A Research Study of an Investigational Drug ALO-02 (Oxycodone Hydrochloride and Naltrexone Hydrochloride) in Patients With Moderate to Severe Chronic Low Back Pain: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Research Study of an Investigational Drug ALO-02 (Oxycodone Hydrochloride and Naltrexone Hydrochloride) in Patients With Moderate to Severe Chronic Low Back Pain (NCT01571362), a randomized phase 3 study. The primary outcome is Change in Weekly Average Electronic Diary (eDiary) Numeric Rating Scale -Pain (NRS-Pain) Score From Randomization Baseline to Final 2 Weeks (Average of Weeks 11 and 12).

2. Clinical Question

The trial record describes a phase 3, multicenter, randomized, double-blind, placebo-controlled withdrawal study. Its stated primary objective was to determine the analgesic efficacy and safety of ALO-02 compared with placebo in people with moderate to severe chronic low back pain. The endpoint focuses on change in pain scores rather than a single measurement at the end of treatment. That approach allows the comparison to reflect pain scores recorded repeatedly in daily diaries and then averaged over the specified period.

The key distinction is between continuing ALO-02 and withdrawing to placebo in the randomized period. This is not a direct comparison of starting ALO-02 with never receiving it: the treatment sequences are described as ALO-02 to ALO-02 and ALO-02 to placebo. Details of the preceding treatment period and the criteria used to enter randomization are not reported in the registry record. The question therefore needs to be read in the context of the withdrawal design, rather than as a general comparison of initial treatment choices.

3. Trial Design

The registry identifies NCT01571362 as a completed phase 3 study. The official title describes a multicenter, 12-week, double-blind, placebo-controlled, randomized withdrawal study of extended-release ALO-02 capsules for moderate to severe chronic low back pain. Allocation was randomized, masking was double, and the model was parallel. The registry lists enrollment as 410. The study start is recorded as 2012-06 and primary completion as 2013-06.

The two randomized sequences were ALO-02 to placebo and ALO-02 to ALO-02. The primary outcome population is described as the intent-to-treat population comprising participants randomized into the double-blind treatment period who received at least one dose of study drug after randomization. The record also identifies an open ALO-02 safety group. The precise timing and procedures for any open-label period, eligibility for randomization, and treatment adherence are not reported in the registry record.

The primary endpoint was the change in weekly average eDiary NRS-Pain score from randomization baseline to the final two weeks, averaging Weeks 11 and 12. The study also recorded disability, patient global assessment, pain-response thresholds, and Brief Pain Inventory outcomes. The record lists 101 outcome measures and 159 analyses, while 100 secondary outcomes are posted in the analyzed record. These counts describe the breadth of recorded outcomes and analyses; they do not establish how many were designated as confirmatory or how the investigators controlled error from multiple testing.

Although the record labels the design randomized withdrawal, it does not report the original sample-size justification, the planned allocation ratio, detailed analysis plan, or interim analysis procedures. Those details should not be inferred from the observed group sizes or from the illustrative planning calculations.

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: A total of 410 participants were enrolled into the Open-Label Conversion and Titration Period and 281 participants were randomized into the Double-Blind Treatment Period, of which, 280 participants received study treatment.

Open-Label Conversion + Titration Period

Reasons for not completing this period:

Double-Blind Treatment Period

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent-to-Treat (ITT) Population: all participants who were randomized into the Double-Blind Treatment Period and received at least 1 dose of study drug after randomization; the averaged value for each participant from the 100 imputed datasets were used. Hybrid multiple and single imputation were applied, depending on reason for discontinuation.

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 in Weekly Average Electronic Diary (eDiary) Numeric Rating Scale -Pain (NRS-Pain) Score From Randomization Baseline to Final 2 Weeks (Average of Weeks 11 and 12), measured over Weeks 11 and 12.

