)
 75% of observed effect: n per arm = 52 (80% power), 70 (90% power)
 50% of observed effect: n per arm = 117 (80% power), 157 (90% power)
power.t.test: 30.2 
Simulated power at n = 30 per arm: 0.797 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists one outcome measure and no posted secondary outcomes. It records no analyses. The posted primary comparison has an unadjusted p-value of 0.002. The record does not report whether investigators considered additional outcomes, time points, positioning dimensions, subgroup comparisons, or alternative analyses that could create multiplicity. It also does not report a multiplicity adjustment or a prespecified hierarchy for testing outcomes. Without that information, it is not possible to assess whether the posted p-value was selected from a larger set of analyses.

Multiplicity is a general concern when many hypotheses are tested: the chance of at least one apparently notable finding can increase as the number of tests grows. Adjustments or prespecified testing strategies may address this, but the registry does not describe such methods. The single listed outcome does not establish that only one analysis was conducted; it means the record does not provide a fuller accounting of analyses.

The record does not describe the amount or pattern of missing outcome data, how missing measurements were handled, or whether any imputation was used. The registry enrollment is 83, while the posted outcome counts are 32 for GTC Frame and 37 for BrainLab Thermoplastic Mask. The reason for the difference between enrollment and the arm-level outcome counts is not reported in the registry record. These counts should not be treated as a complete explanation of participant flow.

The analysis population is also not specified. The record does not state whether the comparison included all randomized participants, only those with available measurements, or another defined group. It does not report exclusions after randomization, adherence to assigned immobilization, or handling of protocol deviations. These details matter because analysis choices can affect comparability created by randomization. The trial’s original statistical analysis plan and its provisions for multiplicity, missing data, and analysis populations are not reported in the registry record.

14. 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.

The registry reports zero serious adverse events among the 32 participants at risk in the GTC Frame group and zero among the 37 participants at risk in the BrainLab Thermoplastic Mask group. It also reports zero other adverse events among the participants at risk in each group. These are the posted counts for the registry’s adverse-event categories.

Zero reported events means that no events in those categories are recorded in the posted results for the listed at-risk groups. It does not demonstrate that the risk is absent, and it does not provide a precise estimate of how often events might occur in a larger population. The record does not give event definitions, ascertainment procedures, observation duration for safety, or an explanation of whether all participants were assessed in the same way.

The death count is not reported in the registry record. The safety information should therefore be read as a limited description of the adverse-event counts recorded, not as a comprehensive assessment of all possible harms or a comparison establishing that the approaches have equivalent safety.

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

The posted result supports a descriptive conclusion about the registry’s positioning measure: the mean was 2.0 millimeters with GTC Frame and 3.17 millimeters with BrainLab Thermoplastic Mask, with a reported mean difference of 1.17 millimeters for mask minus frame. The confidence interval lies above zero, and the unadjusted p-value is 0.002. Together, these statistics indicate a difference in the reported mean positioning measure under the analysis represented in the registry.

The result does not answer whether the difference is clinically important. The record does not specify a threshold for meaningful positioning differences, explain how the distance measure was assembled from directional deviations and rotations, or connect the measure to tumor control, side effects, or other patient-centered outcomes. A difference in a technical measure is not automatically a difference in clinical benefit.

The result also does not show that every participant in one group had a better or worse positioning result than every participant in the other. The reported standard deviations indicate variation within both groups, and group averages do not describe the full pattern of individual measurements. The reported variability was 1.04 millimeters in the GTC Frame group and 1.95 millimeters in the BrainLab Thermoplastic Mask group.

Interpretation is limited by incomplete reporting of the outcome construction, analysis method, analysis population, and missing-data handling. The trial was unmasked, and the registry does not describe masking of measurement assessors. The posted outcome counts do not account for all enrolled participants, and the reason is not reported in the registry record. These issues do not negate the reported comparison, but they constrain how confidently it can be generalized or connected to broader clinical claims.

