This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Comparing Acute Pain Management Protocols for Patients With Sickle Cell Disease (NCT02222246), a randomized phase 4 study. The primary outcome is Difference in Pain Score as Measured by a Visual Analogue Scale (VAS).
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT02222246 |
| Phase | Phase 4 |
| Condition | Sickle Cell Disease |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None (open label) |
| Enrollment | 106 |
| Arms in the primary outcome | 2 |
| Lead sponsor | Duke University |
| Results posted | Yes |
2. Clinical Question
Population
The trial concerned adults with sickle cell disease experiencing vaso-occlusive pain and receiving care in an emergency department. The registry describes the primary-outcome unit of analysis as each emergency-department study visit, rather than each person. The registry record lists enrollment of 106, but the primary-outcome arm summaries each report 26 observations. The relationship between enrollment and those outcome summaries is not reported in the registry record.
Intervention
The patient-specific strategy used a protocol developed by the patient's healthcare team after randomization for use during emergency-department visits for vaso-occlusive crises during the study period. The intervention could use morphine sulfate or hydromorphone. Further details of the protocol are not reported in the registry record.
Comparator
The comparison strategy was a standardized analgesic protocol based on recent NHLBI recommendations. It used morphine sulfate or hydromorphone, with dosage based on weight. The registry lists intravenous or sub-cutaneous delivery in the arm description. These were active protocol strategies, not a comparison with no analgesic treatment.
Primary question
The primary question was whether pain-score change from emergency-department arrival to discharge differed between the two approaches. Pain was measured on a 100 mm visual analogue scale, and change was calculated as arrival score minus discharge score. Thus, a larger positive change represented a greater reduction in pain score during the visit. The timeframe extended from arrival to discharge, up to 6 hours.
3. Trial Design
NCT02222246, titled “Comparing Acute Pain Management Protocols for Patients With Sickle Cell Disease,” is recorded as a completed, randomized, parallel-group, phase 4 pilot study sponsored by Duke University. The registry lists a start date of 2015-03-15, primary completion on 2016-05-31, and enrollment of 106. Masking is listed as none, so the study is described as open-label. The registry record does not report further detail about who knew treatment assignment or whether outcome assessment was blinded.
The two randomized strategies were a standardized, weight-based analgesic protocol and a patient-specific analgesic protocol. Either protocol could use morphine sulfate or hydromorphone. The standardized arm description refers to a protocol based on recent NHLBI recommendations; the patient-specific description says the healthcare team developed a protocol after randomization for use during future emergency-department visits for vaso-occlusive crises, up to 5 visits. The detailed protocol text is incomplete in the registry record, so further operational details are not reported there.
The primary outcome was the difference between arrival and discharge pain scores. Each emergency-department study visit was the unit of analysis for the primary-outcome methods. The arm summaries report 26 observations for the standardized strategy and 26 for the patient-specific strategy. The registry also lists 106 enrolled, but it does not explain the relationship between that enrollment figure, the reported arm summaries, and the number of analyzed visits. The number of visits contributed by each person and how repeated visits were handled in the primary analysis are not fully reported in the registry record.
The registry reports a mixed models analysis for non-inferiority, with p = 0.031, but gives no corresponding effect estimate, confidence interval, non-inferiority margin, or full model description. The trial's original design assumptions, allocation details beyond randomization, and statistical analysis plan are not reported in the registry record. The posted completion status alone does not establish how complete the outcome data were or which participants were included in each analysis.
4. Participant Flow and Accounting
The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.
Before assignment: 106 subject were enrolled and consented to participation during all future Emergency Department (ED) visits for vaso-occlusive crises (VOC), should they occur, over the study period. 53 subjects were randomized to each arm.
