Safety and Efficacy of Buprenorphine Transdermal System in Subjects With Moderate to Severe Osteoarthritis of Hip or Knee: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety and Efficacy of Buprenorphine Transdermal System in Subjects With Moderate to Severe Osteoarthritis of Hip or Knee (NCT00313846), a randomized phase 3 study. The primary outcome is The Time (Days) From First Administration of Double-blind Treatment to the Development of Inadequate Analgesia at the Primary Osteoarthritis Pain Site..

2. Clinical Question

3. Trial Design

This was a randomized, double-blind, placebo-controlled, parallel-group, multicenter phase 3 study. The registry lists 529 enrolled participants and a completed trial. The study began in 2003-04, and primary completion was recorded in 2004-06. The double-blind treatment intervention was planned for 4 weeks. Participants received either a buprenorphine transdermal patch or a matching placebo patch, with acetaminophen available to all participants as supplemental analgesia.

The registry describes BTDS strengths of 5, 10 or 20 micrograms per hour and matching placebo patches. It does not report the allocation ratio or how participants were assigned among the listed patch strengths. The primary outcome analysis population was the Full Analysis Population, defined as randomized participants who received at least one dose of double-blind study drug and had at least one primary efficacy observation during the double-blind phase. That population is reported as N = 326; the outcome results list 162 participants in the placebo group and 164 in the BTDS group.

The registry also reports an open-label run-in period. Its adverse-event information is listed separately from the double-blind comparison, so those figures should not be treated as another randomized comparison. The randomization sequence, concealment procedures, detailed visit schedule, and trial-specific sample-size rationale are not reported in the registry record. The record identifies Purdue Pharma LP as the sponsor. These design details establish the comparison being described, but do not by themselves explain the choices made in the original statistical plan.

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: 529 subjects began the run-in period with buprenorphine transdermal system (BTDS) 5 and their dose was titrated to BTDS 10 or 20 to achieve effective pain control. Subjects meeting criteria for adequate analgesia within 21 days were randomized into the double-blind phase (the number of subjects \[N\] = 328 completed run-in period).

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The Full Analysis Population (N = 326) consisted of subjects who were randomized and received at least 1 dose of double-blind study drug and had at least 1 primary efficacy observation during the double-blind phase.

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 The Time (Days) From First Administration of Double-blind Treatment to the Development of Inadequate Analgesia at the Primary Osteoarthritis Pain Site., measured over Double-blind phase ( 28 days): reaching "inadequate analgesia" on any 2 days of the 7-day dosing periods.

The primary endpoint was the time, in days, from the first administration of double-blind treatment until inadequate analgesia developed at the primary osteoarthritis pain site. Inadequate analgesia could be reached by meeting any of the registry’s stated criteria: an average pain score over the last 24 hours of at least 5 on the 0 to 10 scale on any 2 days in a 7-day dosing period; use of more than 1000 mg per day of acetaminophen for pain at that site for at least 2 days in a 7-day period; or use of nonstudy opioid analgesic medication for pain at that site.

This endpoint combines reported pain intensity with escalation or substitution of pain medication. It is intended to capture the time until the participant no longer met the study’s definition of adequate analgesia, rather than pain intensity alone. A longer time means that the threshold was reached later; it does not mean that every participant had the same pain experience or reached the endpoint through the same criterion. The registry states that the primary phase lasted 28 days and describes the analysis as the mean time to inadequate analgesia. The registry’s reported analysis used a Kaplan-Meier estimate of the mean, while the posted arm summaries give means and standard errors.

7. Results Posted in the Registry

• Double-blind BTDS: mean 17.20 versus 12.30 on control; difference 4.90 (95% CI 2.59 to 7.21), standardized effect d = 0.46; z-test p = <0.001.

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.