The primary endpoint is change in weekly average electronic-diary Numeric Rating Scale pain score from randomization baseline to the final two weeks of double-blind treatment, defined as the average of Weeks 11 and 12. The diary score is based on an 11-point scale, from 0 for no pain to 10 for worst possible pain; higher scores indicate greater pain. Daily scores were averaged over the last 7 days to derive each weekly average.

Averaging repeated diary entries can represent the participant’s pain over a period rather than relying on a single day’s report. Comparing change from randomization baseline also accounts for the starting score in describing the treatment-period trajectory. The endpoint is continuous, so the reported group summaries are least-squares means with standard errors, and the sponsor’s analysis compares adjusted least-squares means using ANCOVA.

The endpoint is clinically interpretable as a change in reported pain intensity, but it is not itself a direct measure of function, overall well-being, or safety. Those domains have separate outcomes in the registry, including the Roland-Morris Disability Questionnaire and Patient Global Assessment. The record’s endpoint description does not specify the sign convention used to construct change. The observed estimates should therefore be interpreted alongside the scale direction and the reported comparison, not by treating the sign alone as a universal indicator of improvement.

7. Results Posted in the Registry

• ALO-02 To ALO-02: mean 0.60 versus 1.23 on control; difference -0.63 (95% CI -1.11 to -0.15), standardized effect d = -0.31; z-test p = 0.010.

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 in Roland-Morris Disability Questionnaire (RMDQ) Total Score From Randomization Baseline to the End of Double-Blind Week 12 (or Final Visit).

• 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 errors.
• Result:
  • ALO-02 To ALO-02: mean 0.67 versus 0.50 on control; difference 0.17 (95% CI -0.93 to 1.27), standardized effect d = 0.04; z-test p = 0.763.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between ALO-02 To ALO-02 and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Participants With Improvement in Weekly Average eDiary NRS-Pain Scores From Screening to Final 2 Weeks of the Double-Blind Treatment Period (Average of Weeks 11 and 12) by Cumulative Percent Reduction of Greater or Equal to (≥) 20%

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • ALO-02 To ALO-02: 62.3% (91/146) versus 48.5% (65/134) on control; risk difference 13.8 percentage points (95% CI 2.3 to 25.4), risk ratio 1.28 (1.04–1.59); unadjusted two-proportion z-test p = 0.020.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for ALO-02 To ALO-02 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Percentage of Participants With Improvement in Weekly Average eDiary NRS-Pain Scores From Screening to Final 2 Weeks of the Double-Blind Treatment Period (Average of Weeks 11 and 12) by Cumulative Percent Reduction ≥30%

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • ALO-02 To ALO-02: 57.5% (84/146) versus 44.0% (59/134) on control; risk difference 13.5 percentage points (95% CI 1.9 to 25.1), risk ratio 1.31 (1.03–1.66); unadjusted two-proportion z-test p = 0.024.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for ALO-02 To ALO-02 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The posted secondary outcomes cover disability, patient global assessment, pain-response thresholds, and Brief Pain Inventory measures. The available results show a mixed picture across the measures that have detailed results posted: the disability comparison is imprecise and its sponsor-reported confidence interval includes zero, while the posted pain-response threshold analyses favor the continued-ALO-02 sequence in their reported percentages and have sponsor-reported p-values below 0.05. The registry lists many additional outcomes, but detailed results for every listed outcome are not included in the registry record for this analysis.

This pattern is broadly compatible with the primary endpoint’s direction for pain-related measures, but it is not a substitute for the primary analysis. Secondary outcomes may capture related but distinct constructs: a continuous average pain score, crossing a defined improvement threshold, disability, and patient global assessment do not measure exactly the same thing. Agreement across related measures can add context, while differences may reflect the outcome definition, measurement variability, or analysis population.

Unadjusted secondary findings carry less weight because multiple outcomes and analyses create more opportunities for apparently notable results by chance. The registry lists 101 outcome measures and 159 analyses, and 100 secondary outcomes are posted. The record does not report a prespecified hierarchy, a family-wise error strategy, or another multiplicity adjustment for these results. Without that information, individual secondary p-values should be treated as supportive or exploratory rather than as independent confirmatory evidence. This caution does not imply that the results are false; it reflects uncertainty about how the full set of tests was planned and interpreted.

