A Study to Investigate the Analgesic Efficacy of AZD2423 Compared With Placebo After 28 Days Treatment in Patients With Painful Diabetic Polyneuropathy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Investigate the Analgesic Efficacy of AZD2423 Compared With Placebo After 28 Days Treatment in Patients With Painful Diabetic Polyneuropathy (NCT01201317), a randomized phase 2 study. The primary outcome is Change From Baseline to Days 24-28 in Numerical Rating Scale (NRS) Average Pain Score..

2. Clinical Question

In practical terms, the endpoint compared average-pain ratings near the end of treatment with ratings collected before treatment. Baseline was the mean of Day -5 to Day -1, and the end-of-treatment value was the mean of Day 24 to Day 28. Participants rated average pain during the past 12 hours twice daily. The registry record describes the primary analysis as using last observation carried forward (LOCF) and the intention-to-treat set.

The study included two active dose groups, which were each compared with placebo. These comparisons address the same primary outcome but do not establish that the doses differ from each other. The registry record does not report a direct comparison between the AZD2423 doses. It also does not report a clinically meaningful threshold for a difference on the pain scale, so the numerical estimates alone cannot establish whether a difference would be important to participants.

3. Trial Design

This was a randomized, quadruple-masked, parallel-group Phase 2 study sponsored by AstraZeneca. The registry record lists the study as completed, with an enrollment of 134. The listed start was 2010-09 and the primary completion was 2011-06. The condition listed was neuropathic pain, and the study title identifies the population as patients with painful diabetic polyneuropathy.

Three groups are described: AZD2423 150 mg, AZD2423 20 mg, and placebo. Each was administered as a tablet once daily in the morning. The parallel design means participants were assigned to one group rather than crossing through several treatments. Randomization is listed, but the registry record does not report the sequence-generation method, allocation concealment procedures, or the allocation ratio. It also does not explain which people were masked under the quadruple-masking label.

The registry record describes 28 days of treatment. The primary endpoint spans baseline, defined as the mean of Day -5 to Day -1, through the mean of Day 24 to Day 28. The primary population is the intention-to-treat set. The record does not provide a fuller definition of that set or detail how participants with missing observations were handled beyond identifying LOCF for the primary endpoint.

The record lists five outcome measures, including the primary outcome and four secondary outcomes. It lists no reported analyses. The registry record does not report the original sample-size justification, intended statistical model, hypothesis-testing hierarchy, planned interim analyses, or the trial's original analysis plan. Those omissions limit how fully the trial's design intentions can be reconstructed from the posted record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: The study had an enrolment phase of up to 35 days (including wash-out and baseline periods), a 28-day treatment phase, and a follow-up phase of 7-14 days. Participants were randomly assigned to blinded treatment in a 1:1:1 ratio either to AZD2423 20 mg, AZD2423 150 mg or placebo.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intention-to-treat set (ITT)

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 Change From Baseline to Days 24-28 in Numerical Rating Scale (NRS) Average Pain Score., measured over Baseline (mean of Day -5 to Day -1) to the mean of Day 24 to Day 28.

The primary endpoint was change from baseline to Days 24-28 in Numerical Rating Scale (NRS) average pain score. Baseline was the mean rating from Day -5 to Day -1, and the later value was the mean rating from Day 24 to Day 28. Participants rated average pain intensity during the past 12 hours twice daily on a 0-10 scale, where 0 meant no pain and 10 meant worst pain imaginable. The registry record states that LOCF was used and identifies the intention-to-treat set as the analysis population.

A change score places each participant's later pain rating in relation to their own baseline rating. A negative group mean therefore represents a reduction in average score from baseline; it is not itself the contrast between AZD2423 and placebo. The treatment comparison is the difference between group mean changes. A positive contrast in the posted comparisons means the AZD2423 group's mean change was numerically less negative than the placebo group's mean change. The comparisons do not show individual trajectories or explain why a participant's score changed.

