Randomized Trial of Diphenhydramine Versus Continued Midazolam in "Difficult-to-sedate" Patients Undergoing Colonoscopy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Randomized Trial of Diphenhydramine Versus Continued Midazolam in "Difficult-to-sedate" Patients Undergoing Colonoscopy (NCT01769586), a randomized na study. The primary outcome is Number of Patients Who Achieve Adequate Sedation to Allow Colonoscopy (Defined as MOAA/S ≤3).

2. Clinical Question

The clinical question is about the relative performance of two sedation strategies in a specific situation: initial standard doses had not produced what the trial described as adequate sedation. It is not a comparison of the strategies as initial sedation for all colonoscopy patients. Nor does the primary outcome measure every aspect of the procedure or recovery. It records whether participants reached the stated sedation threshold within the stated assessment timeframe.

The posted results report 27 participants achieving adequate sedation in the diphenhydramine group and 65 in the midazolam group, with 100 participants in each group. These are counts for the registry-defined binary outcome. The registry record does not report further patient characteristics, details of the colonoscopy experience, or other primary questions. The result therefore addresses the stated threshold in the trial population, rather than broader questions about sedation or outcomes not recorded as endpoints.

3. Trial Design

The registry identifies this as a randomized, parallel-group trial with quadruple masking. Participants were assigned to diphenhydramine or continued midazolam after not achieving adequate sedation with initial standard sedation. The record lists an enrollment of 200 and reports outcome results for 100 participants in each group. It marks the study phase as not applicable and the overall study status as completed.

The two strategies used different incremental dosing schedules. Diphenhydramine was given in increments of 25 mcg, up to 3 times, for a maximum total of 75 mcg. Midazolam was given in increments of 1.5 mg, up to 3 times, for a maximum of 4.5 mg. The initial standard sedation is described as midazolam 5 mg and fentanyl 100 mcg. The registry record does not give further details about how clinicians decided whether initial sedation was inadequate, how additional doses were timed, or whether the strategy could be changed during the procedure.

The study is listed as sponsored by VA Connecticut Healthcare System. The record lists a start date of 2013-02 and a primary completion date of 2015-06. It does not report a detailed protocol, recruitment setting, eligibility criteria, or the identities of the people masked. “Quadruple” is the masking designation in the registry, but which roles it covers is not reported in the registry record.

Random assignment is intended, in general, to support a fair comparison by allocating participants to strategies through a chance-based process. A parallel-group design means the comparison is between groups assigned to different strategies rather than a planned sequence of both strategies for each participant. The registry record does not state the allocation sequence method, allocation concealment procedures, or the original design assumptions. Those details should not be inferred from the design labels alone.

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 Number of Patients Who Achieve Adequate Sedation to Allow Colonoscopy (Defined as MOAA/S ≤3), measured over Approximately 10 minutes or less.

The primary endpoint counts participants who achieved adequate sedation to allow colonoscopy. Adequate sedation is defined in the registry record as MOAA/S 3 or less, assessed approximately 10 minutes or less after the sedation strategy. The endpoint is binary for each participant: the stated threshold was either achieved or not achieved. The posted group results are counts, not mean scores on the MOAA/S scale.

The MOAA/S scale is described as ranging from 0 to 5. A score of 0 denotes general anesthesia, with no response to painful stimuli; a score of 5 denotes minimal sedation, with the patient fully awake. The endpoint’s cutoff focuses on whether the participant reached the registry’s specified level during the assessment period. It does not by itself describe how long that level lasted, the full sedation course, or every clinically relevant feature of the procedure.

In general, a binary endpoint can be straightforward to interpret when its definition and assessment time are explicit: the reader can compare how many participants in each group met the criterion. It also necessarily compresses a range of possible experiences into a yes-or-no outcome. Participants with different scores or experiences may fall on the same side of the cutoff, and the result does not communicate the distribution of scores. The registry record does not report a separate continuous MOAA/S result or further detail on how the assessment was conducted.

This endpoint was chosen to address whether sedation reached a threshold intended to allow colonoscopy in patients who were not adequately sedated after the initial regimen. Its interpretation should remain tied to that population, threshold, and assessment timeframe rather than being extended to outcomes the registry does not report.