Daily Maximum "Pain Right Now" Score for the Primary Osteoarthritis (OA) Pain Site

• 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:
  • Double-blind BTDS: mean 3.20 versus 3.80 on control; difference -0.60 (95% CI -1.00 to -0.20), standardized effect d = -0.32; z-test p = 0.003.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Double-blind BTDS than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry lists one secondary outcome: daily maximum Pain Right Now score at the primary osteoarthritis pain site. Its posted group summaries and recalculated comparison give a lower mean score in the BTDS group than in the placebo group during the specified assessment period. This direction is broadly consistent with the primary outcome’s longer mean time to the defined inadequate-analgesia endpoint in the BTDS group, but the outcomes measure different aspects of the treatment period and should not be treated as interchangeable evidence.

The primary endpoint is a time until a composite definition is met, while the secondary endpoint is a pain score over a specified 7-day period. The latter may be based on the last dosing period or the last period before inadequate analgesia or discontinuation, according to the registry description. That assessment rule means the secondary summary is not simply a score for every participant at one common fixed point, and the record does not give further details about how missing scores were handled.

The secondary comparison’s unadjusted p-value is 0.003. An unadjusted p-value does not account for the possibility of testing multiple outcomes or analyses. Because the record does not state whether a multiplicity procedure was planned or applied, the secondary result carries less confirmatory weight than a clearly prespecified and multiplicity-controlled result would. It is best read as supportive context for the reported primary pattern, rather than as an independent confirmation of it. The registry does not list a sponsor-reported analysis for this secondary measure.

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

• Daily Maximum "Pain Right Now" Score for the Primary Osteoarthritis (OA) Pain Site

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is a time-to-event measure: the event is the first occasion on which a participant meets the trial’s definition of inadequate analgesia. The registry describes this as time from first double-blind treatment administration to endpoint development. Time-to-event methods are useful when the outcome is not simply whether an event occurred, but also when it occurred. They can account for participants whose follow-up ends before the event is observed, provided the analysis handles such observations appropriately.

What does the Kaplan-Meier method contribute?

In general, the Kaplan-Meier method estimates the event-time experience over follow-up while retaining information from participants who have incomplete event-time observation. A Kaplan-Meier estimate of the mean can summarize time to event over a defined observation period. The registry identifies the reported primary analysis as a Kaplan-Meier estimate mean and gives its p-value as .0026. It does not report the detailed handling of censoring, the test used to obtain that p-value, the estimated means by group, or a confidence interval for the sponsor’s analysis. Those details are therefore not reported in the registry record.

The posted recalculation instead uses the arm-level mean summaries and standard errors. It reports the placebo mean as 12.3 days with standard error 0.77 and the BTDS mean as 17.2 days with standard error 0.89. The recalculated unadjusted mean difference is 4.9 days, and its 95% confidence interval is 2.593 to 7.207 days. This is a simple comparison of posted summary statistics, not a reconstruction of the full participant-level time-to-event analysis. The recalculated p-value is <0.001.

Why distinguish a mean difference from the reported analysis?

Different methods can use different information and assumptions, even when they address the same endpoint. A comparison based on arm means and standard errors does not recreate a Kaplan-Meier analysis, which is designed for event-time data and may account for censoring. A difference between the recalculated and registry-reported p-values would not, by itself, identify an error or explain why the values differ. The registry does not provide enough analysis detail to determine the reason. The discrepancy flag is false, so the registry entries do not identify a discrepancy requiring the sponsor’s analysis to take precedence over the recalculation.

For the continuous secondary pain score, the posted summaries give a mean of 3.8 with standard error 0.15 for placebo and a mean of 3.2 with standard error 0.14 for BTDS. The recalculated unadjusted mean difference is -0.6 units, with a reported unadjusted p-value of 0.003. A mean difference expresses the average separation between the two groups on the stated scale, with the sign depending on the order of subtraction. A standard error describes uncertainty in an estimated mean; it is not the spread of individual participant scores.

General statistical explanation: standard errors quantify sampling uncertainty in an estimate under the analysis assumptions. They should not be interpreted as showing that individual observations cluster closely around the group mean. A standard deviation describes variation among observations, whereas a standard error describes precision of an estimated quantity. The registry gives standard errors for the arm means but does not provide all underlying participant-level distributions.

What is the role of a p-value?

