A Trial of Corticosteroids for Low Back Pain: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Trial of Corticosteroids for Low Back Pain (NCT00290589), a randomized phase 3 study. The primary outcome is Numerical Rating Scale (0-10), an Interval Pain Scale, on Which 0 Indicates no Pain and 10 Indicates the Worst Pain Imaginable.

2. Clinical Question

3. Trial Design

This was a Phase 3, randomized, parallel-group trial listed as completed. The registry reports enrollment of 82. Participants were assigned to intramuscular methylprednisolone acetate or placebo, and the masking classification was quadruple. The record names Montefiore Medical Center as the sponsor. The listed start was 2003-06, and primary completion was 2007-08.

The posted primary outcome summaries include 39 participants in the methylprednisolone acetate group and 43 in the placebo group. These are the numbers attached to the primary outcome summaries; the registry does not explain how they relate to enrollment or whether there were exclusions, missing measurements, or other analysis-population rules. Such details are not reported in the registry record.

The parallel design compares the two assigned groups over the same trial period. Random assignment is intended, in general, to make groups comparable on average at the start of a study. Masking can reduce the influence of expectations on participant reports and other trial procedures, although the registry does not describe who was masked or how masking was maintained. The record does not report the sequence-generation method, allocation concealment procedures, treatment adherence, or the timing and method of any unmasking.

The study's primary outcome was a continuous pain-improvement measure at 1 month. The registry also lists one secondary outcome concerning analgesic use, with its assessment at 1 month. There are three outcome measures listed in the record, but only one secondary outcome is posted in the posted results record; how the remaining measures were defined or analyzed is not reported. The registry record lists no reported analyses. It therefore does not establish whether the posted summary comparisons were part of a prespecified analysis plan or whether any other analysis was conducted.

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.

Overall Study

Reasons for not completing this 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 Numerical Rating Scale (0-10), an Interval Pain Scale, on Which 0 Indicates no Pain and 10 Indicates the Worst Pain Imaginable, measured over 1 month.

The primary endpoint is described as improvement in pain on an 11-point Numerical Rating Scale, ranging from 0 for no pain to 10 for the worst pain imaginable. It is based on the difference between a rating at the emergency department visit and a rating collected at a telephone call 1 month later. The registry reports the endpoint as a mean with standard deviation, so the posted summaries describe the average reported improvement and the spread of individual values within each group.

A continuous pain scale can capture gradations in reported symptoms rather than reducing each person's experience to a yes-or-no category. In general, the mean gives a concise summary of the group, while the standard deviation indicates how dispersed the measurements are around that mean. Neither summary describes every participant's experience. The scale's anchors give the number a clinical context, but the registry record does not report a threshold for a clinically important change or a responder definition.

The endpoint is relevant to the stated question because the study aimed to examine pain change after the assigned injection. The registry reports group means of 7.1 for methylprednisolone acetate and 5.8 for placebo. Its improvement description implies that higher values represent greater reported improvement. A difference in group means is an average contrast, not a guarantee that each participant experienced a change of that size.

7. Results Posted in the Registry

• Intramuscular Methylprednisolone Acetate: mean 7.10 versus 5.80 on control; difference 1.30 (95% CI -0.11 to 2.71), standardized effect d = 0.40; z-test p = 0.070.

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.

Number of Patients Using Analgesics

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Intramuscular Methylprednisolone Acetate: 21.6% (8/37) versus 42.5% (17/40) on control; risk difference -20.9 percentage points (95% CI -41.1 to -0.6), risk ratio 0.51 (0.25–1.04); unadjusted two-proportion z-test p = 0.051.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Intramuscular Methylprednisolone Acetate and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists one posted secondary outcome: the number of participants using analgesics at 1 month. The direction of its unadjusted comparison is broadly aligned with the primary mean-improvement comparison, but alignment in direction does not make the two outcomes interchangeable. Pain ratings and analgesic use measure different aspects of the clinical experience, and the registry does not report details about medication use beyond the posted count outcome.