Average pain over repeated ratings can reduce reliance on a single rating at one moment. The registry record does not state why this outcome was selected over the other pain measures, nor does it specify a threshold for a clinically important change. The endpoint is therefore useful for comparing the reported average changes, but interpretation of practical importance requires context not reported in the registry record.

7. Results Posted in the Registry

• AZD2423 150 mg: mean -1.35 versus -1.61 on control; difference 0.26 (95% CI -0.64 to 1.16), standardized effect d = 0.13; z-test p = 0.569.
• AZD2423 20 mg: mean -1.50 versus -1.61 on control; difference 0.11 (95% CI -0.78 to 1.00), standardized effect d = 0.05; z-test p = 0.808.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Change From Baseline to Days 24-28 in Numerical Rating Scale (NRS) Worst Pain Score.

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • AZD2423 150 mg: mean -1.57 versus -1.80 on control; difference 0.23 (95% CI -0.73 to 1.19), standardized effect d = 0.10; z-test p = 0.640.
  • AZD2423 20 mg: mean -1.66 versus -1.80 on control; difference 0.14 (95% CI -0.86 to 1.14), standardized effect d = 0.06; z-test p = 0.783.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between AZD2423 150 mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between AZD2423 20 mg and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants With at Least 30% Decrease From Baseline in Numerical Rating Scale (NRS) Average Pain Score at Day 28.

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • AZD2423 150 mg: 34.8% (16/46) versus 38.5% (15/39) on control; risk difference -3.7 percentage points (95% CI -24.2 to 16.9), risk ratio 0.90 (0.52–1.58); unadjusted two-proportion z-test p = 0.725.
  • AZD2423 20 mg: 45.2% (19/42) versus 38.5% (15/39) on control; risk difference 6.8 percentage points (95% CI -14.7 to 28.2), risk ratio 1.18 (0.70–1.97); unadjusted two-proportion z-test p = 0.537.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between AZD2423 150 mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between AZD2423 20 mg and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants With at Least 50% Decrease From Baseline in Numerical Rating Scale (NRS) Average Pain Score at Day 28.

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • AZD2423 150 mg: 21.7% (10/46) versus 25.6% (10/39) on control; risk difference -3.9 percentage points (95% CI -22.1 to 14.3), risk ratio 0.85 (0.39–1.82); unadjusted two-proportion z-test p = 0.673.
  • AZD2423 20 mg: 19.0% (8/42) versus 25.6% (10/39) on control; risk difference -6.6 percentage points (95% CI -24.7 to 11.5), risk ratio 0.74 (0.33–1.69); unadjusted two-proportion z-test p = 0.476.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between AZD2423 150 mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between AZD2423 20 mg and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary outcomes cover several related but distinct aspects of pain: worst pain over the measurement period, threshold-based reductions in average pain, and neuropathic pain symptoms. Across the available results, the between-group estimates are not consistently in one direction. Some active-group summaries are numerically closer to placebo, while the threshold-based comparisons differ by dose and threshold. The unadjusted confidence intervals for the posted comparisons are broad and include values on either side of no difference.

This overall pattern does not provide a clear, consistent signal that adds precision to the primary comparison. Worst-pain change is conceptually related to the primary average-pain outcome but uses a different rating. Threshold outcomes recast change as a participant count meeting a specified decrease, rather than preserving the full scale of scores. Those outcome definitions can lead to different patterns even when they concern the same underlying symptom domain.

Secondary results are generally interpreted with less weight than a prespecified primary outcome, especially when several outcomes and comparisons are considered. In general, multiple unadjusted tests increase the chance that some result may appear notable by chance. The registry record does not report a multiplicity adjustment or a hierarchy that would establish which secondary outcomes were confirmatory. The listed neuropathic pain symptom inventory outcome has no posted result in the record. Consequently, the secondary outcomes offer descriptive context but do not resolve the uncertainty in the primary comparison or establish a separate confirmatory conclusion.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Change From Baseline to Days 24-28 in Numerical Rating Scale (NRS) Worst Pain Score.
• Number of Participants With at Least 30% Decrease From Baseline in Numerical Rating Scale (NRS) Average Pain Score at Day 28.
• Number of Participants With at Least 50% Decrease From Baseline in Numerical Rating Scale (NRS) Average Pain Score at Day 28.
• Change From Baseline to Day 29 in Neuropathic Pain Symptom Inventory Scale (NPSI) Total Score.