In general, a p-value measures how unusual data at least as incompatible with a specified null hypothesis would be if that hypothesis and the analysis assumptions were true. It is not the probability that the null hypothesis is true, nor does it measure the magnitude or clinical importance of an effect. The unadjusted p-values reported with the recalculations do not account for possible multiplicity across outcomes or analyses. The registry does not report a multiplicity adjustment or a detailed analysis plan, so the extent to which the posted p-values were prespecified or adjusted cannot be determined.

General statistical explanation: time-to-event outcomes can be affected by censoring, meaning that the event time is not observed for every participant during the observation period. The registry does not report the number or timing of censored observations, nor does it state the assumptions used to interpret censoring. This limits independent assessment of the reported Kaplan-Meier analysis. The posted group summaries remain interpretable as summaries, but they cannot provide the missing participant-level information or substitute for the original analysis details.

11. Confidence Intervals and Precision

The recalculated primary mean difference is 4.9 days, with a 95% confidence interval from 2.593 to 7.207 days. A confidence interval describes a range of values compatible with the estimate under the method and assumptions used. Its span conveys uncertainty about the mean difference; it is not a range containing a specified proportion of individual participants’ times. The interval is for the unadjusted recalculation from posted summary statistics, not the Kaplan-Meier estimate mean analysis reported by the registry.

The registry’s primary analysis gives a p-value of .0026 but does not report a confidence interval or the underlying group-specific Kaplan-Meier mean estimates. As a result, its precision cannot be assessed from a reported interval for that analysis. The standard errors in the arm summaries, 0.77 days for placebo and 0.89 days for BTDS, describe the precision of those arm means, not a direct interval for the sponsor’s analysis.

For the secondary pain score, the recalculated mean difference is -0.6 units. The comparison’s reported confidence limits are -1.002 and -0.198; those are registry-posted recalculation values. The registry does not report a sponsor analysis for this secondary outcome. Intervals and p-values address statistical uncertainty under their methods, not whether a difference matters to participants. Neither outcome summary reveals the range of responses among individuals, and no threshold for a clinically important difference is reported in the registry record.

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 figures are illustrations based on the observed recalculated effect, not the trial’s original sample-size design. For these continuous outcomes, the figures are participants per arm, not total event counts. The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The target effect is a stated fraction of the observed effect. The trial’s own original design assumptions and sample-size justification are not reported in the registry record.

The 80% power column describes the number of participants per arm that the planning calculation associates with an 80% chance of detecting the specified effect under its assumptions, if that effect and the assumptions are correct. The 90% power column gives the corresponding planning quantity for a 90% chance. These are prospective planning concepts, not guarantees about a future study. The entries change when the assumed effect changes: planning for a smaller effect requires more participants per arm to reach the same target power.

The rows express planning for 100%, 75% or 50% of the observed effect. They should be read as scenarios rather than as evidence that any of those effects is the true underlying effect. The primary endpoint’s planning figures are based on the recalculated mean difference; the secondary outcome has its own effect and planning figures. The two outcomes therefore do not share one planning result.

These calculations are not observed power. Observed power is not a useful substitute for the estimated effect and its uncertainty after results are known; the planning figures instead show how sample size would vary under stated hypothetical effect assumptions. They do not establish what sample size the completed trial was designed to use, whether the trial met its original target, or whether the original study design had adequate power for the endpoints.

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 = 4.900; pooled standard deviation σ = 10.634.
2. n = 2 × (1.96 + 0.842)² × 10.634² / 4.900² = 73.9, rounded up to 74 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 <- 4.9                      # observed difference in means
sigma <- 10.6339827678144                     # 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(313846)
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 = 74 (80% power), 99 (90% power)
 75% of observed effect: n per arm = 132 (80% power), 176 (90% power)
 50% of observed effect: n per arm = 296 (80% power), 396 (90% power)
power.t.test: 74.9 
Simulated power at n = 74 per arm: 0.793 (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 registry reports a primary analysis using a Kaplan-Meier estimate mean and gives a comparison p-value of .0026. It does not provide that analysis’s group-specific estimates or confidence interval. The recalculated result uses posted arm-level means and standard errors: the mean difference is 4.9 days, its 95% confidence interval is 2.593 to 7.207 days, and its unadjusted p-value is <0.001.