The secondary findings also do not establish effects on outcomes whose results are not reported in the registry record. In particular, outcome titles alone cannot be used to infer results for patient global assessment or Brief Pain Inventory measures. The primary endpoint remains the clearest focal comparison, with secondary results offering qualified context rather than replacing it.

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

• Change in Roland-Morris Disability Questionnaire (RMDQ) Total Score From Randomization Baseline to the End of Double-Blind Week 12 (or Final Visit).
• Percentage (%) of Participants With Shift in Patient Global Assessment (PGA) by Category With Baseline PGA Score of Very Good (1), Good (2), Fair (3), Poor (4), Very Poor (5) From Randomization Baseline to End of Double-Blind Week 12 (or Final Visit).
• Percentage of Participants With Improvement in Weekly Average eDiary NRS-Pain Scores From Screening to Final 2 Weeks of the Double-Blind Treatment Period (Average of Weeks 11 and 12) by Cumulative Percent Reduction of Greater or Equal to (≥) 20%
• Percentage of Participants With Improvement in Weekly Average eDiary NRS-Pain Scores From Screening to Final 2 Weeks of the Double-Blind Treatment Period (Average of Weeks 11 and 12) by Cumulative Percent Reduction ≥30%
• Percentage of Participants With Improvement in Weekly Average eDiary NRS-Pain Scores From Screening to Final 2 Weeks of the Double-Blind Treatment Period (Average of Weeks 11 and 12) by Cumulative Percent Reduction ≥40%
• Percentage of Participants With Improvement in Weekly Average eDiary NRS-Pain Scores From Screening to Final 2 Weeks of the Double-Blind Treatment Period (Average of Weeks 11 and 12) by Cumulative Percent Reduction ≥50%
• Change From Screening Period to End of Open-Label Treatment in Brief Pain Inventory - Short Form (BPI-sf): Worst Pain, Least Pain, Average Pain, Pain Right Now, Pain Severity Index, Pain Interference Index
• Change From Screening Period to Randomization Baseline in BPI-sf: Worst Pain, Least Pain, Average Pain, Pain Right Now, Pain Severity Index, Pain Interference Index

9. Statistical Interpretation

10. Statistical Methods Explained

What does the primary comparison estimate?

The primary result compares average change in pain scores between two randomized treatment sequences. A least-squares mean is a model-based estimate of a group mean, often adjusted for variables included in the analysis model. The registry reports ANCOVA and a difference of least-squares means for the primary comparison. The covariates and other model specifications are not reported in the registry record, so the adjustment cannot be reconstructed from the posted summary alone.

The sponsor’s reported difference is -0.62, with a 95% confidence interval from -1.11 to -0.14 and p = 0.011. The unadjusted recalculation uses the posted group summaries and gives -0.63 with p = 0.01. A recalculation based on summary statistics is useful as a consistency check, but it cannot incorporate information or adjustments unavailable in those summaries. The sponsor’s own analysis takes precedence for the trial’s reported inference.

Why use an analysis of covariance?

In general, analysis of covariance models an outcome while accounting for selected covariates. For a change-from-baseline endpoint, an adjusted model can help compare groups while incorporating measured baseline or design factors, if those are included in the prespecified model. The registry identifies ANCOVA but does not report the complete model specification or its covariates. It would therefore be inappropriate to claim which factors were adjusted for or to attribute the estimate to any particular adjustment.

What does a p-value say?

In general, a p-value measures how compatible the observed data are with a specified null model, under the assumptions of the analysis. It is not the probability that the null hypothesis is true, nor a measure of the size or clinical importance of an effect. The primary sponsor-reported p-value is 0.011. The recalculated unadjusted p-value is 0.01. Their similarity is reassuring as a numerical check, but it does not make the two analyses methodologically identical.