The findings should be understood as a result about measured positioning during SRT in this randomized comparison. They do not establish that one approach is suitable for every patient or setting, and they should not be used alone to make an individual treatment decision.

16. Why This Trial Matters Statistically

This trial illustrates how a randomized design can be used to compare two technical approaches through a continuous positioning outcome. Random allocation provides a basis for comparing groups, while the reported means and mean difference put the contrast on the outcome’s original millimeter scale. The standard deviations add information about variation that a single group average cannot convey.

It also shows why statistical significance is only one part of interpretation. The unadjusted p-value is 0.002 and the confidence interval lies above zero, but neither statistic determines whether a positioning difference matters clinically. That question depends on a meaningful threshold and on evidence connecting the measurement to outcomes important to patients, neither of which is described in the registry record.

Measurement definition is central in technical trials. The registry mentions lateral, longitudinal, and vertical deviations and head rotations, but does not explain how those components yield the posted distance. When the construction of an endpoint is unclear, readers may not know exactly what quantity the group means represent or how it relates to the underlying measurements. Reporting the measurement procedure and the unit of analysis would help readers judge both validity and precision.

The record also demonstrates the value of transparent analysis reporting. Details about the test, repeated measurements, participant flow, missing observations, and analysis population would make the comparison easier to evaluate. The posted results offer a useful statistical contrast, but interpretation of its scope depends on these design and analysis details as well as the numerical estimate.

17. 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 are specific to what the registry record reports. The endpoint title describes distance from ideal to the center of the frame and mask, while the description refers to multiple directional deviations and head rotations. The record does not explain the formula linking those components to the posted mean, whether they were combined, or whether separate component results exist. This makes the precise construct represented by the outcome uncertain.

The outcome was measured during SRT, but the registry does not describe the number or timing of measurements per participant, how repeated observations were summarized, or whether the analysis accounted for within-participant correlation. It does not specify the unit of analysis. If each participant contributed repeated images, treating those images as independent would require careful justification; the record does not provide enough detail to assess this issue.

The trial was unmasked, and the registry does not report whether outcome assessors were masked or whether measurement procedures were standardized across groups. Because the endpoint depends on imaging measurements, information about image assessment, calibration, and reproducibility would help evaluate the possibility of measurement bias. Such procedures are not reported in the registry record.

Although enrollment is listed as 83, the posted outcome counts are 32 and 37. The reason for the difference is not reported. The record also does not provide participant flow, missing-data methods, the analysis population, baseline characteristics, adherence information, or the original sample-size rationale. These omissions restrict assessment of attrition, group comparability, and the extent to which the analysis followed the randomized assignment.

Finally, the registry does not report a clinically meaningful threshold for the outcome or patient-centered results that connect positioning to treatment benefit or harm. The posted comparison therefore supports interpretation of the measured technical outcome, but its clinical implications and generalizability remain uncertain.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Comparing Two Forms of Head Immobilization for Stereotactic Radiotherapy — NCT00328510.
• Registry record last updated 2016-11-29; status last verified 2016-10.
• Results first posted 2013-08-19.
• 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.

21. Record Summary

This completed phase 3 trial randomized participants in a parallel design to GTC Frame or BrainLab Thermoplastic Mask immobilization during fractionated stereotactic radiotherapy. The registry lists 83 enrolled participants and one primary continuous outcome measured during SRT using ExacTRAC imaging.

The posted means were 2.0 millimeters for GTC Frame and 3.17 millimeters for BrainLab Thermoplastic Mask. The reported mean difference for mask minus frame was 1.17 millimeters, with a confidence interval above zero and an unadjusted p-value of 0.002. Standard deviations were 1.04 and 1.95 millimeters, respectively.

The registry reports zero serious and zero other adverse events in the listed at-risk groups. It does not report a death count. Important details about the outcome calculation, analysis method, participant flow, missing data, and original sample-size rationale are not reported in the registry record. The result is therefore best understood as a comparison of the posted positioning measure, not as evidence about broader clinical outcomes.