9. Statistical Interpretation

10. Statistical Methods Explained

The registry record identifies the primary outcome as a continuous measure summarized by a mean and standard deviation. A continuous score can take values across a scale, rather than only categories such as yes or no. The primary summary is a mean change from baseline, while the standard deviation describes how spread out participant-level changes were within each group. Standard deviation is not the uncertainty in the group mean; confidence intervals for contrasts address uncertainty in the estimated group difference.

How the treatment contrasts are defined

Each posted primary comparison is a mean difference between an AZD2423 group and placebo. The point estimates are 0.26 for AZD2423 150 mg versus placebo and 0.11 for AZD2423 20 mg versus placebo. Because these are contrasts of changes, they should be distinguished from the individual group means. The group mean changes were -1.35, -1.5, and -1.61 for AZD2423 150 mg, AZD2423 20 mg, and placebo, respectively. A positive mean difference corresponds to an AZD2423 mean change that is less negative than the placebo mean change.

The reported confidence intervals for these comparisons are -0.635 to 1.155 and -0.779 to 0.999, respectively. In general, a confidence interval describes the range of values reasonably compatible with the data under the statistical method used. A wide interval indicates limited precision. When an interval crosses zero for a difference, the data do not distinguish the estimate from no difference at the corresponding confidence level. This does not prove that the true difference is zero.

What the p-values express

The posted p-values are unadjusted: 0.569 for AZD2423 150 mg versus placebo and 0.808 for AZD2423 20 mg versus placebo. In general, a p-value measures how unusual results at least as discrepant as the observed data would be under a specified null hypothesis and statistical model. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of a difference. These values should be considered alongside the point estimates and confidence intervals, not in isolation.

The registry record lists no reported analyses, while the displayed contrasts are recalculations from posted arm-level summaries. They are simple unadjusted comparisons and should not be treated as a substitute for a sponsor's planned analysis. The record does not report the original model, covariate adjustments, or multiplicity procedure. The discrepancy indicator is false, so there is no noted disagreement with a sponsor analysis in the record.

Standardized effects and variability

The posted comparisons also include standardized effect estimates, d, of 0.126 for AZD2423 150 mg and 0.054 for AZD2423 20 mg. In general, standardizing an effect expresses a mean difference relative to a measure of variability, which can help describe the magnitude on a scale separate from the original score units. Such a standardized estimate is still uncertain and does not by itself determine clinical importance. The posted pooled standard deviations are 2.068 and 2.021, respectively; the comparisons use the observed variation in the groups rather than assuming no variability.

In general, greater variability makes a group mean harder to estimate precisely when the number of observations is fixed. That is one reason standard deviations and confidence intervals matter alongside means. Conversely, the same apparent mean difference may have different precision in studies with different numbers of participants or different variability. No additional statistical model details are reported in the registry record, so these general principles should not be mistaken for a description of a more elaborate analysis.

Missing observations and LOCF

The primary outcome description specifies last observation carried forward. In general, LOCF substitutes a participant's last recorded observation for a later missing observation, carrying that value forward to the analysis time point. This is an assumption about the unobserved value, not a measurement of it. LOCF can affect estimates if participants with missing observations differ from those with complete observations or if scores would have changed after the last measurement. The registry record does not report the amount or pattern of missing data, reasons for missingness, or sensitivity analyses that would show how conclusions might change under alternative assumptions.

The record identifies the intention-to-treat set for the primary endpoint, but does not give a detailed definition of membership in that set. The primary group summaries have participant counts of 46, 42, and 39. Reasons for the counts and any exclusions are not reported in the registry record. A reader should not infer the amount or cause of missingness from those counts alone.