7. Results Posted in the Registry

• Midazolam: 65.0% (65/100) versus 27.0% (27/100) on control; risk difference 38.0 percentage points (95% CI 25.2 to 50.8), risk ratio 2.41 (1.69–3.43); unadjusted two-proportion z-test p = <0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary: each participant either achieved the registry-defined MOAA/S threshold or did not. The posted group counts are 27 of 100 for diphenhydramine and 65 of 100 for continued midazolam. For this kind of outcome, several statistical summaries can describe the comparison. They answer related but distinct questions, so the scale should be named when interpreting a result.

What does the risk difference describe?

The risk difference subtracts the proportion with the outcome in one group from the proportion in the other. In the recalculation, the comparison is oriented as midazolam minus diphenhydramine, giving 0.38. This is an absolute difference in the probability of meeting the endpoint, expressed on the risk scale. It is not a relative increase and should not be read as a count of additional participants in a future group. The estimate is calculated from the posted group summaries without adjustment for other variables.

The registry’s own analysis also reports a risk difference, .38, with a 95% confidence interval of .24 to .50. It used a chi-squared method and reports p < 0.001. The unadjusted recalculation reports a 95% confidence interval from 0.252 to 0.508 and p < 0.001. Differences in the reported interval endpoints can arise from differences in calculation methods or rounding. The registry record does not give enough detail to determine the precise interval calculation used for its reported estimate.

What does the chi-squared test assess?

In general, a chi-squared test compares observed counts across categories with the counts expected under a null model of no association between group and outcome. For a two-group binary outcome, this tests whether the outcome distribution differs between the groups. The test result depends on the observed counts and the assumptions of the test, including appropriate independent observations. The registry identifies the method but does not report further implementation details.

A p-value describes how unusual data at least as incompatible with the null model as the observed data would be if that model and the test assumptions held. It is not a measure of the size of the effect, and it is not the probability that a finding is due to chance. The reported value p < 0.001 is below the stated threshold, but it does not replace interpretation of the estimated difference and its confidence interval. The record does not state whether the reported p-value was adjusted for multiple testing.

How do relative measures differ?

The recalculated relative risk compares the outcome risks as a ratio, and the recalculated odds ratio compares the outcome odds as a ratio. The relative risk is 2.407, with a 95% confidence interval from 1.692 to 3.426. The odds ratio is 5.021, with a 95% confidence interval from 2.747 to 9.178. Both summaries are oriented to compare midazolam with diphenhydramine. They use the same group outcome counts as the risk difference but express the contrast on different scales.

In general, a risk is the proportion of participants with an event, while odds compare the proportion with the event to the proportion without it. An odds ratio is not the same as a risk ratio, particularly when the event is not rare. The absolute risk difference is often more directly interpretable as a difference in outcome probability, while relative measures describe proportional contrasts. No one scale conveys every aspect of the result.

Why report a confidence interval?

In general, a confidence interval communicates statistical uncertainty around an estimate. A 95% confidence interval is generated by a method designed to cover the true parameter in 95% of repeated studies under its assumptions. It is not a statement that there is a 95% probability that this particular interval contains the true value. The interval’s location indicates the direction and range of values compatible with the analysis, while its width reflects precision.

The recalculated interval for the risk difference is entirely above zero, consistent with the reported p-value for a test of no difference. The registry interval is also reported entirely above zero. This agreement is a property of these reported results, not proof that every possible source of bias or uncertainty has been addressed. Confidence intervals do not account for unmeasured limitations, errors in outcome assessment, or analysis choices not represented in the calculation.

What is not established by these methods?

The registry reports one analysis for the primary outcome and lists zero posted secondary outcome measures. It does not report whether the analysis was intention-to-treat, per-protocol, or based on another population; nor does it report how missing outcomes were handled. The recalculation uses the posted group counts and is unadjusted. It does not reconstruct a fuller analysis from participant-level data, account for covariates, or resolve details absent from the registry record.

In general, statistical significance and precision address sampling uncertainty under specified methods. They do not establish generalizability, clinical importance, or completeness of reporting. Those judgments require attention to the trial population, intervention definitions, endpoint, conduct, and missing information as well as the numerical estimates.