These are not identical analysis outputs. The recalculation is a simple, unadjusted comparison from summary statistics; the registry’s reported method is identified as a Kaplan-Meier estimate mean. The registry does not provide enough detail to reproduce the reported analysis or explain the relationship between its p-value and the recalculated one. The discrepancy flag is false. Accordingly, the posted record does not indicate a discrepancy requiring the sponsor’s analysis to take precedence, but the recalculation should still be understood as a separate summary-statistic analysis rather than as a replacement for the reported method.

For the secondary pain score, the registry posts arm summaries and the recalculated unadjusted comparison, but no sponsor-reported analysis is listed. The registry does not report a multiplicity strategy or a complete statistical analysis plan for either result.

15. Multiplicity, Missing Data and Analysis Populations

The registry record reports one primary outcome and one secondary outcome, with an unadjusted recalculated comparison for each. It does not report whether the outcomes were arranged in a testing hierarchy, whether a multiplicity adjustment was planned, or how the family of analyses was defined. Without that information, the nominal p-values should not be read as if the record established control of the overall false-positive risk across outcomes and analyses.

The registry defines the Full Analysis Population for the primary outcome as randomized participants who received at least one dose of double-blind study drug and had at least one primary efficacy observation during that phase. It reports N = 326 for this population and lists 162 placebo participants and 164 BTDS participants in the outcome summaries. The registry does not report a separate safety population definition, an intention-to-treat analysis description, or how participants excluded from the Full Analysis Population were handled in the efficacy analyses.

The outcome descriptions refer to treatment-period observations, but detailed information about missing primary endpoint data, censoring, withdrawals, protocol deviations, or imputation is not reported in the registry record. The Kaplan-Meier method is identified for the reported primary analysis, but the record does not state the censoring rules, give numbers censored, or describe sensitivity analyses. For the secondary pain score, the registry gives the assessment window but does not describe the amount or pattern of missing data or its analytic treatment.

The trial’s original statistical analysis plan, prespecified handling of multiplicity, interim analyses, and sample-size justification are not reported in the registry record. These gaps do not invalidate the posted summaries, but they limit assessment of whether the analyses followed prespecified procedures and how robust the results are to alternative assumptions. The recalculated p-values and intervals are unadjusted summaries and should be interpreted with those limitations in mind.

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 4.50% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: transient ischemic attack (0.6%), numbness left side of face (0.6%), worsening of hypertension (0.6%), intermittent complete heart block (0.2%), pneumonia (0.2%).

During the double-blind period, the registry lists 0 serious adverse events among 162 participants at risk in the placebo patch group and 1 among 164 participants at risk in the BTDS group. For other adverse events, the corresponding listed counts are 15 and 28. These are event figures reported by treatment group; they do not by themselves establish whether an event was caused by study treatment.

The open-label run-in period is reported separately, with 6 serious adverse events and 229 other adverse events among 529 participants at risk. Those figures should not be combined with the double-blind groups to interpret the randomized comparison. The registry lists serious-event terms including transient ischemic attack, numbness left side of face, worsening of hypertension, intermittent complete heart block, pneumonia, and left foot cellulitis, with percentages as recorded in the adverse-event information. Death counts are not reported in the registry record.

Adverse-event counts are descriptive and do not establish comparative risk or causation. The registry does not provide further detail about event timing, severity beyond the serious-event classification, or relatedness assessments in the information summarized.

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

The results describe a randomized comparison of BTDS patches with matching placebo patches during a 4-week double-blind treatment period in participants with moderate to severe osteoarthritis pain of the hip or knee. The posted primary summaries show a longer mean time to the study-defined inadequate-analgesia endpoint in the BTDS group. The secondary summaries show a lower average daily maximum pain score in that group during the specified assessment window.

These results do not show that every participant experienced the same outcome, that all pain-related medication use followed the same pattern, or that a particular individual would have a similar response. The primary endpoint could be reached through different pain-score or medication-use criteria, and the registry does not report the distribution of those routes. The group means also do not describe the full spread of participant experiences.