For the binary secondary outcomes, the registry reports Cochran-Mantel-Haenszel analyses. In general, this method can compare outcomes across treatment groups while accounting for stratification. The stratification variables and detailed model implementation are not reported in the registry record. The sponsor-reported p-values for the two posted pain-response thresholds are 0.021 and 0.025. The posted record does not identify a multiplicity adjustment for these analyses.

What are the unadjusted recalculations?

The recalculated continuous comparison uses the posted least-squares mean summaries and standard errors, with a pooled standard deviation described as approximate. The recalculated binary comparisons use the posted participant counts and proportions to estimate an unadjusted risk difference. These are simple, unadjusted calculations from summary information, not a reproduction of the sponsor’s complete analysis. They do not account for unavailable participant-level information, covariates, stratification, or imputation procedures beyond what the summary statistics encode.

For the primary endpoint, the recalculated mean difference is -0.63, compared with the sponsor’s ANCOVA estimate of -0.62. For the secondary disability outcome, the recalculated difference is 0.17, while the sponsor reports 0.18. These close estimates do not prove that the underlying models are identical. For the binary pain-response outcomes, the recalculated risk differences are 0.138 and 0.135; the corresponding sponsor analyses report p-values but no effect estimate or confidence interval in the posted results.

What is the role of the scale and measurement period?

The NRS-Pain score is bounded by its stated anchors, with higher values indicating greater pain. A weekly average is derived from daily ratings over the last 7 days, and the primary outcome averages Weeks 11 and 12. In general, repeated patient-reported measurements can reduce reliance on one potentially atypical measurement, but they also depend on diary completion and the analytic approach to missing entries. Diary completion rates and rules for forming weekly averages when data are incomplete are not reported in the registry record.

The registry states that each participant’s averaged value from 100 imputed datasets was used, and describes the approach as hybrid multiple and single imputation. This establishes that imputation was part of the primary endpoint analysis, but the description is truncated and does not give enough detail to evaluate the imputation assumptions, variables, or sensitivity analyses. The missing-data mechanism assumed by the analysis is not reported in the registry record.

How should the analysis population be understood?

The primary endpoint population is called the intent-to-treat population in the registry, but its operational definition is randomized participants who received at least one dose after randomization. This is narrower than a definition that includes every randomized participant regardless of post-randomization dosing. The number of randomized participants who did not receive a dose and their outcomes are not reported in the registry record. The distinction matters because analysis populations define which participants contribute to the estimate and may affect how broadly the result applies.

In general, randomization supports a fair comparison by balancing measured and unmeasured factors on average at assignment, but post-randomization exclusions or missing outcome data can complicate that interpretation. The record does not provide enough information to assess the extent of missing outcome data or the robustness of the result to alternative analysis populations. These limitations do not erase the reported result; they define the information available for judging how it was obtained.

11. Confidence Intervals and Precision

The sponsor’s primary estimate is a difference of least-squares means of -0.62, with a 95% confidence interval from -1.11 to -0.14. A confidence interval describes a range of values that are compatible with the data and the statistical model at the stated confidence level. It is not a guarantee that the true effect lies within the particular interval, nor does it describe the range of individual participant responses.

The interval does not include zero, and the sponsor reports p = 0.011. The bounds are on the change-score scale, so they express uncertainty in the group contrast in NRS-Pain units. The interval spans estimates with different magnitudes; the point estimate alone should not be read as exact. No threshold for a clinically important difference is reported in the registry record, so the interval cannot be classified against such a threshold from the record alone.

The recalculated primary estimate is -0.63, close to the sponsor’s -0.62. The recalculation uses summary statistics and is unadjusted; it should not replace the sponsor’s model-based interval. For the disability outcome, the sponsor reports a 95% confidence interval from -0.94 to 1.29, which includes zero. For the binary secondary outcomes, the posted sponsor analyses do not include confidence intervals or effect estimates, limiting direct assessment of their precision from those analyses.

