Phase 2 Chronic Low Back Pain Study: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Phase 2 Chronic Low Back Pain Study (NCT01364922), a randomized phase 2 study. The primary outcome is Change From Double-blind Baseline in Chronic Lower Back Pain (CLBP) Intensity by Visual Analog Scale (VAS).

2. Clinical Question

3. Trial Design

This was a Phase 2, randomized, double-blind, parallel-group study of hydrocodone/acetaminophen extended release compared with placebo for chronic low back pain. The registry record describes the study as a randomized withdrawal study and lists an open-label active-treatment arm alongside double-blind active-treatment and placebo arms. The posted descriptions say that the open-label arm received 2 tablets twice daily, while the double-blind active and placebo arms each received 1 tablet twice daily. Further details about the withdrawal process, eligibility for double-blind randomization, or treatment sequence are not reported in the registry record.

The study enrolled 168 participants and is marked completed. Its recorded start was 2011-06, and primary completion was 2011-10. Those dates describe the registry record and do not establish the duration of treatment for each participant. The primary endpoint timeframe is double-blind baseline to Day 29.

The primary outcome analysis included participants randomized to the double-blind period who received at least 1 dose of study drug in that period and had at least 1 assessment there. This analysis population was 95 participants in the active-treatment arm and 46 in the placebo arm. The registry lists 99 participants at risk for adverse events in the double-blind active arm and 47 in the double-blind placebo arm. These are distinct posted counts for different outcome reporting and should not be treated as interchangeable.

The sponsor is listed as AbbVie, with Abbott noted as the prior sponsor. The registry identifies two secondary outcomes: the participant’s global assessment of back pain status at final evaluation and the participant’s global assessment of study drug at final evaluation. Numerical results for those outcomes are not reported in the registry record. The original sample-size justification, detailed randomization procedures, and protocol-specific analysis plan are also not reported in the registry 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.

Open-label (OL) Period

Reasons for not completing this period:

Double-Blind (DB) Period

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The analysis of the primary outcome measure included all randomized participants who received at least 1 dose of study drug during the double-blind period (double-blind intent-to-treat) and had at least 1 assessment during the double-blind period.

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 Double-blind Baseline in Chronic Lower Back Pain (CLBP) Intensity by Visual Analog Scale (VAS), measured over Double-blind baseline to Day 29.

The primary endpoint is change from double-blind baseline in chronic lower back pain intensity, measured with a visual analog scale. The scale ranges from 0 mm for no pain to 100 mm for the worst pain imaginable. The baseline is defined as the last assessment before the first dose in the double-blind period; the final assessment is evaluated through Day 29.

Change scores focus the comparison on how participants’ measured pain intensity shifted from their own double-blind baseline, rather than only comparing their final scores. This can be useful when starting scores vary between participants. The registry reports least-squares means and standard errors from an ANCOVA model, so the arm estimates are model-based means rather than simple raw averages. The model’s covariates and other details are not reported in the registry record.

The posted least-squares mean change was 10.9 in the double-blind hydrocodone/acetaminophen extended-release arm and 19.9 in the placebo arm. The reported mean difference, active treatment minus placebo, was -9.0 points. These are changes on the stated pain-intensity scale. A negative contrast indicates that the estimated change was lower in the active-treatment arm than in the placebo arm; it does not, by itself, establish the precision or statistical significance of the result.

The endpoint is a continuous measure, so the comparison concerns average change rather than the number of participants reaching a particular threshold. A threshold-based responder outcome is not reported in the registry record.

7. Results Posted in the Registry

• Double-blind Hydrocodone/Acetaminophen Extended Release: mean 10.90 versus 19.90 on control; difference -9.00 (95% CI -18.27 to 0.27), standardized effect d = -0.33; z-test p = 0.057.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is continuous: it records a numerical change in pain-intensity score. For a continuous outcome, an arm’s mean summarizes the central level of measured change, while a contrast between arm means estimates how far apart the groups are on that measure. Here the registry reports least-squares means, an estimated mean difference, and standard errors. The posted values are model-based estimates rather than a report of individual participant measurements.

What does a least-squares mean represent?

A least-squares mean is an adjusted mean estimated from a statistical model. In general, adjustment allows the model to account for variables included in its specification. The registry identifies the model as ANCOVA, but does not report the covariates or further model details. It is therefore possible to describe the estimates as model-based, but not to say which factors were adjusted for or how they affected the result.

The posted least-squares mean change was 10.9 in the double-blind hydrocodone/acetaminophen extended-release arm and 19.9 in the placebo arm. The reported contrast is active treatment minus placebo, with an estimate of -9.0. The sign identifies the direction of the contrast under that ordering: the estimated change was lower in the active-treatment arm. The sign alone does not indicate whether the estimate is precise or whether it meets a clinical importance criterion.