The secondary result should be viewed as supporting context rather than as a decisive confirmation. Its uncertainty and its status as a secondary outcome matter, especially because the primary comparison itself has an interval that includes no difference. The registry record does not report whether the secondary outcome was prespecified with a particular inferential role, whether an analysis hierarchy was used, or whether multiplicity was addressed. Without that information, a secondary p-value should not be interpreted as if it were an independent, confirmatory test.

In general, examining multiple outcomes increases the opportunity to observe a seemingly notable result by chance, particularly when each is tested separately without adjustment. This is why secondary outcomes are often interpreted cautiously and in the context of the full analysis plan. The record lists three outcome measures but posts only one secondary outcome, and does not explain how the other listed measures relate to the posted results. No broader pattern across unposted outcomes can therefore be assessed from the registry record.

The secondary measure may add a complementary view of the trial, but it does not resolve the precision limits of the primary estimate or establish a causal explanation for any difference in medication use. The posted information supports a restrained reading: the secondary direction is compatible with the primary pattern, while the lack of reported multiplicity and analysis details limits the weight it can carry.

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

• Number of Patients Using Analgesics

9. Statistical Interpretation

10. Statistical Methods Explained

What does a mean difference represent?

The primary outcome is continuous, and the registry reports a mean for each group. A mean difference subtracts one group's average from the other's according to a stated comparison direction. In this record, the reported contrast is methylprednisolone acetate minus placebo: 7.1 minus 5.8 is reported as a mean difference of 1.3 units. The positive estimate means the intervention group's reported mean improvement was higher. It is an average contrast, not a description of how much every participant changed.

The posted group means are accompanied by standard deviations of 3.0 and 3.5. In general, a standard deviation describes the spread of observations around a group mean. It is not the uncertainty of the mean itself and it is not a confidence interval. A broad spread relative to the average contrast means there can be considerable overlap in participant-level values even when group means differ.

What does an unadjusted comparison mean?

The mean-difference estimate and its confidence interval are simple unadjusted recalculations from the posted summary statistics. An unadjusted comparison uses the group summaries without a reported covariate adjustment. The registry lists zero reported analyses, so it does not present a sponsor-reported model or explain whether an adjusted analysis was planned or performed. The recalculation should therefore be understood as a way to describe the posted summaries, not as a substitute for an unreported trial analysis.

In general, adjustment can account for measured baseline characteristics or other specified features of a study, while an unadjusted comparison does not do so. Whether adjustment was planned, which variables might have been used, and how the trial would handle departures from its intended analysis are not reported in the registry record. The record also does not report whether the outcome comparison follows an intention-to-treat principle or another analysis-population definition.

How should the p-value be read?

The primary unadjusted p-value is 0.07. In general, a p-value describes how surprising data at least as inconsistent with a null model would be if the null model and the assumptions used for the calculation were correct. It is not the probability that the null hypothesis is true, and it does not state the chance that the observed estimate is due to chance. It also does not measure the size or clinical importance of the difference.

The value is above 0.05, a commonly used reference threshold, but that threshold is not a boundary separating meaningful from meaningless evidence. The estimate, its interval, the outcome's clinical context, the analysis plan, and the quality of the measurements all matter. The registry record does not report the trial's original significance threshold or provide a sponsor analysis to compare with the recalculation.

Why does the confidence interval matter?

A confidence interval expresses uncertainty around an estimated contrast. In general, an interval constructed using a stated method gives a range of values compatible with the observed data under that method and its assumptions. The primary interval extends from just below zero to 2.707. Zero corresponds to no mean difference. Because zero falls within the interval, the results do not rule out no average difference at the interval's stated confidence level. The registry does not report the confidence level or interval method in the record, so the interval should be read as the posted comparison's uncertainty range without assuming further details.

The interval also reaches values representing a positive mean difference, so the result is not precise enough to distinguish a small or absent contrast from a larger positive one. In general, more information can narrow an interval, but the record does not establish why the interval has its particular width or whether the original trial was designed around a specific target precision.