In general, precision depends on factors such as variability, the amount of information available, and the analysis method. Standard errors are reported alongside the primary group estimates, but the record does not report enough information to assess all sources of uncertainty, including missing-data assumptions and model sensitivity. Confidence intervals quantify sampling and model uncertainty under the stated analysis; they do not resolve limitations in endpoint selection, multiplicity, or external applicability.

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 planning illustrations report participants per arm, not total participants, for the continuous and binary outcomes in this record. The planning assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. Each illustration uses a target effect equal to the stated fraction of the observed effect. The trial’s own original sample-size assumptions and justification are not reported in the registry record.

The 80% power column represents the estimated participants per arm for a future study to have an 80% chance of detecting the specified effect if that effect is present and the planning assumptions hold. The 90% column describes the corresponding planning target for a 90% chance. These are design calculations, not a guarantee that a future study will produce a significant result. For the primary endpoint, the illustration gives 166 participants per arm when planning for 100% of the observed effect, 294 for 75%, and 661 for 50% at the 80% target. At the 90% target, the corresponding figures are 221, 393, and 884 participants per arm.

Smaller effects are harder to distinguish from random variation, so planning to detect a smaller fraction of the observed effect requires more participants per arm under these assumptions. The illustrations use the observed effect as an input and should not be treated as evidence that the original trial was designed around those assumptions. They are not observed power. For time-to-event outcomes, the stated convention would instead report a total number of events, but the primary endpoint in this record is continuous.

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 = -0.630; pooled standard deviation σ = 2.043.
2. n = 2 × (1.96 + 0.842)² × 2.043² / -0.630² = 165.1, rounded up to 166 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 <- -0.63                      # observed difference in means
sigma <- 2.04286396363366                     # 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(1571362)
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 = 166 (80% power), 221 (90% power)
 75% of observed effect: n per arm = 294 (80% power), 393 (90% power)
 50% of observed effect: n per arm = 661 (80% power), 884 (90% power)
power.t.test: 166 
Simulated power at n = 166 per arm: 0.800 (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. Analyses Reported by the Sponsor

The sponsor’s primary analysis is ANCOVA and reports a difference of least-squares means of -0.62, a 95% confidence interval from -1.11 to -0.14, and p = 0.011. The simple unadjusted recalculation from posted summary statistics gives a mean difference of -0.63 and p = 0.01. The close values indicate that the summary-based calculation is numerically consistent with the direction and approximate magnitude of the sponsor result.

They are not interchangeable analyses. The recalculation does not reproduce ANCOVA or its unavailable model details and uses only summary statistics. The registry does not report enough information to recreate the sponsor’s adjusted model, including its full covariate specification. The sponsor’s own analysis therefore takes precedence for inference about the trial. The recalculation is best viewed as a transparent, limited check, rather than as a competing estimate.

For the disability outcome, the sponsor’s estimate is 0.18 with a 95% confidence interval from -0.94 to 1.29 and p = 0.755; the unadjusted recalculation is 0.17. For the two pain-response outcomes, the sponsor reports Cochran-Mantel-Haenszel p-values of 0.021 and 0.025, while the displayed unadjusted recalculations are based on the posted counts and proportions. The registry does not report sponsor effect estimates and confidence intervals for these binary comparisons. Accordingly, the unadjusted effect summaries do not replace the sponsor’s reported analysis.

15. Multiplicity, Missing Data and Analysis Populations

The record describes a broad outcome and analysis program, including 101 outcome measures and 159 analyses, with 100 secondary outcomes posted. Multiple testing raises the chance that at least one result may appear statistically notable even when the tested effects are absent. In general, prespecifying a primary endpoint and controlling the testing family can help maintain the intended error rate. The registry record does not report a multiplicity adjustment, a testing hierarchy for secondary outcomes, or how decisions about confirmatory testing were made.