Interpreting secondary outcomes

The secondary outcomes include a continuous worst-pain score, participant counts crossing average-pain decrease thresholds, and a neuropathic pain symptom inventory score. The registry record posts results for the first three of these secondary outcomes, but the listed neuropathic pain symptom inventory outcome has no posted results in the record. For the posted secondary outcomes, the comparisons are also unadjusted recalculations from arm-level summaries. They do not come with reported analyses or a documented testing hierarchy. General statistical practice treats multiple unadjusted comparisons cautiously because examining several outcomes can increase the chance of finding an apparently notable result by chance. The record does not report whether a multiplicity adjustment was planned.

The sample-size illustrations use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. They vary the effect to detect as a stated fraction of the observed effect. These are planning illustrations based on the observed estimates, not the trial's original design assumptions. In general, over-dispersion means variability exceeds what a chosen model expects; if present, it can make standard errors too small and confidence intervals too narrow unless the analysis accounts for it. The registry record does not report an over-dispersion assessment, and the primary outcome is continuous, so this general issue should not be taken as a claim about the trial's actual analysis.

11. Confidence Intervals and Precision

The primary comparisons have confidence intervals that extend from negative to positive values. For AZD2423 150 mg versus placebo, the mean difference was 0.26, with a 95% confidence interval from -0.635 to 1.155. For AZD2423 20 mg versus placebo, the mean difference was 0.11, with a 95% confidence interval from -0.779 to 0.999. Both intervals include zero, the value corresponding to no difference in mean change between the compared groups.

In general, a confidence interval describes uncertainty around an estimate under the statistical method used. The data are compatible with a range of possible contrasts, not just the point estimate. Because both intervals include values on either side of zero, the posted summaries do not pin down the direction of a difference with precision. They also do not demonstrate equivalence: equivalence requires a prespecified margin and an analysis designed to evaluate whether the interval lies within that margin. Such a margin is not reported in the registry record.

The arm standard deviations for the primary outcome range from 1.77 to 2.26, while the analyzed group counts range from 39 to 46. These summaries provide context for why precision is limited: the participant-level scores vary, and the analyzed groups are modest in size. The interval width should be considered alongside the estimate, rather than reading the point estimate as an exact treatment effect.

The p-values are unadjusted and do not replace the confidence intervals. The intervals also do not address every source of uncertainty, including possible bias from missing observations or choices in the analysis plan that are not reported. The registry record does not provide the sponsor's inferential analyses, so the recalculated intervals are best understood as unadjusted estimates based on the posted summary statistics.

12. 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 sample-size figures are planning illustrations, not counts of participants enrolled in this trial. For these non-time-to-event outcomes, the figures are participants per arm; the time-to-event exception described in the planning note does not apply to the pain score outcomes. The illustrations assume a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. They use the observed effect as a starting point and show planning for 100%, 75%, or 50% of that observed effect.

The 80% power column gives the participants per arm estimated to provide an 80% chance of detecting the stated effect under those planning assumptions, if that effect is the true one and the assumptions hold. The 90% power column gives the corresponding estimate for a 90% chance. These are prospective design calculations, not an estimate of the probability that the trial's result is correct. This is not observed power.

The displayed values are tied to the particular observed contrast and its variability, so they can be very large when the estimated effect is small relative to the spread in the outcome. Planning for a smaller fraction of the observed effect requires more participants per arm than planning for the full observed effect. The figures are illustrations based on the posted results; the trial's original sample-size justification and design assumptions are not reported in the registry record. They should not be read as the number of participants that the investigators actually planned to enroll or as evidence that the trial was designed around these estimates.