Why report standard errors?

A standard error describes the uncertainty of an estimated quantity due to sampling variability under the analysis model. It is not the same as the spread of individual participants’ outcomes. The registry reports standard errors of 2.86 for the active-treatment least-squares mean and 3.77 for the placebo least-squares mean. Those values help describe uncertainty in the arm estimates. They should not be interpreted as the range of participant scores or as standard deviations.

Standard errors are also used in statistical inference, including construction of confidence intervals and tests of hypotheses. However, the comparison’s confidence interval and unadjusted p-value are reported directly, and those are the most direct summaries for interpreting the between-arm estimate. Reconstructing the trial’s model from arm summaries would not reproduce all features of its analysis, because model covariates and other details are not reported in the registry record.

What does ANCOVA do in general?

In general, analysis of covariance models an outcome while accounting for one or more covariates. For a baseline-and-follow-up setting, an ANCOVA approach can use baseline information in estimating the comparison between groups. The exact implementation varies with the specified model. The registry says that least-squares means and standard errors came from an ANCOVA model; it does not state the covariates, assumptions checked, or whether the reported model included baseline score as a covariate. Those specifics should not be inferred from the model name alone.

Model-based estimates can be useful because the analysis can account for prespecified features of the data, but interpretation depends on a suitable model and the data being analyzed. General considerations include whether the model form fits the outcome, whether observations satisfy the model’s assumptions, and how missing observations are handled. Results of diagnostics and sensitivity analyses are not reported in the registry record.

How should the p-value be read?

The unadjusted p-value is 0.057. In general, a p-value measures how unusual data at least as discrepant as those observed would be under a specified null hypothesis and model. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of an effect. The term unadjusted indicates that the posted value is not described as adjusted for multiple comparisons. The registry does not report a multiplicity procedure or an explanation of how the primary analysis was designated in relation to other analyses.

A p-value should be interpreted alongside the estimate and confidence interval. A value near a commonly used cutoff does not turn an uncertain estimate into a certain one, and a cutoff does not distinguish a meaningful effect from a trivial one. Here the confidence interval includes zero, while the point estimate is -9.0. Both the direction and the uncertainty matter.

Why can over-dispersion matter in general?

Over-dispersion is a general concern when observed variability is greater than the model’s assumed variability, particularly for outcomes modeled as counts or proportions. If a model does not account for excess variation, standard errors may be too small and uncertainty may be understated. The primary endpoint in this trial is a continuous change score analyzed using ANCOVA, and the registry does not report an over-dispersion assessment. This concept is a general statistical consideration, not a reported finding about this analysis.

Finally, the primary analysis population is defined by randomization, receipt of at least 1 double-blind dose, and at least 1 double-blind assessment. In general, analysis populations determine which participants contribute to an estimate and can affect how results are interpreted. The registry’s definition should be kept distinct from the overall enrollment and from the adverse-event at-risk counts. Details about missing-data methods, model diagnostics, covariate selection, and sensitivity analyses are not reported in the registry record.

10. Confidence Intervals and Precision

The estimated active-treatment-minus-placebo difference in change was -9.0 points. Its 95% confidence interval ranges from -18.27 to 0.275. A confidence interval is an uncertainty summary: in general, it describes a range of effect estimates compatible with the observed data and the analysis assumptions. It is not a guarantee that the true effect lies in that particular interval, nor does it show the distribution of individual participants’ scores.

Because the interval includes zero, the posted results do not rule out a zero difference under this interval-based assessment. The interval also extends to estimates in both directions around zero, so the data are compatible with a range of possible group contrasts. The point estimate is negative, but it should not be reported without the interval that conveys its uncertainty.

The interval is relatively broad on the scale of the point estimate, which indicates limited precision in the estimated contrast. The registry does not report a threshold for clinical importance, so the interval cannot be judged against a registry-defined meaningful difference. It also does not report the model’s covariates, diagnostic checks, or sensitivity analyses; those omissions limit assessment of how robust the interval is to analytic choices.

The standard errors for the arm-level least-squares means were 2.86 for double-blind hydrocodone/acetaminophen extended release and 3.77 for placebo. These quantify uncertainty in the respective estimated means, not the dispersion of individual outcomes. The confidence interval for the difference is the more direct measure for assessing precision of the between-arm comparison.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm, not total event counts. They illustrate sample sizes for a two-arm comparison under the stated assumptions: a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. The effect to detect is set as a stated fraction of the observed effect. The planning values are 143 participants per arm for 80% power and 192 per arm for 90% power when planning for 100% of the observed effect; 254 and 340 per arm when planning for 75% of the observed effect; and 572 and 765 per arm when planning for 50% of the observed effect.

