Neuroimaging in Patients Undergoing TMS for Depression: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Neuroimaging in Patients Undergoing TMS for Depression (NCT02974296), a randomized na study. The primary outcome is Change in Functional Connectivity Measured by Resting MRI.

2. Clinical Question

3. Trial Design

The registry identifies this as a randomized, double-masked, parallel study of transcranial magnetic stimulation. It lists New Target TMS and Standard TMS as the comparison arms. The former is described as MRI-guided; the latter as standard TMS. The registry lists enrollment as 33 and the study status as completed. The recorded start date is 2017-04-30, and the primary completion date is 2019-08-20.

The stated project goals include using functional MRI to guide TMS for people with depression who did not respond to standard TMS, and exploring whether resting-state functional connectivity maps can predict responsiveness among people beginning standard TMS. These goals provide context, but they do not establish that every goal was evaluated in the posted randomized comparison. The registry record does not report detailed eligibility criteria, the allocation procedure, the masking roles, or the timing and conduct of treatment sessions.

The primary outcome is change in resting MRI correlation coefficients, defined as post-treatment minus baseline. The timeframe is Week 1 before the first TMS treatment and week 36 after completion of TMS treatments. The arm-level outcome record lists 10 participants for New Target TMS and 22 for Standard TMS. These are the outcome-specific group counts shown in the record; they should not be treated as a description of every participant’s analysis status or of unreported populations.

The registry lists one secondary outcome: change in depressive symptoms measured using the 17-item Hamilton Depression Rating Scale. It also lists two outcome measures overall, zero reported analyses, and a completed study. Further details about the study’s original statistical design and analysis procedures are 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.

Overall Study

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Change in Functional Connectivity Measured by Resting MRI, measured over Week 1 prior to first TMS treatment and week 36 after completion of TMS treatments.

The primary endpoint is change in resting MRI correlation coefficients between the left dorsolateral prefrontal cortex and the subgenual anterior cingulate. The registry defines change as the post-treatment value minus the baseline value. Measurements are described for Week 1 before the first TMS treatment and week 36 after completion of TMS treatments. The endpoint is therefore a change score for a functional-connectivity measure, rather than a direct rating of depressive symptoms.

Functional connectivity describes correlation between activity in specified brain regions. In general, a change score can summarize how a measured relationship differs between two assessment times. Comparing the change scores between randomized groups addresses whether the recorded change differs between the MRI-guided targeting strategy and standard TMS in this study. The registry’s stated project rationale is that individualized functional MRI may help guide targeting; the endpoint evaluates a connectivity measure relevant to that rationale.

This measure does not, by itself, establish how participants felt or functioned, nor does it establish a clinical benefit. The registry lists depressive symptoms as a separate secondary outcome using the 17-item Hamilton Depression Rating Scale. A brain-connectivity measure and a symptom measure describe different aspects of an outcome, so a finding on one should not be substituted for a result on the other. The record does not report a threshold for interpreting a change in connectivity as clinically important.

7. Results Posted in the Registry

• New Target TMS: mean 0.10 versus -0.76 on control; difference 0.87 (95% CI 0.76 to 0.98), standardized effect d = 6.89; 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.

Change in Depressive Symptoms Measured by the Hamilton Depression Rating Scale (17-item HDRS)

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • New Target TMS: mean 64.94 versus 27.73 on control; difference 37.21 (95% CI 22.37 to 52.05), standardized effect d = 1.52; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for New Target TMS than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry lists one secondary outcome, change in depressive symptoms measured using the 17-item Hamilton Depression Rating Scale. The posted group means and unadjusted comparison show a larger average percentage change for New Target TMS than for Standard TMS. The comparison’s reported interval is above zero and the unadjusted p-value is < 0.001. These are the posted descriptive and inferential results for the secondary measure; they do not change what the primary connectivity endpoint measures.