13. 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 = 0.260; pooled standard deviation σ = 2.068.
2. n = 2 × (1.96 + 0.842)² × 2.068² / 0.260² = 992.7, rounded up to 993 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 <- 0.26                      # observed difference in means
sigma <- 2.06763149521379                     # 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(1201317)
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 = 993 (80% power), 1,330 (90% power)
 75% of observed effect: n per arm = 1,765 (80% power), 2,363 (90% power)
 50% of observed effect: n per arm = 3,971 (80% power), 5,317 (90% power)
power.t.test: 993.7 
Simulated power at n = 993 per arm: 0.797 (20000 replicates)

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

14. Multiplicity, Missing Data and Analysis Populations

The registry record lists one primary outcome and four secondary outcomes, but reports no analyses. It does not describe a hierarchy for testing the primary and secondary outcomes, a family-wise error procedure, or another approach to multiplicity. The posted comparisons are unadjusted recalculations from summary statistics. When several comparisons are examined, unadjusted p-values do not account for the increased chance of an apparently notable finding across the full set. This matters especially for the secondary outcomes, which use different measures and thresholds.

The record identifies the intention-to-treat set for the primary outcome. It does not give a detailed definition of that set, identify the analysis populations for each secondary outcome, or report how participants were handled in each analysis. The primary outcome description names LOCF, but the amount and pattern of missing data, reasons for missingness, and results under alternative missing-data assumptions are not reported in the registry record. The posted participant counts alone do not explain why the counts differ across groups or from the overall enrollment.

The original statistical analysis plan, sample-size justification, interim analyses, covariate adjustments, and sponsor-reported inferential results are not reported in the registry record. It is therefore not possible to assess from the record whether the posted comparisons were the originally planned analyses or whether the endpoint definitions and analysis choices were finalized before results were known. These omissions do not establish that a particular practice was or was not used; they limit what can be concluded from the registry entry.

In general, a prespecified analysis plan helps distinguish confirmatory tests from exploratory ones and documents how missing data and multiple outcomes will be addressed. Without those details, the posted estimates can still be described, but their inferential status is less clear. The unadjusted secondary findings should be treated as exploratory context rather than independent confirmation of the primary result.

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 4% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: asthma (2.3%), clostridium difficile colitis (2.3%), myocardial infarction (2.3%), sudden cardiac death (2.3%).

The registry record lists serious adverse event figures for 47 participants in the AZD2423 150 mg group, 44 in the AZD2423 20 mg group, and 41 in the placebo group. The corresponding counts with serious adverse events are 0, 3, and 0. These figures describe the recorded counts in each group; they do not by themselves establish that an event was caused by a study treatment. The registry record does not report deaths in these groups; death counts are not reported in the registry record.

The record also lists 14, 6, and 11 participants with other adverse events in the respective groups. These are separate from the serious adverse event figures and should not be combined with them. The listed top serious adverse event terms are asthma, Clostridium difficile colitis, myocardial infarction, and sudden cardiac death, each reported as 2.3%. The record does not provide enough context alongside those terms to assign them to a particular treatment group from this listing.

These counts are descriptive. The registry record does not report a formal statistical comparison of safety outcomes or enough detail to interpret differences as treatment-related risks. Small group counts also mean that the figures alone provide limited information about uncommon events.

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

The posted primary results describe mean changes in average pain scores over the specified treatment period and unadjusted comparisons of each AZD2423 dose with placebo. The point estimates for both comparisons are positive, and their confidence intervals include zero and extend in both directions. This means the posted summaries do not precisely establish the direction or size of a difference between either active group and placebo.

That uncertainty is not the same as proof that the groups were identical. A result that does not distinguish a contrast from zero can reflect a small effect, limited precision, or both. Establishing equivalence or ruling out clinically important differences would require a prespecified clinical margin and an analysis designed for that purpose; neither is reported in the registry record. No threshold for a clinically meaningful change is reported either.

The mean change in each group is also not a standalone estimate of treatment effect. Participants in the placebo group had a negative mean change, as did participants in both active groups. The between-group comparisons are the relevant summaries for assessing how the changes differed, but those estimates remain uncertain. Secondary outcomes add related descriptive information, but their unadjusted status and the lack of a reported testing hierarchy make them less suitable as independent confirmation.