In plain terms, the 80% power column gives the illustrated number of participants per arm for an analysis to have an 80% chance of detecting the assumed effect under those planning conditions. The 90% column gives the corresponding number for a 90% chance. These are planning probabilities under the assumptions, not guarantees about a future study’s result. The illustrations use the observed effect as their starting point; choosing a smaller fraction means planning to detect a smaller effect, which requires more participants per arm in these calculations.

These figures are planning illustrations, not observed power for this completed trial. They were calculated from the observed effect and are not the trial’s original design justification. The original design assumptions and sample-size justification are not reported in the registry record. The listed assumptions describe the illustration only and should not be attributed to the trial’s protocol.

12. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -9.000; pooled standard deviation σ = 27.151.
2. n = 2 × (1.96 + 0.842)² × 27.151² / -9.000² = 142.9, rounded up to 143 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 <- -9                      # observed difference in means
sigma <- 27.1505966652775                     # 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(1364922)
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 = 143 (80% power), 192 (90% power)
 75% of observed effect: n per arm = 254 (80% power), 340 (90% power)
 50% of observed effect: n per arm = 572 (80% power), 765 (90% power)
power.t.test: 143.8 
Simulated power at n = 143 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.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome and two secondary outcomes: participant’s global assessment of back pain status at final evaluation and participant’s global assessment of study drug at final evaluation. It reports 3 outcome measures, but the number of analyses listed is 0. The numerical secondary-outcome results are not reported in the registry record.

Multiplicity arises when several outcomes, comparisons, or analyses are examined. In general, conducting multiple tests can increase the chance of at least one result appearing statistically unusual by chance. Researchers may address this through a prespecified testing sequence or another adjustment strategy. Whether such a strategy was planned or used in this study is not reported in the registry record. The primary comparison’s p-value is explicitly unadjusted, so it should not be described as corrected for multiple testing.

The registry gives a primary analysis population definition: randomized participants who received at least 1 dose during the double-blind period and had at least 1 assessment during that period. This is described as the double-blind intent-to-treat population. The definition is important because it specifies who contributes to the primary endpoint analysis; it is not identical to the overall enrollment count. How many randomized participants did not meet the stated assessment requirement is not reported in the registry record.

The record does not describe how missing outcome measurements after an initial assessment were handled, whether any imputation was used, or whether sensitivity analyses examined alternative missing-data assumptions. The amount and pattern of missing primary-outcome data are not reported in the registry record. Without those details, the impact of missing observations on the estimated difference cannot be assessed from the registry results alone.

The registry identifies ANCOVA as the model used for least-squares means and standard errors but does not report the covariates, model specification, diagnostics, or a detailed analysis plan. It also does not report whether interim analyses were planned or conducted. These omissions do not establish that such procedures were absent; they mean the relevant details are not reported in the registry record.

14. 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: chest pain (2.1%).

The registry reports serious adverse events by listed group. In the open-label hydrocodone/acetaminophen extended-release group, 0 of 168 participants at risk were affected by a serious adverse event. In the double-blind active-treatment group, the figure was 0 of 99; in the double-blind placebo group, it was 1 of 47. The registry lists CHEST PAIN at 2.1% among the top serious-event terms, but does not state which group that term belongs to in the posted summary.

These are counts of participants affected, not counts of all events, and they do not describe the full safety profile. The registry also lists other, non-serious adverse-event counts of 70 of 168 in the open-label group, 7 of 99 in the double-blind active-treatment group, and 8 of 47 in the double-blind placebo group. The categories and reporting windows should be interpreted as recorded, not combined into a single comparison.

Death counts are not reported in the registry record. The differing at-risk group sizes and the open-label versus double-blind periods matter when reading these figures. They are descriptive counts and do not, on their own, establish comparative safety or causation.

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

The posted primary-outcome estimate describes a lower mean change in pain-intensity score in the double-blind hydrocodone/acetaminophen extended-release group than in the placebo group. The estimated difference was -9.0 points. Its confidence interval, from -18.27 to 0.275, includes zero, and the unadjusted p-value was 0.057. Together, these figures show an estimated difference with uncertainty that includes values close to no difference.

The result does not establish that the groups were identical. A confidence interval that includes zero is not proof of no effect; it indicates that the data and analysis do not exclude zero at the interval’s stated level. Conversely, a negative point estimate is not proof of a definitive difference. Interpretation needs to account for the range of plausible estimates and the study’s analysis population and reporting limitations.