10. Confidence Intervals and Precision

The recalculated risk difference is 0.38, with a 95% confidence interval from 0.252 to 0.508. This interval describes uncertainty around the absolute group difference under the unadjusted calculation. Its lower and upper limits are both above zero, so the range of values compatible with this calculation points toward a higher observed endpoint rate in the midazolam group. The interval is not a guarantee about the true effect and does not include uncertainties that the calculation does not model.

The registry’s own result is .38, with a 95% confidence interval from .24 to .50. Its interval is somewhat wider at the reported precision than the recalculated interval, and the endpoints differ. The registry result and the recalculation both report p < 0.001 and the same direction of comparison. The record does not describe the interval method in enough detail to explain the endpoint differences; no cause should be assumed from the values alone.

In general, a confidence interval gives a range of estimates compatible with the observed data under a statistical procedure. A narrower interval indicates greater precision under that procedure, while a wider interval indicates more uncertainty. The interval should be read alongside the point estimate: the point estimate alone can make an uncertain result seem more exact than it is. Conversely, a confidence interval does not by itself establish whether a difference matters clinically.

The group counts and endpoint definition also matter for precision. The estimate concerns the registry’s binary threshold, not the full MOAA/S score distribution. The record does not report outcome data for other timepoints or an alternative endpoint, so precision cannot be assessed for those outcomes. It also does not report missing-data details or analysis-population definitions, which limits assessment of whether the posted counts capture all randomized participants as intended.

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 illustrations are expressed as participants per arm, not as total events. They use the observed risk difference as a reference and consider planning for 100%, 75%, or 50% of that observed effect. For planning for 100% of the observed effect, the illustrations give 26 participants per arm for 80% power and 35 participants per arm for 90% power. For planning for 75% of the observed effect, they give 46 participants per arm for 80% power and 61 participants per arm for 90% power. For planning for 50% of the observed effect, they give 100 participants per arm for 80% power and 133 participants per arm for 90% power.

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 80% power column describes the planned number per arm under which the calculation would have an 80% chance of detecting the specified effect at the stated significance level, if the assumptions hold. The 90% column uses the higher target probability of detection under those same assumptions.

Smaller effects require more participants in these illustrations: planning for a smaller fraction of the observed effect produces larger per-arm figures. These are planning illustrations, not observed power. The registry record does not report the trial’s original sample-size assumptions or justification, so the illustrations should not be read as the study’s own design calculations.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.270; treatment proportion p₁ = 0.650; average p̄ = 0.460.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7048 = 1.3815.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6516 = 0.5484.
4. n = (1.3815 + 0.5484)² / (0.3800)² = 25.8, rounded up to 26 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
p0 <- 0.27                          # control proportion
p1 <- 0.65                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(1769586)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 26 (80% power), 35 (90% power)
 75% of observed effect: n per arm = 46 (80% power), 61 (90% power)
 50% of observed effect: n per arm = 100 (80% power), 133 (90% power)
power.prop.test: 25.8 
Simulated power at n = 26 per arm: 0.799 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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. Analyses Reported by the Sponsor

The registry’s own analysis reports a risk difference of .38, with a 95% confidence interval from .24 to .50 and p < 0.001, using a chi-squared method. The unadjusted recalculation from the posted group counts gives a risk difference of 0.38, with a 95% confidence interval from 0.252 to 0.508 and p < 0.001. Both analyses point in the same direction and report the same risk-difference magnitude at the displayed precision, although their confidence-interval endpoints differ.

The discrepancy flag is false, so the recalculation does not identify a flagged conflict with the registry result. The registry’s own analysis remains the trial’s reported analysis; the recalculation is a simple unadjusted result derived from the posted summary counts. It should not be treated as a replacement for an analysis based on participant-level data or for methodological details not reported in the registry.

The registry record does not explain the exact method used to form its confidence interval or provide further details of the chi-squared analysis. The observed difference in interval endpoints should therefore be described rather than attributed to a particular statistical choice. Both sets of reported results indicate that the estimated risk difference is above zero under their respective analyses.

