Bupivacaine for Benign Headache in the ED: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Bupivacaine for Benign Headache in the ED (NCT01785459), a randomized phase 2 study. The primary outcome is Length of Stay.

2. Clinical Question

In statistical terms, the comparison asks whether average length of stay differed between participants assigned to the injection approach and those assigned to standard care. The posted group means were 131.0 minutes for treatment and 158.0 minutes for standard care. The estimated mean difference, treatment minus standard care, was -27.0 minutes. This estimate is accompanied by a wide confidence interval, so the observed difference alone does not establish a precise estimate of the underlying group difference.

The registry also lists symptomatic relief of headache as a secondary outcome. The brief study description refers to a pain scale and symptom reevaluation, but the posted primary result analyzed is length of stay. The registry record does not provide a posted estimate for the secondary outcome. Accordingly, the length-of-stay result should not be treated as a direct numerical summary of headache relief, nor should it be interpreted as answering questions that were not reported.

3. Trial Design

The registry describes a randomized, parallel-group, Phase 2 trial with no masking. It lists enrollment as 23 and status as terminated. The study was sponsored by Wake Forest University Health Sciences. The record gives a start date of 2013-10 and a primary completion date of 2016-07. It does not report why the trial was terminated or provide a detailed account of recruitment and follow-up.

Participants were assigned to 0.5% bupivacaine or standard care. The treatment arm received bilateral injections into the cervical paraspinous musculature; standard care was described as intravenous Prochlorperazine. The study description says participants completed a validated pain scale before treatment and 20 minutes after injection, followed by symptom reevaluation. It further states that participants could then be discharged or receive standard treatment at the treating clinician’s discretion. The registry record does not report additional procedural details or how that discretion was applied.

For the primary outcome, the registry lists length of stay, measured in minutes, with a mean and standard deviation. The description refers both to total encounter time and to the interval from doctor encounter to the disposition decision. The posted results give one mean and standard deviation for each group, without identifying which of the described time intervals those values represent. This limits how specifically the result can be tied to the operational definition of length of stay.

The posted group summaries include 11 participants in standard care and 12 in treatment. These are the group sizes associated with the reported outcome summaries; the registry record does not explain whether any enrolled participants were excluded from the outcome analysis or whether all assigned participants contributed data. The analysis populations and handling of missing outcome values are not reported in the registry record.

The registry lists three outcome measures and one posted secondary outcome, symptomatic relief of headache. It records no analysis entries. The original statistical analysis plan, the trial’s own design assumptions, and any reason for the recorded termination are not reported in the registry record. The planning illustrations calculated from the observed result should therefore be kept separate from the original trial design.

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

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 Length of Stay, measured over enrollment day.

The primary endpoint is length of stay, a continuous measure expressed in minutes. A continuous outcome can take a range of values rather than being limited to categories. Reporting a group mean gives a summary of the average recorded time, while the standard deviation describes how spread out individual times were around that mean.

Length of stay is used as an operational measure of the time spent in the emergency department. The registry description defines it in relation to total encounter time and, separately, the interval from doctor encounter to the disposition decision. These intervals represent related but not necessarily identical aspects of a visit. The posted group means do not specify which interval was used, so the result cannot be assigned more precisely to one definition on the basis of the registry record.

The reported mean was 158.0 minutes in standard care and 131.0 minutes in treatment. The estimated mean difference was -27.0 minutes, with the negative direction indicating a lower observed mean in the treatment group under the stated treatment-minus-standard-care comparison. This is a summary of the posted outcome data, not a statement about why the times differed or whether the result would apply in other settings.

Length of stay is not itself a direct measure of pain relief. Although the registry lists symptomatic relief as a secondary outcome and describes pain-scale assessment, the primary result is about time. The registry record does not provide a posted numerical result for the listed secondary outcome. The distinction matters when interpreting what the primary endpoint can answer.