What assumptions underlie the planning illustrations?

The planning illustrations are based on the observed effect and use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. They show planning for the stated fractions of the observed effect, rather than estimates of the trial's original design. Smaller effects require more participants to detect under otherwise similar assumptions because they are harder to distinguish from random variation.

In general, a normal approximation uses a simplified mathematical model for the sampling distribution of an estimate. Such planning calculations are illustrations, not guarantees that a future study will achieve its target. The registry record does not report the original sample-size justification, the trial's own design assumptions, or whether the planning assumptions used in these illustrations match the assumptions used by the trial team.

Why are the analysis details important?

For a continuous outcome, choices about the analysis population, missing measurements, distributional assumptions, and covariate adjustment can affect an estimate or its uncertainty. Randomization supports a comparison of assigned groups, but analysis decisions still matter. The registry does not report how missing primary outcome values were handled, whether any imputation was used, or whether participants were analyzed according to their assigned group.

These omissions do not change the posted group summaries, but they limit what can be inferred about the full analysis process. The outcome summaries and unadjusted recalculation support a descriptive comparison of the available posted data. They do not supply a complete account of prespecified methods or the sponsor's inferential analysis.

11. Confidence Intervals and Precision

The primary mean-difference estimate is 1.3 units, with a confidence interval whose lower limit is slightly below zero and whose upper limit is 2.707. This interval is important because it shows that the estimated contrast is not precise enough to exclude no mean difference. It also includes positive differences, so the data do not pin down a single plausible size of contrast. The posted estimate should not be read as though 1.3 were known without uncertainty.

In general, a confidence interval communicates precision: a narrower interval indicates less statistical uncertainty about the estimate under the method used, while a wider interval indicates more. The interval's inclusion of zero is consistent with the unadjusted p-value of 0.07. These are related summaries of the same comparison, not separate pieces of evidence. The registry record does not report the confidence level or the method used for the posted interval, which limits further assessment of its construction.

The group standard deviations, 3.0 and 3.5, also provide context for the mean contrast by showing that individual measurements vary within each arm. They do not replace the interval for the difference. The group sizes attached to the primary summaries are 39 and 43, but the registry does not report the number with missing outcome data or explain the analysis population. That gap matters because the precision and interpretation of a comparison depend on which observations contributed.

The interval is compatible with a range of possibilities rather than a single definitive conclusion. Clinical importance cannot be judged from its statistical limits alone: the registry does not report a threshold for a meaningful improvement. The result is best described as an uncertain estimate of the average difference in reported improvement, with its lower bound close to and below no difference.

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 participants per arm, not total event counts. They describe how many participants might be planned in each of two equally allocated groups under the stated assumptions. The illustrations use a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. They are based on the observed effect, with scenarios planning for 100%, 75%, or 50% of that effect. The original trial's own sample-size assumptions and justification are not reported in the registry record.

The 80% power column represents a planning target of an 80% chance of detecting the specified effect under the calculation's assumptions if that effect is present. The 90% power column uses a higher target chance of detection, which generally requires more participants for the same effect. For example, planning for 100% of the observed effect gives 100 participants per arm for 80% power and 134 participants per arm for 90% power; planning for 50% gives 398 and 533 participants per arm, respectively.

The figures rise as the assumed effect becomes smaller: a smaller contrast is harder to distinguish from random variation, so more participants are needed under otherwise similar conditions. These are planning illustrations computed from the observed effect, not observed power. They do not establish what the trial was designed to detect, whether its original assumptions were appropriate, or whether a future study would achieve the stated power.