At a broad level, the primary and secondary comparisons point in the same direction: each has a larger recorded change in the New Target TMS arm. This directional consistency may be of interest because one outcome measures functional connectivity and the other measures depressive symptoms. However, the endpoints are distinct, use different units, and address different aspects of the study. Their effect sizes cannot be compared directly, and agreement in direction does not establish that one change caused the other.

The registry lists two outcome measures overall, but only one secondary outcome is described, and it lists zero reported analyses. It does not report a hierarchy of secondary hypotheses, a multiplicity adjustment, or a plan for interpreting secondary findings. Without such information, unadjusted secondary p-values carry less weight than a result tied to a clearly specified primary analysis: if several outcomes or analyses are considered, the chance of finding a small p-value somewhere can increase. The record does not establish which multiplicity framework, if any, was intended.

The symptom result should therefore be read as a posted secondary comparison that is compatible with the direction of the primary result, rather than as independent confirmation or a comprehensive assessment of clinical outcomes. The registry does not report other symptom, functioning, or quality-of-life results in the record described, and it does not provide a threshold for clinically meaningful score change.

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

• Change in Depressive Symptoms Measured by the Hamilton Depression Rating Scale (17-item HDRS)

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: it is a change in correlation coefficients. The record reports an arm mean and standard deviation for each group, followed by a mean difference and confidence interval for the comparison. These quantities answer different questions. A mean locates the group’s average change; a standard deviation describes how dispersed individual changes are around that average; and the mean difference contrasts the two group averages.

What does the mean difference represent?

The posted contrast is New Target TMS minus Standard TMS. A positive value means the average change is higher in the New Target TMS group, using the registry’s post-minus-baseline definition. For the primary outcome, the group means have opposite signs, so the direction of the contrast reflects a positive average change in one group and a negative average change in the other. The contrast does not mean that every participant changed in the same direction, and group averages do not describe the full distribution of individual responses.

The mean difference is an unadjusted comparison based on the posted summary statistics. It does not account for covariates unless an adjustment is specified, and an adjustment is not reported in the registry record. The record contains no reported analyses, so a fuller account of the analysis model, assumptions, or analysis population is not available. These recalculated summary-based values should be distinguished from a documented sponsor analysis; the record does not flag a discrepancy.

What does the standard deviation tell the reader?

In general, a standard deviation describes variability among observed values within a group. It is not the standard error of the mean and is not itself a measure of uncertainty around the group average. A smaller standard deviation indicates that the recorded values are more concentrated around the mean, while a larger one indicates more spread. Comparisons of standard deviations should be interpreted with the group sizes and outcome context in mind. The registry reports standard deviations for both arms, but does not provide participant-level distributions or plots in the record described.

Because the primary outcome is a continuous measure, the mean and standard deviation are natural summary statistics. For counts or event rates, different models may be appropriate. Over-dispersion is a general concern when count data vary more than a count model expects; it is not the central issue for this continuous change-score endpoint. The registry record does not report a model-based analysis that would allow further assessment of distributional assumptions.

How should the p-value be read?

The primary unadjusted p-value is < 0.001. In general, a p-value measures how incompatible the observed data are with a specified null model, assuming that model and the calculation’s assumptions hold. It is not the probability that there is no group difference, and it does not quantify clinical importance. The registry does not report the full testing procedure or any adjustment for multiple outcomes. The unadjusted label therefore matters: it describes the reported comparison without an identified multiplicity correction.

A p-value does not replace the estimated difference and its confidence interval. The estimate describes the direction and scale of the observed contrast, while the interval conveys uncertainty about the contrast under the calculation’s assumptions. Both are relevant to interpretation. The posted interval for the primary comparison lies above zero, consistent with the positive estimated difference and small unadjusted p-value.

What does a standardized effect add?

The record also gives a standardized effect size, d, for each comparison. In general, a standardized effect expresses a mean difference relative to a measure of variability, making the contrast unitless. The primary standardized effect is 6.89. Such a value can help describe the contrast relative to observed dispersion, but it does not establish that the endpoint is clinically meaningful. In small groups, standardized estimates can be especially sensitive to the observed variability, and the registry does not report an adjusted or bias-corrected effect-size analysis.