14. Multiplicity, Missing Data and Analysis Populations

The registry record lists one primary outcome measure and one reported analysis, with no secondary outcome measures posted. It does not report whether additional outcomes or analyses were prespecified in the protocol, whether any interim analyses were planned or conducted, or whether the reported analysis was selected from a larger set of possible analyses. Those details are not reported in the registry record.

Multiplicity arises in general when researchers test several outcomes, timepoints, subgroups, or models. Repeated testing can increase the chance of finding at least one statistically significant result by chance, unless the analysis plan accounts for the testing strategy. The record’s listing of a single reported primary analysis does not establish that no other analyses occurred. It also does not state whether any adjustment for multiple testing was used. The reported p-value is therefore best described as the registry reports it, without assuming a multiplicity correction.

The registry reports 100 participants in each group for the primary outcome, but does not state how missing outcome data were handled or whether any participants were excluded from the analysis. The posted counts do not, by themselves, show whether all randomized participants were included in the primary analysis or whether any outcome values were unavailable. The handling of withdrawals, protocol deviations, and missing assessments is not reported in the registry record.

The analysis population is also not identified as intention-to-treat, per-protocol, or another population. In general, an intention-to-treat analysis keeps participants in their randomized groups, while other analysis populations may use different inclusion rules. The registry record does not provide enough information to classify this result or evaluate the impact of exclusions.

These omissions do not change the reported counts, but they limit assessment of how the analysis was conducted and how robust it is to missingness or analytic choices. The outcome is recorded as a count for each group; no further breakdown of participant flow or analysis-set membership is reported.

15. 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 non-serious adverse event terms

The posted serious adverse event figures show 0 serious adverse events among 100 participants at risk in the diphenhydramine group and 0 among 100 participants at risk in the midazolam group. No serious adverse event terms are listed. The registry also reports other adverse events affecting 2 participants in the diphenhydramine group and 8 in the midazolam group, with 100 participants at risk in each group for that measure.

These are counts in the registry’s adverse-event summary. The record does not provide event descriptions for the other adverse events, their severity, timing, relationship to study medication, or methods of collection. Death counts are not reported in the registry record. The serious-event figures should therefore be read as the posted counts, not as a complete account of all possible harms or as evidence that serious events cannot occur.

In general, safety counts need context about how events were defined, monitored, and reported, and how long participants were followed. The registry record does not provide that detail. The adverse-event counts are also distinct from the primary sedation endpoint and do not establish an overall comparison of safety between the strategies.

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

The primary outcome result shows that 27 participants in the diphenhydramine group and 65 in the continued-midazolam group achieved the registry-defined adequate-sedation threshold, with 100 participants reported in each group. The unadjusted recalculated risk difference, oriented as midazolam minus diphenhydramine, is 0.38. Its 95% confidence interval ranges from 0.252 to 0.508, and the reported p-value is p < 0.001. The registry’s analysis reports the same risk-difference magnitude at its displayed precision and also reports p < 0.001.

This finding applies to the trial’s stated population: patients undergoing colonoscopy who were not adequately sedated after initial standard sedation with midazolam 5 mg and fentanyl 100 mcg. It compares the study’s specified incremental diphenhydramine strategy with its specified continued-midazolam strategy. The endpoint is achievement of MOAA/S 3 or less, assessed approximately 10 minutes or less. It is not a result about other sedation regimens, populations, assessment times, or outcomes.

The result does not describe the full distribution of MOAA/S scores, the duration of sedation, procedural completion, recovery, or patient experience. Those outcomes are not reported as primary results in the registry record. Nor does the endpoint alone establish a broader judgment about the overall benefit-harm balance of either strategy. The posted adverse-event counts are limited and lack details about event types and follow-up.

A statistically small p-value does not mean that the result is free of bias or that its clinical importance is settled. The confidence interval conveys uncertainty under the calculation, but does not account for information absent from the registry, such as missing-data handling, analysis-population definitions, or detailed masking and allocation procedures. The result is best understood as evidence of a difference in the recorded binary endpoint within this trial, alongside limits in what the registry reports.