The registry does not report a clinically important difference threshold, so statistical estimates cannot be classified as clinically important or unimportant. The result also concerns a specific measure, population definition, comparison, and timeframe: change in the visual analog scale from double-blind baseline to the final assessment through Day 29. It should not be generalized from this record to other outcomes or time periods.

The primary analysis population excludes randomized participants who did not receive a double-blind dose or did not have at least 1 double-blind assessment, according to the registry’s definition. Further information about missing data and analysis choices is not reported. The listed secondary outcomes have no numerical results in the registry record, and safety counts are descriptive rather than a complete assessment of comparative risk.

These results are statistical evidence from a Phase 2 study, not medical advice. They do not by themselves determine treatment choices, establish a complete benefit-risk profile, or answer questions beyond the measured endpoint and reported follow-up.

16. Why This Trial Matters Statistically

This trial illustrates why an estimated effect should be read together with its precision. The point estimate is -9.0 points, but the 95% confidence interval runs from -18.27 to 0.275 and includes zero. The unadjusted p-value is 0.057. Looking only at the point estimate would leave out uncertainty; looking only at the p-value would omit the estimated size and direction of the contrast.

It also demonstrates the importance of distinguishing model-based estimates from raw group summaries. The primary result uses least-squares means and standard errors from ANCOVA. Such estimates depend on the model specification, while the registry does not report its covariates or diagnostics. Transparent reporting of those details supports assessment of how the analysis was conducted and how its assumptions relate to the result.

Analysis population definitions matter as well. Overall enrollment was 168, whereas the primary outcome analysis included 95 active-treatment and 46 placebo participants under the stated double-blind intent-to-treat definition. The adverse-event at-risk counts differ from those primary-outcome counts. Keeping these populations distinct prevents the mistaken impression that every reported outcome uses the same denominator.

Finally, the registry includes secondary outcomes and adverse-event summaries but does not report numerical results for the secondary outcomes or details of any multiplicity procedure. It also does not report the original sample-size justification or missing-data methods. These gaps limit what can be concluded from the registry record and reinforce a central statistical lesson: estimates, uncertainty, analysis populations, and reporting completeness all shape interpretation.

17. Limitations

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

The registry record leaves important details about design and analysis unspecified. Although the study is identified as a randomized withdrawal study and lists open-label and double-blind groups, the sequence by which participants entered those groups and the withdrawal procedures are not reported in the registry record. The original sample-size justification and the assumptions used in the trial’s own planning are also not reported.

The primary analysis population requires randomization, at least 1 dose during the double-blind period, and at least 1 assessment during that period. The record does not report how many participants failed to meet these conditions or describe the handling of subsequent missing measurements. Missing-data methods and sensitivity analyses are not reported in the registry record.

The primary outcome is based on change in a visual analog scale, and the reported estimates come from ANCOVA. The registry does not state the model covariates, full specification, diagnostic results, or whether additional analyses assessed robustness. It does not provide a threshold for a clinically important difference. These omissions restrict assessment of clinical interpretation and model dependence.

The p-value is unadjusted, while two secondary outcomes are listed without numerical results. The multiplicity procedure, if any, is not reported in the registry record. The adverse-event summary gives group counts, but death counts are not reported, and the listed top serious-event term is not attributed to a specific group in the posted summary.

The outcome timeframe is double-blind baseline to Day 29. The registry results therefore do not describe longer-term outcomes in this record. These are limitations of what the registry record reports; they should not be taken as proof that unreported procedures or information did not exist.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Phase 2 Chronic Low Back Pain Study — NCT01364922.
• Registry record last updated 2014-01-29; status last verified 2013-12.
• Results first posted 2013-12-23.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

21. Record Summary

NCT01364922 is a completed Phase 2 randomized, double-blind, parallel-group study of hydrocodone/acetaminophen extended release compared with placebo for chronic low back pain. The record lists enrollment of 168, a start date of 2011-06, and primary completion in 2011-10. It also describes an open-label active-treatment arm and identifies the study as a randomized withdrawal study.

The primary outcome was change in chronic lower back pain intensity from double-blind baseline to the final assessment through Day 29, measured using a visual analog scale from 0 mm to 100 mm. The estimated active-treatment-minus-placebo difference was -9.0 points, with a 95% confidence interval from -18.27 to 0.275 and an unadjusted p-value of 0.057.

The registry lists two secondary outcomes but does not report their numerical results. It reports serious adverse events in 0 of 168 open-label participants, 0 of 99 double-blind active-treatment participants, and 1 of 47 double-blind placebo participants. Important details about the original sample-size justification, missing-data handling, and analysis plan are not reported in the registry record.