Overall Study
| Stage | Standard Dose of Morphine Sulfate or Hydromorphone | Patient Specific Dose of Morphine Sulfate or Hydromorphone |
|---|---|---|
| STARTED | 53 | 53 |
| COMPLETED | 26 | 26 |
| NOT COMPLETED | 27 | 27 |
Reasons for not completing this period:
| Reason | Standard Dose of Morphine Sulfate or Hydromorphone | Patient Specific Dose of Morphine Sulfate or Hydromorphone |
|---|---|---|
| No Emergency Department Visits | 27 | 26 |
| Physician Decision | 0 | 1 |
Analysis population for the primary outcome, as defined in the registry: Each Emergency Department study visit was the unit of analysis for the statistical methods addressing the primary outcome.
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.
| Characteristic | Unit | Standard Dose of Morphine Sulfate or Hydromorphone | Patient Specific Dose of Morphine Sulfate or Hydromorphone | Total |
|---|---|---|---|---|
| Participants analysed | 26 | 26 | 52 | |
| Age, Continuous | years | 27.0 (23.0 to 33.0) | 28.5 (23.0 to 32.0) | 27.0 (23.0 to 32.5) |
| Sex: Female, Male: Female | Participants | 11 | 11 | 22 |
| Sex: Female, Male: Male | Participants | 15 | 15 | 30 |
| Ethnicity (NIH/OMB): Hispanic or Latino | Participants | 2 | 5 | 7 |
| Ethnicity (NIH/OMB): Not Hispanic or Latino | Participants | 21 | 21 | 42 |
| Ethnicity (NIH/OMB): Unknown or Not Reported | Participants | 3 | 0 | 3 |
| Race/Ethnicity, Customized: Race/Ethnicity – African American/Black | Participants | 23 | 23 | 46 |
| Race/Ethnicity, Customized: Race/Ethnicity – Caucasian | Participants | 3 | 3 | 6 |
| Region of Enrollment: United States | Participants | 26 | 26 | 52 |
6. Primary Endpoint
The registry lists the primary outcome as Difference in Pain Score as Measured by a Visual Analogue Scale (VAS), measured over Arrival in ED to discharge from the ED, up to 6 hours.
| Element | Registry entry |
|---|---|
| Result type | Mean |
| Measure of spread | Standard Deviation |
| Unit | Units on a 100 mm VAS |
| Endpoint classification | Continuous Endpoint |
The primary endpoint was change in pain score from emergency-department arrival to discharge. Pain severity was assessed at both time points using a 100 mm visual analogue scale. Its range was 0 to 100, with 0 indicating “no pain” and 100 indicating “pain as bad as it could be” or “worst imaginable pain.” The registry defines the change as arrival score minus discharge score. Consequently, a positive difference indicates that the measured pain score was lower at discharge than at arrival, and a larger positive difference indicates a greater reduction over that visit.
Discharge was defined by whichever occurred first: a decision to admit to hospital, the patient physically leaving the emergency department for home, or six hours of observation. This definition means that the endpoint captures change over the observed emergency-department episode, which could end for different reasons. The registry's timeframe is arrival to discharge, up to 6 hours.
Using a change score focuses the comparison on the difference between two measurements for a visit rather than the pain score at only one time point. As a general statistical point, a change score is interpretable only in relation to how and when both measurements were obtained and how missing measurements were handled. Those missing-data details are not reported in the registry record. The endpoint also represents reported pain on a scale, not every dimension of pain care, treatment experience, or safety. The registry lists several secondary outcomes, including change in pain scores over time and adverse-event measures, but these do not replace the primary arrival-to-discharge comparison.
7. Results Posted in the Registry
- Patient Specific Dose of Morphine Sulfate or Hydromorphone: mean 43.00 versus 26.40 on control; difference 16.60 (95% CI 8.37 to 24.83), standardized effect d = 1.10; z-test p = <0.001.