7. Results Posted in the Registry

• Treatment: mean 131.00 versus 158.00 on control; difference -27.00 (95% CI -108.71 to 54.71), standardized effect d = -0.27; z-test p = 0.517.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous: length of stay is recorded in minutes, and the registry reports a mean and standard deviation for each group. A mean difference compares the two group averages on the original measurement scale. Here, the comparison is treatment minus standard care, and the estimate is -27.0 minutes. A negative value means the treatment-group mean was lower in the posted summaries. It does not, by itself, explain the cause of the difference or establish that the difference is reproducible.

What does the standard deviation describe?

A standard deviation summarizes the spread of individual observations around a group mean. It is not the uncertainty interval for the mean, and it does not describe the range containing every participant’s result. The reported standard deviations were 85.5 minutes in standard care and 113.5 minutes in treatment. These values indicate considerable variation in recorded length of stay within both groups. When individual outcomes vary substantially, a group mean may be estimated less precisely, particularly when the group sizes are limited.

The posted group sizes are 11 for standard care and 12 for treatment. The standard errors listed for the group means are 25.78 and 32.76, respectively. A standard error describes the sampling uncertainty of an estimated mean under the statistical model; it is distinct from the standard deviation, which describes spread among individual observations. The registry does not report participant-level observations, so readers cannot assess the distribution of times, unusual values, or whether the data contain influential observations.

How should the mean difference be read?

The mean difference is an effect estimate expressed in minutes. It summarizes the contrast between the observed group means, with the stated direction of treatment minus standard care. Its confidence interval runs from -108.7 to 54.71 minutes. The interval gives a range of values compatible with the data and the method’s assumptions; it is not a range of individual participant outcomes. Because it spans zero, it includes both directions of difference. The estimate is therefore uncertain about whether the average length of stay is lower or higher under treatment in the underlying population.

The unadjusted p-value is 0.517. A p-value measures how unusual the observed data, or data more extreme, would be under a particular no-difference model and its assumptions. It is not the probability that the null hypothesis is true, nor is it a measure of the probability that the finding occurred by chance. The registry does not report an adjusted analysis or a prespecified significance threshold for the primary comparison. The planning assumptions list a two-sided significance level of 0.05, but that assumption belongs to the planning illustration and should not be mistaken for a reported original trial analysis plan.

Why does variability matter?

Over-dispersion is a general statistical concern when observed data vary more than a model assumes. If a model understates variability, uncertainty estimates can be too narrow and statistical tests can be misleading. The registry provides standard deviations for the continuous endpoint, but it does not report a distributional assessment or a model-checking procedure. It is therefore not possible to establish from the record whether assumptions about the outcome distribution were examined. In general, examining the distribution and sensitivity to influential observations can be useful for continuous outcomes, especially when samples are limited.

The comparison is described as unadjusted in the recalculation of the posted summary statistics. An unadjusted comparison uses the group summaries without accounting for other variables. Adjustment can be useful when prespecified factors are relevant to the outcome, but no covariates, adjusted estimates, or adjustment plan are reported in the registry record. The unadjusted estimate should not be presented as if it controlled for baseline differences or other sources of variation.

What can randomization and masking establish?

Random allocation is a design method intended to reduce systematic differences between groups at assignment. In a small trial, randomization does not guarantee that all characteristics will be balanced in the realized groups. The registry does not provide baseline characteristics, details of sequence generation, or allocation concealment. These features are therefore not reported in the registry record. The study is listed as having no masking. In general, lack of masking can influence participant reports, care decisions, or outcome assessment, depending on the endpoint and trial procedures. The registry does not specify whether outcome assessment was independently performed.

The record identifies the primary outcome but does not describe the original statistical analysis plan, model specification, missing-data method, analysis population, or treatment of protocol deviations. It lists no analysis entries. These omissions restrict the ability to evaluate whether the reported result followed a prespecified approach or whether the analysis was sensitive to decisions made after data collection.