The primary endpoint is identified, and its sponsor-reported analysis is ANCOVA. That identification helps distinguish the main question from the many secondary questions, but it does not establish how analyses were selected or whether secondary testing was adjusted. The secondary p-values should therefore be interpreted with caution. The available record does not establish which outcomes or analyses were prespecified in the protocol or statistical analysis plan.

Missing-data handling is only partly described. The primary outcome population is defined as randomized participants who received at least one dose after randomization, and the registry says participant averages from 100 imputed datasets were used. It calls the approach hybrid multiple and single imputation, but the description ends before the procedure is fully specified. The imputation model, assumptions about why data were missing, amount of missingness, sensitivity analyses, and handling of incomplete diaries are not reported in the registry record.

These details matter because imputation relies on assumptions about unobserved outcomes. In general, different plausible assumptions can affect estimates and uncertainty, especially when missingness differs by treatment sequence or relates to outcomes. There is not enough information in the registry record to assess the direction or size of any such effect in this trial.

The record’s operational definition of the primary analysis population also deserves attention: randomized participants needed at least one post-randomization dose. The number and outcomes of randomized participants who did not meet that condition are not reported in the registry record. The protocol, full statistical analysis plan, original sample-size justification, interim analyses, and detailed handling of protocol deviations are likewise not reported in the registry record.

16. 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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: intermittent claudication (0.7%), peripheral arterial occlusive disease (0.7%), road traffic accident (0.7%), arthritis (0.7%), costochondritis (0.7%).

The registry lists serious adverse events and other adverse events for the open ALO-02 group and the two randomized sequences. In the open ALO-02 group, the record lists 3 participants affected by serious adverse events among 410 at risk and 186 affected by other adverse events among 410 at risk. For ALO-02 to placebo, it lists 2 affected by serious adverse events among 134 at risk and 26 affected by other adverse events among 134 at risk. For ALO-02 to ALO-02, it lists 5 affected by serious adverse events among 146 at risk and 41 affected by other adverse events among 146 at risk.

These are registry counts, not a complete account of event timing, duration, severity, relatedness, or differences in exposure time. The record lists intermittent claudication, peripheral arterial occlusive disease, road traffic accident, arthritis, costochondritis, and atrial flutter among the serious-event terms, each reported as 0.7%. The registry does not report death counts in the posted safety fields. These summaries should be read as descriptive safety information; they do not establish that a treatment caused or prevented an event, and the groups should not be compared by calculating percentages from these counts alone.

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

The primary sponsor analysis reports a difference in average change in eDiary pain scores between the two randomized treatment sequences. The estimate is -0.62, with a 95% confidence interval from -1.11 to -0.14 and p = 0.011. The posted group means and scale direction indicate that the continued-ALO-02 sequence had a smaller average change than the ALO-02 to placebo sequence over the specified final treatment weeks. This is a group-level result for the endpoint and analysis population described in the registry.

The estimate does not mean every participant had the same response. A mean summarizes the group and does not describe the spread of individual outcomes or identify who might experience a particular result. Nor does statistical evidence against a zero difference establish that the magnitude is important to patients. The registry record does not report a clinical-importance threshold for the primary change score.

The result is specific to a randomized withdrawal comparison after ALO-02 exposure, not necessarily to initiating ALO-02 versus never receiving it. It concerns average pain scores during the final two weeks of a 12-week study, not longer-term outcomes. It also does not by itself answer questions about function, quality of life, or every safety outcome; those require their own results and interpretation.

The posted secondary outcomes offer some context, but the record does not report a multiplicity strategy, and several secondary results are not available in the summarized results. Missing-data methods are only partly described. The interpretation should therefore preserve the distinction between the reported primary finding and the uncertainty that remains about model details, missingness, other outcomes, clinical importance, and applicability beyond the recorded study population and treatment sequence.