Mean difference: Patient Specific Dose of Morphine Sulfate or Hydromorphone
95% CI: 8.37 to 24.83
Recalculated from the posted summary statistics
8. Statistical Interpretation
For the standardized protocol, the posted mean pain-score reduction was 26.4 units on the 100 mm VAS, with a standard deviation of 10.6. For the patient-specific protocol, the posted mean reduction was 43.0, with a standard deviation of 18.6. These are visit-level summaries as described for the primary outcome; the registry does not fully explain how visits and people contributed to the arm summaries.
The simple unadjusted recalculation gives a mean difference of 16.6 units, calculated as patient-specific minus standardized. The recalculated 95% confidence interval is 8.371 to 24.83, and the unadjusted p-value is <0.001. In this direction, the positive estimate means the posted patient-specific group summary had a larger mean arrival-to-discharge reduction. The interval describes uncertainty around this recalculated difference under its analysis assumptions. It is not the sponsor's reported model-based result, and it does not account for any features of repeated visits or covariate adjustment that may have been handled in the sponsor's analysis.
The registry separately reports a mixed models analysis labeled non-inferiority, with p = 0.031. It does not report the model's effect estimate, confidence interval, non-inferiority margin, or enough detail to determine the exact hypothesis tested and how the result should be interpreted against that margin. A p-value alone does not state the size or precision of a treatment contrast, and a non-inferiority conclusion depends on a prespecified margin and the direction of the test. Those details are not reported in the registry record. The sponsor's own analysis therefore takes precedence over the simple recalculation, while the missing details limit independent interpretation of that analysis.
The randomized design is relevant because random allocation can support a comparison of strategies, but the registry information does not establish all features needed to assess the analysis. In particular, each emergency-department visit was the stated primary-outcome unit, while patient-specific protocols could apply during future visits. The number and distribution of repeated visits, their contribution to the model, and any correlation adjustments are not fully described. If visits from the same person are correlated, treating them as independent can misstate uncertainty; a suitable mixed model may account for such dependence, but the model details are unavailable.
The posted summaries should therefore be read as evidence about the measured pain-score change among the reported outcome observations, not as a complete account of every enrolled participant or every clinical outcome. The trial was described as a pilot, and the registry record does not give a definitive account of its design assumptions or missing data. No conclusion about the broader clinical value of either strategy follows from the numerical comparison alone. The estimates address one recorded outcome over a defined emergency-department period; other outcomes, patient experiences, and implementation considerations are separate questions.
9. Statistical Methods Explained
What does the mean difference represent?
The primary outcome is continuous: it is a numerical change in VAS pain score. The posted arm values are means, and the spread values are standard deviations. A mean summarizes the average observed score change within an arm; a standard deviation describes variation among observations around that mean. Neither value by itself describes every visit or person. Since the outcome is arrival minus discharge, positive changes correspond to reduced measured pain scores during the visit.
The recalculated comparison subtracts the standardized-arm mean from the patient-specific-arm mean. The result is a difference in VAS units, not a percentage or a ratio. Its sign depends on that chosen order. The recalculated difference of 16.6 is therefore read as a larger average reduction for the patient-specific group among the observations summarized. This simple comparison uses the posted arm summaries and is unadjusted. It is not a substitute for the sponsor's model-based analysis.
How should the confidence interval and p-value be read?
A confidence interval gives a range of estimates compatible with the data and the assumptions of the method. The recalculated 95% interval, 8.371 to 24.83, conveys the uncertainty in the unadjusted mean difference; it does not show the range of individual pain changes. A p-value measures how surprising a result at least this extreme would be under a specified null hypothesis and model assumptions. The recalculated p-value is <0.001. It is not the probability that the null hypothesis is true, nor does it measure clinical importance.
These inferential quantities come from a simple recalculation of posted summaries. The registry's own reported result is a mixed models analysis for non-inferiority with p = 0.031. These are different analyses and the p-values should not be treated as interchangeable. The registry does not report the non-inferiority margin or the sponsor model's estimate and interval, which are needed to interpret the sponsor result fully.
Why can repeated visits matter?