The recalculated mean difference and interval are unadjusted calculations from the posted summary statistics. The registry record does not indicate a discrepancy between those calculations and a sponsor-reported analysis. Such a recalculation remains a summary-based analysis rather than a substitute for participant-level data or the trial’s complete analysis plan. It cannot address missing observations, outcome distribution, adjustment choices, or adherence to the study protocol beyond what the record reports.

10. Confidence Intervals and Precision

The mean difference in length of stay was -27.0 minutes, with a confidence interval from -108.7 to 54.71 minutes. A confidence interval describes uncertainty around an estimate under the method’s assumptions. It is not a guarantee that the true effect lies in the interval, and it is not a range that contains a specified proportion of individual patients’ times.

This interval is wide relative to the point estimate and crosses zero. It includes a potentially lower mean length of stay with treatment and a potentially higher mean length of stay with treatment, relative to standard care. The data therefore leave considerable uncertainty about both direction and magnitude. The interval does not establish equivalence: demonstrating equivalence would require an appropriate design and a prespecified margin, neither of which is reported in the registry record.

The precision is consistent with the small posted outcome groups and the observed spread in times. The standard deviations were 85.5 minutes for standard care and 113.5 minutes for treatment. Substantial participant-to-participant variability makes a mean difference harder to estimate precisely, while limited group sizes constrain the information available. The registry does not report whether any observations were missing or excluded, so the effect of those possibilities on precision cannot be assessed.

The unadjusted p-value of 0.517 should be read alongside the estimate and interval rather than as a stand-alone verdict. It does not establish that there is no difference, and it does not quantify the chance that the observed estimate is correct. The interval is more informative about the range of effect sizes compatible with the data, while still depending on the statistical method and its assumptions.

The record does not report a prespecified clinically important difference for length of stay. Without that context, the interval cannot be judged against a threshold for practical importance. The posted data support describing the estimate as imprecise; they do not support a precise claim about the size of the average difference.

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. They are planning illustrations based on the observed effect, rather than the trial’s own original design assumptions. The listed 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 set to the stated fraction of the observed effect. The original trial’s sample-size justification and design assumptions are not reported in the registry record.

For planning at 100% of the observed effect, the figures are 221 participants per arm for 80% power and 295 participants per arm for 90% power. For planning at 75% of the observed effect, they are 392 participants per arm for 80% power and 525 participants per arm for 90% power. For planning at 50% of the observed effect, they are 882 participants per arm for 80% power and 1,180 participants per arm for 90% power.

In plain language, the 80% power column gives the number of participants per arm estimated to provide an 80% chance of detecting the chosen effect under the stated assumptions if that effect is the true one. The 90% power column uses a higher target chance of detection and therefore calls for more participants under otherwise identical assumptions. Smaller effects are harder to distinguish from random variation, so the planning figures rise as the fraction of the observed effect gets smaller.

These are hypothetical planning calculations using the observed estimate, not observed power. They are not evidence that the terminated trial had a particular chance of detecting an effect, and they do not show that the original design used these assumptions. The normal approximation and equal allocation are stated planning assumptions; the registry record does not establish that they were part of the original protocol.

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 = -27.000; pooled standard deviation σ = 101.138.
2. n = 2 × (1.96 + 0.842)² × 101.138² / -27.000² = 220.3, rounded up to 221 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 <- -27                      # observed difference in means
sigma <- 101.138106896791                     # 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(1785459)
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 = 221 (80% power), 295 (90% power)
 75% of observed effect: n per arm = 392 (80% power), 525 (90% power)
 50% of observed effect: n per arm = 882 (80% power), 1,180 (90% power)
power.t.test: 221.2 
Simulated power at n = 221 per arm: 0.804 (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 three outcome measures and identifies symptomatic relief of headache as a secondary outcome. The primary endpoint is length of stay. It records no analysis entries and provides no posted numerical result for the listed secondary outcome. The registry record does not report whether all planned outcomes were analyzed, whether outcomes were added or changed, or how outcome selection was handled.