17. Why This Trial Matters Statistically

This trial poses a clear statistical comparison: two randomized groups, a binary endpoint, and a direct contrast in the number of participants meeting a defined sedation threshold. Such settings make the distinction between absolute and relative measures especially useful. The risk difference communicates the absolute contrast on the outcome scale, while the recalculated relative risk and odds ratio express the comparison as ratios. Each answers a different interpretive question, and no single measure replaces the endpoint definition.

The result also illustrates why effect estimates and precision should be read together. The recalculated risk difference is 0.38, with a 95% confidence interval from 0.252 to 0.508. The registry reports a risk difference of .38, with a 95% confidence interval from .24 to .50. Both intervals are reported zero, and both analyses report p < 0.001. The estimates communicate a substantial observed separation in the posted counts, while the interval endpoints express uncertainty under the respective methods.

The trial is statistically specific in an important way: it concerns patients who had not reached adequate sedation after a stated initial regimen. That restriction means the comparison should not be generalized automatically to all patients undergoing colonoscopy or to initial sedation decisions. The endpoint’s threshold also makes interpretation concrete but leaves information about score distributions and the wider procedural experience unaddressed.

Finally, the registry illustrates the value and limits of registry reporting. It provides group counts, an analysis method, an estimate, an interval, and a p-value, allowing a simple unadjusted recalculation. It does not report several details needed to assess the full analysis process, including missing-data handling, analysis population, multiplicity, and original sample-size assumptions. Statistical interpretation is strongest when the numerical result is considered together with both the design information and these reporting limits.

18. 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 gives limited information about participant characteristics, eligibility criteria, recruitment setting, and the specific process used to determine inadequate sedation after the initial regimen. Without those details, it is difficult to assess how closely the trial population represents other groups undergoing colonoscopy. The result should remain anchored to the stated difficult-to-sedate population rather than extended to patients who respond to initial sedation.

Although randomization and quadruple masking are recorded, the registry does not describe the allocation sequence, concealment procedures, or which roles were masked. The record also does not provide details about adherence to assigned dosing, rescue treatment, treatment switching, or the timing of individual assessments. These details could affect interpretation of the comparison but are not reported in the registry record.

The primary outcome is a binary threshold on the MOAA/S scale. The registry does not report the distribution of scores, how the assessment was performed, or further information about assessor training. The result therefore cannot show whether groups differed in other ways across the sedation scale or describe outcomes outside the stated threshold and assessment timeframe.

Missing outcome data, participant flow, protocol deviations, and analysis-population definitions are not reported. The registry does not state whether the primary analysis included all randomized participants or how unavailable assessments were handled. It also does not report whether the analysis plan accounted for multiple testing, or whether additional analyses were conducted. The trial’s original sample-size assumptions and justification are not reported in the registry record; the planning illustrations are separate from those design details.

The adverse-event summary contains counts but little context about event types, monitoring, attribution, or follow-up. Death counts are not reported. These limitations restrict conclusions about safety and about the overall experience of the compared strategies. They do not invalidate the posted endpoint counts, but they narrow the scope of conclusions that the registry results can support.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Randomized Trial of Diphenhydramine Versus Continued Midazolam in "Difficult-to-sedate" Patients Undergoing Colonoscopy — NCT01769586.
• Registry record last updated 2017-02-10; status last verified 2016-12.
• Results first posted 2016-09-30.
• 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.

22. Record Summary

The registry record describes a completed randomized, parallel-group trial with quadruple masking, sponsored by VA Connecticut Healthcare System. It lists an enrollment of 200 and compares diphenhydramine with continued midazolam in patients undergoing colonoscopy who were not adequately sedated after initial standard sedation with midazolam 5 mg and fentanyl 100 mcg.

The primary endpoint was the number achieving MOAA/S 3 or less, assessed approximately 10 minutes or less. The posted counts are 27 of 100 in the diphenhydramine group and 65 of 100 in the midazolam group. The registry reports a risk difference of .38, a 95% confidence interval from .24 to .50, and p < 0.001. An unadjusted recalculation gives a risk difference of 0.38, with a 95% confidence interval from 0.252 to 0.508 and p < 0.001.

The registry lists one primary analysis and no secondary outcome measures. Further details on missing data, analysis populations, multiplicity, and original sample-size assumptions are not reported in the registry record.