The registry explicitly identifies each emergency-department study visit as the unit of analysis for the primary outcome. A person may contribute more than one visit, and repeated observations from the same person can be correlated because they share characteristics. As a general statistical point, correlation among observations affects the uncertainty of a comparison. A method that recognizes the clustering or repeated-measures structure can produce different standard errors from a method that treats every visit as independent. The sponsor's mixed models label suggests a model-based analysis, but its specification and handling of repeated visits are not reported in the registry record. The recalculation from arm-level summaries cannot independently account for within-person correlation.
What does non-inferiority mean in general?
As a general statistical concept, a non-inferiority analysis asks whether a new strategy is not worse than a comparator by more than a prespecified acceptable margin. The margin, direction of the contrast, and confidence interval are central to interpreting such a test. A p-value alone does not reveal the size of a difference or whether the relevant interval stays within the margin. The registry records a non-inferiority analysis and its p-value, but the margin and detailed result are not reported in the registry record. Accordingly, the reported p-value cannot be translated into a complete non-inferiority interpretation from the registry information alone.
What does variation in the outcome indicate?
The standard deviation is larger for the patient-specific arm summary than for the standardized arm summary. This indicates greater observed spread in the reported change scores for that arm, without explaining why the spread occurred. Pain scores can vary across visits and people, but the registry record does not supply participant-level observations or subgroup summaries that would identify sources of variation. General statistical interpretation should distinguish variability in individual outcomes from uncertainty in the estimated group mean. The standard deviation describes the former; the confidence interval for a mean difference addresses the latter under the calculation's assumptions.
What can be inferred from the design?
Randomization is intended to make assigned groups comparable on average, but it does not remove uncertainty or guarantee that the analyzed observations reflect all randomized participants. The registry lists no masking. It does not report allocation concealment, adherence, crossover, protocol deviations, or a detailed analysis population. With a pain-rating outcome in an open-label study, expectations and measurement context may be relevant in general, but the registry does not establish whether or how they affected these observations. The posted summaries do not allow independent assessment of those possibilities.
Finally, the numerical summaries address the recorded endpoint and timeframe only. Statistical significance does not by itself establish that a difference is meaningful to patients, and a large-looking difference in a pilot dataset does not resolve uncertainty about replication or generalizability. The registry does not report a threshold for an important VAS difference, a sample-size justification, or the full statistical plan. Interpretation should therefore keep the recalculated unadjusted contrast distinct from the sponsor's non-inferiority analysis and from broader clinical conclusions.
10. Confidence Intervals and Precision
The recalculated mean difference is 16.6 VAS units, with a 95% confidence interval from 8.371 to 24.83. The entire interval is on the positive side of zero under the unadjusted calculation. In plain terms, the data and calculation are compatible with a positive difference in mean arrival-to-discharge pain-score reduction for the patient-specific group relative to the standardized group. The width of the interval also shows that the exact size of the difference remains uncertain.
A confidence interval is not a range containing a stated proportion of individual outcomes. It concerns the estimated group contrast under the method's assumptions. Here the interval is recalculated from posted summary statistics and does not incorporate any adjustment for repeated visits, covariates, or the sponsor's mixed model. The registry does not provide the sponsor model's confidence interval, so precision for that reported non-inferiority analysis cannot be assessed from the posted result.
The arm standard deviations, 10.6 and 18.6, show that observed visit-level changes varied around their respective means. Such variation is distinct from the uncertainty in the difference estimate. The registry record does not provide the distributions of individual changes or the number of visits per person, so it is not possible to assess skew, outliers, or clustering from the summaries alone. Precision should thus be interpreted as applying to the simple recalculation and not as a complete description of uncertainty in every possible analysis.
Statistical precision is also different from clinical importance. No threshold for a meaningful VAS difference is reported in the registry record. The interval cannot establish how patients would value the range of differences, and the registry does not report a decision rule connecting this scale change to broader outcomes. The sponsor's p-value and the recalculated interval answer distinct statistical questions and should not be combined into a single claim.