The results do not establish clinical importance. The registry does not report a threshold defining a clinically meaningful difference for either endpoint. Statistical significance, including a small p-value, is not itself evidence that the size of a difference is important to participants. The confidence interval for the recalculated primary comparison describes uncertainty under that calculation’s assumptions, not certainty about an individual outcome or future response.

Interpretation is also bounded by the 4-week double-blind period, the selected analysis population, the outcome definitions, and the missing analytic details. The registry record does not provide long-term outcomes, a complete account of missing-data handling, or a detailed multiplicity plan. It therefore cannot answer questions about longer treatment, other populations, or outcomes not reported in the trial record. These are statistical descriptions of posted trial results, not individual medical guidance.

18. Why This Trial Matters Statistically

This trial illustrates why the definition of an endpoint matters. The primary outcome is not simply a pain score at a single visit: it measures time until a prespecified composite threshold for inadequate analgesia is met. Its interpretation depends on the event definition, follow-up, and analysis of incomplete event times. A secondary pain score offers a different view of the outcome and should be considered alongside, not substituted for, the primary endpoint.

It also illustrates the distinction between a registry-reported analysis and a recalculation from summary statistics. The primary registry result is identified as a Kaplan-Meier estimate mean analysis, while the recalculation compares posted means and standard errors. Stating the method and its limits helps readers avoid treating different p-values as if they came from the same calculation.

Finally, planning calculations based on observed effects are not the trial’s original design rationale and are not observed power. They can illustrate how planning needs depend on the assumed effect, while the absent original design assumptions leave the trial’s original sample-size rationale unclear. The registry’s omissions on multiplicity, missing data, and analysis details are also statistically important: they shape how much confidence readers can place in confirmatory interpretations without erasing the descriptive information that is reported.

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.

The registry does not report how randomization was generated or concealed, whether the patch strengths were assigned separately, or how participants were distributed among the 5, 10 and 20 microgram-per-hour strengths. The arm summaries pool BTDS strengths, so they do not establish a separate result for each strength. The record also does not give detailed adherence information or treatment exposure by strength.

The study’s double-blind treatment period was 4 weeks, limiting the time horizon of the posted results. The primary endpoint uses multiple routes to inadequate analgesia, but the registry does not report the number of events attributed to pain scores, acetaminophen use, or nonstudy opioid use. It also does not report detailed censoring, missing-data handling, or sensitivity analyses. These omissions make it difficult to assess how individual endpoint components and incomplete observations influenced the result.

The registry identifies a Kaplan-Meier estimate mean analysis but does not provide its group estimates or confidence interval. The recalculated mean difference is based on arm-level summary statistics and cannot recover participant-level event-time information. The secondary outcome has an assessment window that may depend on inadequate analgesia or discontinuation; further details about missing scores and analysis are not reported.

Multiplicity procedures, interim analyses, the original statistical analysis plan, and the trial’s original sample-size justification are not reported in the registry record. The available safety information is also limited: death counts are not reported, and the posted adverse-event counts alone do not describe timing, relatedness, or the full clinical context of events. These limitations constrain interpretation but do not change what the registry reports.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Safety and Efficacy of Buprenorphine Transdermal System in Subjects With Moderate to Severe Osteoarthritis of Hip or Knee — NCT00313846.
• Registry record last updated 2012-09-03; status last verified 2012-08.
• Results first posted 2010-09-28.
• 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

NCT00313846 is a completed randomized, double-blind, placebo-controlled, parallel-group phase 3 study sponsored by Purdue Pharma LP. It enrolled 529 participants and evaluated buprenorphine transdermal patches at 5, 10 or 20 micrograms per hour against matching placebo patches for moderate to severe osteoarthritis pain of the hip or knee. The double-blind treatment period lasted 4 weeks.

The primary endpoint was time to inadequate analgesia, analyzed in a Full Analysis Population of N = 326. The posted summaries report means of 12.3 days for placebo and 17.2 days for BTDS. The registry reports a Kaplan-Meier estimate mean analysis with p = .0026; the separate unadjusted recalculation gives a mean difference of 4.9 days and p < 0.001. One secondary pain-score outcome is listed. The registry does not report the original sample-size rationale, a detailed multiplicity plan, or several analysis details needed for fuller assessment.