18. Why This Trial Matters Statistically

This trial illustrates why endpoint definitions and design context matter in clinical biostatistics. The primary measure is not a single pain rating: it averages diary reports over the last 7 days of each relevant week and then compares change from randomization baseline over the final two weeks. That construction determines what the estimate means and how it can be interpreted.

The randomized withdrawal design also defines the treatment contrast. The trial compares continued ALO-02 with withdrawal to placebo among sequences that both begin with ALO-02, rather than comparing initial treatment with no prior treatment. Correctly describing that contrast prevents a result from being generalized beyond the question the design addresses.

The primary analysis provides a model-based estimate and interval, while a straightforward recalculation from group summaries yields a similar estimate. Seeing that agreement is useful, but it also demonstrates why a summary-data check does not replace the sponsor’s adjusted analysis. The model’s full details and missing-data procedures are needed to reproduce and assess the result thoroughly.

Finally, the record contains many outcomes and analyses. The primary result can be distinguished from secondary findings, but absent information about multiplicity and prespecification limits how strongly individual secondary p-values can be interpreted. Together, these features make the study a useful example of reading effect estimates in light of trial design, endpoint construction, analysis populations, uncertainty, and the broader testing context.

19. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

Several trial-specific limitations affect interpretation. The withdrawal design compares continued ALO-02 with switching to placebo after ALO-02 exposure, so it does not directly answer every question about starting treatment or using it in other clinical settings. The registry record does not report the detailed procedures or criteria leading into randomization, limiting assessment of how participants entered the withdrawal comparison.

The primary analysis population is described as randomized participants who received at least one dose after randomization. The record does not report the outcomes of randomized participants who did not receive a post-randomization dose or provide enough detail to assess differences between the operational analysis population and all randomized participants. The primary outcome also uses imputation, but the description of the hybrid multiple and single imputation approach is incomplete. Missingness extent, assumptions, and sensitivity analyses are not reported in the registry record.

The ANCOVA model is identified, but its complete specification is not reported. The unadjusted recalculations therefore cannot reproduce the sponsor’s model. The primary confidence interval is available, but sponsor effect estimates and intervals are not posted for the two binary pain-response outcomes described in detail. A clinical-importance threshold for the primary endpoint is also not reported.

The large number of recorded outcomes and analyses creates a multiplicity concern. The registry does not report the testing hierarchy or a multiplicity adjustment. Results for some secondary outcomes are not available in the record, so the evidence cannot be characterized across every listed domain. Finally, the registry safety summaries do not provide comprehensive details on event timing, exposure duration, relatedness, or deaths. These reporting limits constrain the conclusions that can be drawn from the posted information.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Research Study of an Investigational Drug ALO-02 (Oxycodone Hydrochloride and Naltrexone Hydrochloride) in Patients With Moderate to Severe Chronic Low Back Pain — NCT01571362.
• Registry record last updated 2017-04-04; status last verified 2017-02.
• Results first posted 2017-04-04.
• 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.

23. Record Summary

NCT01571362 is a completed phase 3, multicenter, randomized, double-blind, parallel-group, placebo-controlled withdrawal study of ALO-02 in moderate to severe chronic low back pain. The official title describes a 12-week study. The registry lists enrollment as 410, a start of 2012-06, and primary completion of 2013-06.

The primary outcome is change in weekly average eDiary NRS-Pain score from randomization baseline to Weeks 11 and 12. The sponsor’s ANCOVA reports a difference of least-squares means of -0.62, a 95% confidence interval from -1.11 to -0.14, and p = 0.011, comparing ALO-02 to ALO-02 with ALO-02 to placebo. The unadjusted recalculation is similar but does not replace the sponsor’s analysis.

Secondary outcomes span disability, pain response, global assessment, and pain interference. Their interpretation is limited by unreported multiplicity procedures and incomplete results. The primary analysis uses imputation, but its details are not fully reported. The record therefore supports a focused reading of the primary comparison while leaving important questions about analysis assumptions, clinical importance, and broader applicability unresolved.