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.
| Effect to detect | Participants per arm, 80% power | Participants per arm, 90% power |
|---|---|---|
| 100% of the observed effect | 14 | 18 |
| 75% of the observed effect | 24 | 32 |
| 50% of the observed effect | 53 | 70 |
The sample-size figures are planning illustrations derived from the observed effect, not the trial's own original design assumptions. The registry record does not report the trial's original sample-size justification or planning assumptions. For this continuous-outcome comparison, the figures are participants per arm. The general assumptions listed for the illustration are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; the target effect is the stated fraction of the observed effect. The note about total events applies to time-to-event planning tables, not these participant-per-arm figures.
The 80% power column gives the planned number of participants per arm for an illustration designed to have an 80% chance of detecting the specified effect under those assumptions, if that effect and model assumptions were correct. The 90% column gives the corresponding planning number for a 90% chance. These are prospective design concepts; they do not describe the probability that the completed trial detected an effect, nor the probability that a result is true.
The table considers planning for 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect. As the target effect becomes smaller, more participants per arm are needed to distinguish it from random variation at a similar power target. This is a general feature of sample-size planning. The figures are not observed power; observed power is not a useful replacement for reading the estimate and its uncertainty after data collection.
Because the observed effect is used as the starting point, these illustrations depend on an estimate from the trial itself. If that estimate is uncertain or unusually large, planning based on it may be optimistic. The registry record does not report an independently justified effect size for a future study. Nor do these simple planning values establish how repeated visits, missing outcomes, or the sponsor's mixed model would affect a future design.
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 z1−α/2 = 1.960) and the desired power (z1−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
- Observed difference Δ = 16.600; pooled standard deviation σ = 15.138.
- n = 2 × (1.96 + 0.842)2 × 15.1382 / 16.6002 = 13.1, rounded up to 14 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 <- 16.6 # observed difference in means
sigma <- 15.138031576133 # 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(2222246)
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 = 14 (80% power), 18 (90% power)
75% of observed effect: n per arm = 24 (80% power), 32 (90% power)
50% of observed effect: n per arm = 53 (80% power), 70 (90% power)
power.t.test: 14.1
Simulated power at n = 14 per arm: 0.798 (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. Analyses Reported by the Sponsor
| Method | Estimate | 95% CI | P-value |
|---|---|---|---|
| Mixed Models Analysis | not reported | 0.031 |
The sponsor's posted analysis is a mixed models analysis labeled non-inferiority and reports p = 0.031 for the two groups. The registry does not give its effect estimate, confidence interval, non-inferiority margin, or full model specification. Those missing details limit independent interpretation of what contrast was tested and whether the reported result meets the intended non-inferiority criterion.
The recalculated comparison is a simple, unadjusted contrast of the posted group means. It estimates a mean difference of 16.6 VAS units, with a 95% confidence interval from 8.371 to 24.83 and an unadjusted p-value of <0.001. It does not reproduce the sponsor's mixed model or its non-inferiority test. The two p-values therefore refer to different analyses and should not be expected to match or be compared as though they tested the same hypothesis.
There is no discrepancy flag for this record, but that does not make the recalculation equivalent to the sponsor's analysis. For interpretation of the trial's own analysis, the sponsor's analysis takes precedence. The recalculation is useful as a transparent, limited check based on the posted summary statistics, while the absent margin and model details remain important limitations.
14. Multiplicity, Missing Data and Analysis Populations
The registry lists 12 outcome measures and 11 posted secondary outcomes. The primary endpoint is change in pain score from arrival to discharge. Secondary outcomes include pain-score change over time and measures of nausea, vomiting, blood-pressure decreases, oxygen desaturation, respiratory distress, and sedation during emergency-department visits. The registry does not report a multiplicity adjustment, a hierarchy among secondary outcomes, or a prespecified strategy for interpreting multiple tests. Without such information, results across several outcomes may be more likely to include chance findings if each is tested separately. This is a general statistical concern, not evidence that any particular reported result is false.