The findings apply to the outcomes, groups, and follow-up described in the registry record. They do not establish what would happen in other populations or over longer periods. The record does not report the sponsor's inferential analyses, original sample-size rationale, detailed missing-data pattern, or a complete statistical plan. Those gaps limit assessment of robustness and of the intended interpretation. The posted data can be described statistically, but they do not support medical advice or a conclusion about whether treatment should be used.

17. Why This Trial Matters Statistically

This trial illustrates why a treatment comparison should focus on the contrast between groups rather than on change within each group. All three groups had negative mean changes in the primary pain score, but those within-group changes do not isolate the treatment effect. The mean differences versus placebo are the comparisons that address the randomized question.

It also shows how point estimates and precision need to be read together. The primary estimates are close to zero relative to their confidence intervals, and both intervals span zero. A reader who looks only at the point estimates could overstate their certainty; a reader who treats an interval crossing zero as proof of no effect would also overstate what the data establish. The interval conveys a range of plausible contrasts under the calculation used.

Another statistical issue is the distinction between a registry's arm-level results and a planned inferential analysis. The record includes arm summaries and permits simple unadjusted recalculations, but lists no reported analyses. It does not document the original testing hierarchy or multiplicity approach. This makes the results useful for descriptive comparison while limiting conclusions about confirmatory testing.

Finally, the sample-size illustrations show how design requirements depend on the effect targeted: planning for a smaller effect requires more participants. These illustrations are based on observed results and are not the trial's original justification or an assessment of observed power. Together, these features make the study a useful example of interpreting continuous outcomes, uncertainty, multiple endpoints, and the limits of registry-reported analyses.

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 uses LOCF, but the record does not report how many observations were carried forward, why data were missing, or whether alternative missing-data analyses were performed. LOCF relies on an assumption about later unobserved scores, and the available record does not permit assessment of how sensitive the results are to that assumption.

The analyzed group counts for the primary outcome are 46, 42, and 39, while the registry lists overall enrollment of 134. Reasons for the group counts and any exclusions or unavailable outcome measurements are not reported in the registry record. It would be inappropriate to infer a specific attrition pattern or cause from these figures alone.

The record lists no sponsor-reported analyses and does not provide the original statistical analysis plan, sample-size justification, allocation details, or a detailed definition of the intention-to-treat set. The posted comparisons are unadjusted recalculations from summary statistics. This limits the ability to verify whether they align with the investigators' prespecified methods, and the unadjusted secondary comparisons should be interpreted cautiously.

The primary endpoint is measured over Days 24 to 28, and the study title describes a 28-day treatment period. The registry record does not provide longer-term results. It also does not report a threshold for clinically meaningful change, so the practical importance of a numerical contrast cannot be judged from the record alone. The enrollment and analyzed group counts also limit precision, as reflected in the broad primary confidence intervals.

Safety data are presented as counts, without a reported formal comparison or sufficient context to attribute listed serious event terms to a particular group. Death counts are not reported in the registry record. These limitations constrain interpretation of both benefit-related outcomes and safety descriptions.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: A Study to Investigate the Analgesic Efficacy of AZD2423 Compared With Placebo After 28 Days Treatment in Patients With Painful Diabetic Polyneuropathy — NCT01201317.
• Registry record last updated 2014-04-25; status last verified 2014-04.
• Results first posted 2013-09-25.
• 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

This completed randomized, quadruple-masked, parallel-group Phase 2 study compared AZD2423 150 mg, AZD2423 20 mg, and placebo in painful diabetic polyneuropathy. Overall enrollment was 134. The primary outcome was change in NRS average pain score from baseline to the mean of Days 24 to 28, analyzed in the intention-to-treat set with LOCF as described in the registry record.

The posted primary comparisons are unadjusted recalculations from arm-level summary statistics. Their confidence intervals include zero, and the registry record lists no reported analyses. Secondary outcomes provide additional descriptive measures, but no multiplicity procedure or analysis hierarchy is reported. The original design assumptions, detailed missing-data information, and longer-term results are not reported in the registry record.