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 = 1.300; pooled standard deviation σ = 3.272.
2. n = 2 × (1.96 + 0.842)² × 3.272² / 1.300² = 99.4, rounded up to 100 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 <- 1.3                      # observed difference in means
sigma <- 3.27204064766928                     # 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(290589)
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 = 100 (80% power), 134 (90% power)
 75% of observed effect: n per arm = 177 (80% power), 237 (90% power)
 50% of observed effect: n per arm = 398 (80% power), 533 (90% power)
power.t.test: 100.4 
Simulated power at n = 100 per arm: 0.794 (20000 replicates)

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

14. Multiplicity, Missing Data and Analysis Populations

The registry record does not report a statistical plan describing how the primary and secondary outcomes were to be tested, whether an outcome hierarchy was specified, or whether multiplicity adjustments were used. It lists three outcome measures and posts one secondary outcome, but does not explain the role of the other measures in the analysis. The number of reported analyses is zero. Consequently, the posted comparisons do not establish whether they were prespecified, exploratory, or selected from a broader set of analyses.

In general, multiplicity arises when investigators conduct several statistical tests or examine several outcomes. Without a planned approach, the chance of at least one apparently notable result can increase across a collection of tests. Adjustments or a prespecified hierarchy can help control how the findings are interpreted. No such method is reported, so the secondary result deserves particular caution and should not be treated as confirmatory evidence solely because of its numerical appearance.

Missing-data handling is also not reported in the registry record. It does not say how many participants had a primary outcome measurement available, why any measurements might be absent, or whether missing values were imputed or otherwise addressed. The group sizes attached to the primary outcome summaries are known, but they do not reveal the missingness process or the analysis population. The registry also does not identify an intention-to-treat population or another rule for inclusion in the analysis.

In general, missingness can affect interpretation when the availability of an outcome is related to participant characteristics or to the unobserved outcome itself. The direction and size of any such effect cannot be inferred without information about missingness and the analysis. The record does not report sensitivity analyses, protocol deviations, or adherence-based analyses. These gaps prevent a full appraisal of whether the posted comparisons reflect all randomized participants as intended, and of how robust the results are to alternative analysis choices.

15. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 1% reporting threshold.

Most frequent non-serious adverse event terms

The registry reports serious adverse events for participants at risk in each group. The number affected by a serious adverse event is 0 in the methylprednisolone acetate group and 0 in the placebo group. Deaths are also reported as 0 in each group. The record lists no serious adverse event terms.

These figures describe what was posted for the groups and their reported at-risk counts; they do not prove that no serious event could occur or that the groups have identical safety profiles. In general, an absence of reported events in a trial summary still leaves uncertainty about uncommon harms, particularly when the number of participants observed is limited. The registry record does not provide further detail about event definitions, seriousness criteria, the method of ascertainment, or follow-up duration for safety outcomes.

The record also includes non-serious adverse-event counts, but the serious-event figures should be interpreted separately from those counts. Neither the absence of reported serious events nor the posted adverse-event summary is sufficient by itself to establish comparative safety beyond the observation recorded in this trial.

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

The primary result describes a higher mean reported improvement in the methylprednisolone acetate group than in the placebo group. The unadjusted difference is 1.3 units on the pain-improvement scale. Its confidence interval includes zero, and the unadjusted p-value is 0.07, so the posted summaries leave uncertainty about the average difference. The result is not a precise estimate of a single effect size.

This finding does not establish that every participant experienced more improvement, nor does it indicate what proportion had a personally meaningful change. The registry record does not report a clinically important difference threshold, a responder definition, or individual-level outcome distributions. A difference between group means can coexist with varied experiences in both arms.

Nor does the posted comparison by itself prove that the intervention caused the observed difference. Randomization is a design feature intended to support causal comparison, but full interpretation depends on trial conduct, outcome availability, and the analysis population. The registry record does not report enough detail about missing data, protocol deviations, or the sponsor's analysis to assess those elements fully.

The secondary analgesic-use outcome points in a direction broadly consistent with the primary pattern, but it does not remove the primary result's uncertainty. The record does not report a multiplicity plan or explain the inferential status of secondary outcomes. The safety summary reports no serious adverse events or deaths in either group, but it cannot establish the absence of uncommon harms.