The primary-outcome reporting also has important gaps. The registry identifies the visit as the unit of analysis, but does not report the number of visits per participant, the handling of repeated visits in the mixed model, or whether every randomized participant contributed to the primary analysis. Arm summaries report 26 observations in each group, while overall enrollment is listed as 106. The registry does not explain the relationship among those figures. They should not be treated as interchangeable counts.
Missing-data methods, reasons for missing measurements, and the amount of missing primary-outcome data are not reported in the registry record. Nor does the record specify whether the main analysis followed an intention-to-treat population, a per-protocol population, or another analysis set. These details matter in randomized trials, particularly for non-inferiority questions, because exclusions, nonadherence, or incomplete outcome measurements can affect group comparability and the interpretation of a contrast.
The registry does not provide a complete statistical analysis plan, model formula, covariate list, or a description of sensitivity analyses. It also does not report whether interim analyses were planned or conducted. Those elements are not inferable from the completed status or from the posted p-value. A reader can describe the posted estimates and identify these omissions, but cannot reconstruct the full analysis decisions from the registry record.
15. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Patient Specific Dose of Morphine Sulfate or Hydromorphone | 0 | 26 | 0.0% |
| Standard Dose of Morphine Sulfate or Hydromorphone | 0 | 26 | 0.0% |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Patient Specific Dose of Morphine Sulfate or Hydromorphone | 0 of 26 |
| Standard Dose of Morphine Sulfate or Hydromorphone | 0 of 26 |
The posted serious-adverse-event figures report 0 affected in the patient-specific group and 0 affected in the standardized group, with 26 at risk in each group. The other-adverse-event figures also report 0 affected in each group, with 26 at risk in each. These are the registry's reported counts; they do not mean that every possible symptom or safety outcome was necessarily assessed or recorded in the same way.
The registry also lists secondary measures for nausea, vomiting, blood-pressure decreases, oxygen desaturation, respiratory distress, and sedation. Those outcome names indicate what was considered, but the safety counts alone do not provide the results for each named measure. The registry record does not report deaths in the posted safety fields. It also does not provide further detail about event definitions, collection procedures, follow-up beyond the emergency-department visit, or whether all enrolled participants were included in the risk counts.
Zero reported events should be read as zero events recorded in these posted group summaries, not as proof that no risk exists. Safety interpretation depends on the observation period, completeness of reporting, and the number and type of observations monitored. The record does not provide enough information to assess those features fully.
16. What the Result Does, and Does Not, Mean
The posted primary-outcome summaries describe greater mean pain-score reduction from arrival to discharge in the patient-specific group than in the standardized group. The simple unadjusted recalculation estimates a difference of 16.6 VAS units, with a 95% confidence interval from 8.371 to 24.83 and an unadjusted p-value of <0.001. This is a comparison of the reported group means for the measured endpoint; it is not a complete analysis of all trial data or all possible outcomes.
The sponsor's reported mixed models analysis is labeled non-inferiority and has p = 0.031. Without its margin, effect estimate, confidence interval, and model details, the registry result cannot be fully interpreted as a non-inferiority finding. The unadjusted recalculation does not supply those missing elements and should not be used in their place. The sponsor's analysis takes precedence for the trial's own analysis.
These numbers do not, by themselves, establish the practical importance of the difference. The registry does not report a threshold for a meaningful VAS change, nor does the primary endpoint measure every aspect of pain management or the patient's experience. The outcome is a scale change during an emergency-department episode that ended by a specified discharge rule, up to 6 hours. Results for that endpoint should not be generalized automatically to outcomes outside that period.
The record describes this as a pilot study, and it leaves important analysis details unreported, including missing-data handling and the relationship between enrollment and the primary-outcome observations. It also does not provide enough information to assess repeated-visit handling independently. Accordingly, the posted data can be described as a positive unadjusted mean contrast in the observed summaries, but they do not settle broader questions of generalizability, safety, implementation, or comparative effectiveness.