Multiplicity arises when several outcomes, time points, or comparisons are tested. Conducting multiple tests can increase the chance of at least one apparently notable result, so a plan may specify how the overall error rate is handled. The registry record does not report a multiplicity adjustment or a rationale for not using one. The planning illustration specifies a two-sided significance level of 0.05, but that is not evidence of the original trial’s multiplicity plan. No interpretation of the secondary outcome can be made from a numerical result because one is not reported in the registry record.

Missing-data handling is not reported in the registry record. The posted outcome groups contain 11 participants in standard care and 12 in treatment, but the record does not say whether these are all randomized participants, whether any outcome measurements were unavailable, or how missing values were treated. It also does not report whether participants were analyzed according to their assigned group or according to treatment received. Those definitions are important for interpreting a randomized comparison, particularly if participants did not follow their assignment or were excluded from analysis.

The registry record does not identify an intention-to-treat, per-protocol, or other analysis population. It does not describe exclusions, protocol deviations, or sensitivity analyses. The original analysis plan and any imputation approach are also not reported in the registry record. Consequently, the posted means and comparison should be understood as the results recorded for the reported groups, without assuming a particular missing-data strategy or analysis population.

These omissions do not demonstrate that missing data or multiplicity affected the result. They limit the ability to assess those issues. A complete interpretation would require the prespecified analysis plan and reporting of outcome availability, analysis populations, and any adjustments or sensitivity checks.

14. Safety Results

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

Deaths

Other (non-serious) adverse events

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

Most frequent non-serious adverse event terms

The registry reports serious adverse events for 11 participants at risk in standard care and 12 participants at risk in treatment. Serious adverse events affected 0 in each group. Deaths were 0 in each group. The record does not list serious adverse event terms.

For other, non-serious adverse events, the record reports 1 affected participant among 11 at risk in standard care and 2 affected participants among 12 at risk in treatment. These counts are descriptive. The registry record does not provide an inferential comparison or details that would establish whether an event was caused by the assigned treatment. They should not be interpreted as evidence of a difference in risk.

The small group sizes mean that the absence of reported serious events does not rule out uncommon harms. The record also does not provide event descriptions, timing, severity details beyond the serious/non-serious distinction, or information about completeness of adverse-event collection. The posted figures summarize what was recorded, not a comprehensive assessment of safety across broader populations or longer periods.

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

The result describes a randomized comparison in a terminated Phase 2 trial with no masking. In the posted summaries, average length of stay was 131.0 minutes with bupivacaine treatment and 158.0 minutes with standard care. The estimated mean difference was -27.0 minutes, with a confidence interval from -108.7 to 54.71 minutes and an unadjusted p-value of 0.517.

The estimate points toward a lower observed mean in the treatment group, but the interval is broad and spans zero. The data therefore do not give a precise estimate of the underlying difference. They are compatible with effects in either direction under the reported analysis. A result that does not meet a conventional significance threshold is not proof that the treatments have identical average outcomes, just as the negative point estimate is not proof of a reliable reduction in length of stay.

The endpoint is time in the emergency department, not a direct measure of symptom relief. Although symptomatic relief of headache is listed as a secondary outcome and the study description mentions pain-scale assessment, the registry record does not report a numerical result for that outcome. The length-of-stay estimate cannot stand in for an unreported pain or symptom result.

The registry’s description gives more than one way to define length of stay, and the posted result does not specify which measure generated the reported mean. The record also does not explain the reason for termination, the original design assumptions, analysis populations, or missing-data methods. These limits matter when deciding how much weight to place on the posted comparison.