In general, statistical evidence and clinical meaning are related but distinct. A p-value does not measure clinical importance, and a confidence interval alone cannot determine whether a difference matters to patients. The registry supports a cautious description of the observed summaries and their uncertainty; it does not support a definitive claim about benefit, lack of benefit, or safety.

17. Why This Trial Matters Statistically

This trial illustrates why an estimated difference should be read together with its uncertainty. The primary mean contrast is positive, but the confidence interval extends across no difference and the unadjusted p-value is 0.07. Looking only at the direction of the point estimate would overlook how much uncertainty remains about its size.

It also shows the distinction between a continuous outcome summary and individual experience. Group means of 7.1 and 5.8 describe averages, while standard deviations of 3.0 and 3.5 indicate variation around those averages. An average contrast does not mean that all participants had the same response, and the registry does not report responder thresholds that could connect the mean difference to a clinically defined level of improvement.

The record also makes clear that registry summaries may not provide a complete statistical analysis. The posted comparison can be recalculated from the group summaries, yet the record lists no reported analyses and does not provide details about a prespecified model, missing-data handling, or multiplicity. Those omissions limit what can be inferred about the analysis choices behind the result.

Finally, planning calculations based on an observed effect are not a substitute for the original design justification. Their assumptions are stated, but the trial's own assumptions are not reported. Keeping those two purposes separate helps prevent a planning illustration from being mistaken for evidence about the trial's intended power. Together, these features make the study a useful example of reading estimates, precision, secondary outcomes, and reporting gaps as distinct parts of statistical interpretation.

18. Limitations

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

Several details needed to evaluate the primary comparison are not reported in the registry record. The record does not describe how participants were selected, provide full eligibility criteria, explain the randomization sequence or allocation concealment, or identify which parties were masked. It classifies masking as quadruple but does not explain its implementation.

The registry does not report the original sample-size justification or the trial's own statistical assumptions. It also lists zero reported analyses, so the posted summary comparison is not accompanied by a sponsor-reported model or an account of prespecified analysis decisions. The available contrast is an unadjusted recalculation from posted group summaries.

Missing-data handling and the analysis population are not described. The record does not state whether all randomized participants were included in the primary analysis, why the primary outcome group sizes differ from enrollment, or whether any measurements were missing. It also does not report adherence, protocol deviations, or sensitivity analyses. These omissions limit assessment of how robust the comparison is to analysis choices.

The registry describes the pain outcome but does not report a threshold for clinically important improvement or a responder definition. It also does not provide individual-level outcome distributions. As a result, the clinical meaning and spread of participants' experiences cannot be fully assessed from the means and standard deviations alone.

Only one secondary outcome is posted, although three outcome measures are listed. The record does not report multiplicity control or clarify the role of other measures. Safety reporting gives serious-event counts but lacks further detail about ascertainment and follow-up. These limitations do not invalidate the posted summaries, but they constrain how far conclusions can extend beyond those summaries.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: A Trial of Corticosteroids for Low Back Pain — NCT00290589.
• Registry record last updated 2019-08-07; status last verified 2019-08.
• Results first posted 2019-08-07.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

22. Record Summary

The registry record describes a completed Phase 3 randomized, quadruple-masked, parallel-group trial sponsored by Montefiore Medical Center. It compares intramuscular methylprednisolone acetate, described as 160mg, with a placebo intramuscular injection for radicular low back pain.

The primary outcome is improvement on a 0-to-10 Numerical Rating Scale between the emergency department visit and a telephone call at 1 month. The posted group means are 7.1 for methylprednisolone acetate and 5.8 for placebo. An unadjusted recalculation gives a mean difference of 1.3, with a confidence interval that includes zero and an unadjusted p-value of 0.07.

The registry lists analgesic use as a secondary outcome and reports no serious adverse events or deaths in either group. It does not report the trial's original sample-size justification, sponsor analyses, missing-data approach, or multiplicity plan. Interpretation is therefore limited to the posted summaries and their uncertainty.