17. Why This Trial Matters Statistically
This trial raises a statistical issue common in emergency-care research: the outcome may be recorded at the visit level even though randomization and treatment protocols concern people. The registry explicitly names each emergency-department study visit as the primary-outcome unit. A patient-specific protocol could be used during future visits, and observations from the same person may be related. Clarifying the unit of analysis, the number of repeated visits, and how the model handles their dependence is essential for evaluating the estimate's uncertainty.
It also demonstrates why a posted group difference and a sponsor's non-inferiority test need to be kept distinct. The simple recalculation gives an interpretable mean difference and confidence interval from arm summaries. The sponsor reports a mixed models non-inferiority p-value, but without the margin or corresponding estimate and interval, the statistical claim cannot be fully reconstructed. A p-value alone is not enough to communicate the magnitude or precision of a non-inferiority comparison.
Finally, the registry's enrollment figure and its primary-outcome arm summaries are not linked by an explanation of the analysis population or visit counts. This gap matters because readers need to know what observations underlie an estimate. A transparent report of participant flow, repeated measurements, missing data, model specification, and multiplicity procedures would make the result easier to assess. These are statistical reporting questions, not conclusions about the clinical merits of either protocol.
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 primary-outcome summaries report 26 observations for each arm, while the registry lists enrollment of 106. The record does not explain how enrollment relates to the reported arm observations, whether the observations represent participants or visits in the arm summaries, or how many visits were contributed by individuals. Because the primary outcome's unit is each emergency-department visit, this missing information limits assessment of the analysis population and dependence among observations.
The registry reports a mixed models non-inferiority analysis but does not report the model estimate, confidence interval, non-inferiority margin, model specification, or analysis set. The simple recalculation is consequently not a replication of that analysis. The original sample-size justification and trial design assumptions are also not reported in the registry record; planning illustrations based on the observed effect cannot fill that gap.
Other limitations include the absence of reported missing-data methods, outcome completeness, adherence information, and sensitivity analyses. The registry lists no masking, but does not describe outcome-assessment procedures in detail. The endpoint depends on a self-reported pain scale measured at arrival and discharge, and discharge could occur for different reasons, including a decision to admit, leaving for home, or six hours of observation. The registry does not report how these endpoint circumstances were distributed across groups.
The study is described as a pilot, and the available registry information does not establish how representative the observed visits were of other settings or populations. A threshold for clinically meaningful VAS change is not reported. These limitations constrain interpretation of precision and generalizability; they do not change the arithmetic of the posted group summaries.
19. Related Tutorials
20. Related Statistical Calculators
21. Sources
- ClinicalTrials.gov: Comparing Acute Pain Management Protocols for Patients With Sickle Cell Disease — NCT02222246.
- Registry record last updated 2017-08-04; status last verified 2017-06.
- Results first posted 2017-08-04.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
22. Record Summary
NCT02222246 is a completed, randomized, parallel-group, phase 4 pilot study sponsored by Duke University. It compared a patient-specific analgesic protocol with a standardized, weight-based protocol for adults with sickle cell disease receiving emergency-department care for vaso-occlusive pain. The registry lists no masking and enrollment of 106.
The primary endpoint was arrival-to-discharge change in pain on a 100 mm VAS, with positive change indicating reduced pain score. The posted arm means were 26.4 for the standardized strategy and 43.0 for the patient-specific strategy. The recalculated unadjusted mean difference was 16.6, with a 95% confidence interval of 8.371 to 24.83 and p < 0.001. Separately, the registry reports a mixed models non-inferiority analysis with p = 0.031, but does not report its margin or effect estimate.
The registry does not fully explain how enrollment relates to the primary-outcome observations, or report key details about repeated visits, missing data, and analysis populations. These omissions limit independent assessment of the sponsor analysis and the broader interpretation of the result.