Accordingly, the result should be read as an imprecise estimate from the data reported in the registry, not as a definitive answer about the treatments or as guidance for clinical care. It does not establish benefit, lack of benefit, equivalence, or comparative safety. Further interpretation would require fuller methods and outcome reporting, particularly for the listed secondary endpoint and the precise definition of the primary measure.

16. Why This Trial Matters Statistically

This trial illustrates why an effect estimate should be read together with its uncertainty. The observed mean difference is -27.0 minutes, but the confidence interval ranges from -108.7 to 54.71 minutes. Focusing only on the point estimate would obscure that the data remain compatible with differences in either direction and across a broad range of sizes.

The standard deviations, 85.5 minutes in standard care and 113.5 minutes in treatment, show substantial variation in the continuous outcome. This variation is important in a small randomized study: group means can differ because of chance variation, and estimates can remain imprecise even when a point estimate appears notable. Reporting means alone would not communicate that uncertainty.

The record also demonstrates the distinction between a primary endpoint and a listed secondary outcome. Length of stay is the posted primary result, while symptomatic relief of headache is listed secondarily without a numerical result in the registry. A trial can only answer statistical questions for which interpretable outcome data are reported. The endpoint’s operational definition also matters: total encounter time and time from doctor encounter to disposition decision are not necessarily interchangeable, and the record does not identify which definition underlies the posted means.

Finally, trial status and reporting context affect interpretation. The record lists the study as terminated, but not why, and does not report the original statistical plan, missing-data procedures, or analysis population. These gaps do not invalidate the reported summaries, but they limit how fully readers can assess design choices and potential sources of bias. Transparent reporting of estimates, intervals, outcome definitions, and analysis decisions is central to evaluating a clinical comparison.

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.

The registry record leaves several trial-specific limitations unresolved. It lists no masking, so participants and clinicians may have known the assigned approach. The record does not describe whether outcome assessors were masked, how disposition decisions were made, or whether discharge practices were standardized. These details are relevant because the primary endpoint includes time related to disposition.

The length-of-stay definition is not fully clear in the posted result. The registry description refers both to total time of encounter and to time from doctor encounter to disposition decision, while providing a single mean and standard deviation for each group. It does not identify which measure is represented or whether both were analyzed. This makes the endpoint less straightforward to interpret.

The trial is recorded as terminated, but the reason is not reported in the registry record. The record does not state whether termination affected recruitment, follow-up, or outcome availability. It also does not report the original design assumptions or the rationale for the planned enrollment. The planning illustrations based on the observed effect do not fill that gap.

The outcome groups are small, and the standard deviations are large relative to the estimated difference. The confidence interval is wide and includes zero. Baseline characteristics, allocation-concealment details, adherence, protocol deviations, missing observations, and analysis-population definitions are not reported in the registry record. No adjusted analysis or sensitivity analysis is listed.

Finally, the secondary outcome of symptomatic relief has no posted numerical result, and details of adverse events are limited. The registry does not explain whether outcome reporting was complete or whether other results were collected. These limitations restrict the precision and scope of conclusions that can be drawn from the record.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Bupivacaine for Benign Headache in the ED — NCT01785459.
• Registry record last updated 2022-07-14; status last verified 2019-03.
• Results first posted 2019-04-03.
• 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

NCT01785459 is a randomized, parallel-group, unmasked Phase 2 trial of 0.5% bupivacaine injections versus intravenous Prochlorperazine for benign headache in the emergency department. The registry lists enrollment as 23 and status as terminated. The primary outcome is length of stay in minutes.

The posted means were 131.0 minutes for treatment and 158.0 minutes for standard care. The estimated mean difference was -27.0 minutes, with a confidence interval from -108.7 to 54.71 minutes and an unadjusted p-value of 0.517. The interval is wide and includes zero, indicating substantial uncertainty about the difference.

Symptomatic relief of headache is listed as a secondary outcome, but a numerical result is not reported in the registry record. The reason for termination, the original analysis plan, missing-data handling, and analysis populations are also not reported.