The standardized effect should not be mistaken for a change in the original correlation-coefficient units. The original-scale mean difference and its confidence interval retain direct connection to the measured endpoint, while the standardized effect gives a scale-relative description. Neither one explains the mechanism behind the result or makes the comparison generalizable beyond the study population without further evidence.

Overall, these methods support a descriptive reading of the posted group contrast and its uncertainty. They do not recover the full protocol, establish the intended primary analysis, or reveal how missing measurements and multiple outcomes were handled. Those design and analysis details are not reported in the registry record.

11. Confidence Intervals and Precision

The primary comparison is reported with a confidence interval whose lower and upper bounds are both positive. This is consistent with the positive mean difference and the unadjusted p-value of < 0.001. The interval describes a range of values compatible with the observed data under the calculation’s assumptions; it does not mean that there is a 95% probability that the true effect lies within that particular interval.

In general, confidence-interval width reflects precision: narrower intervals indicate less sampling uncertainty under the model, while wider intervals indicate greater uncertainty. Precision depends on factors including group sizes and variability. Here the outcome counts are 10 for New Target TMS and 22 for Standard TMS, and the reported standard deviations differ between the arms. The record does not provide participant-level data or a detailed model description to evaluate how other assumptions affect interval width.

For the secondary outcome, the posted mean-difference interval also lies above zero. Its limits are farther from one another than the primary interval when expressed in their respective outcome units, but the endpoints use different scales and should not be compared directly by raw interval width. The secondary comparison is unadjusted, and the registry does not report multiplicity handling. A confidence interval for an individual contrast does not account for the possibility of selecting among multiple outcomes unless that adjustment is built into the method.

These intervals describe statistical precision, not clinical importance. The registry does not report a clinically meaningful threshold for either the connectivity change or the symptom measure. The observed direction and statistical separation therefore need to be considered alongside the small outcome-group counts, the lack of reported analysis details, and the absence of an established interpretation threshold.

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 events. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set as a stated fraction of the observed effect. The original design assumptions for the trial itself are not reported in the registry record.

For the primary connectivity outcome, planning for 100% of the observed effect gives 1 participant per arm for 80% power and 1 per arm for 90% power. Planning for 75% gives 1 per arm for 80% power and 1 per arm for 90% power. Planning for 50% gives 2 per arm for 80% power and 2 per arm for 90% power. For the secondary symptom outcome, planning for 100% gives 7 participants per arm for 80% power and 10 per arm for 90% power; planning for 75% gives 13 and 17 per arm, respectively; and planning for 50% gives 28 and 37 per arm, respectively.

The 80% power column describes the planned number per arm under which the calculation would have an 80% chance of detecting the specified effect, if the stated assumptions hold. The 90% column describes the corresponding planning target for a 90% chance. Smaller effects require more participants under these assumptions, as illustrated by the increasing requirements when the target is reduced from the full observed effect to half of it. These are planning illustrations based on the observed effects, not observed power and not evidence that the trial was designed with these assumptions or sizes.

13. How the Sample Size Is Derived

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

Formula

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

Worked calculation for this trial

1. Observed difference Δ = 0.867; pooled standard deviation σ = 0.126.
2. n = 2 × (1.96 + 0.842)² × 0.126² / 0.867² = 0.3, rounded up to 1 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- 0.867140265                      # observed difference in means
sigma <- 0.125847388888487                     # 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(2974296)
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 = 1 (80% power), 1 (90% power)
 75% of observed effect: n per arm = 1 (80% power), 1 (90% power)
 50% of observed effect: n per arm = 2 (80% power), 2 (90% power)
power.t.test: 1.9 
Simulated power at n = 1 per arm: NA (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

Multiplicity arises when a study evaluates more than one outcome, time point, subgroup, or analysis and interprets the resulting statistical tests. Repeated testing can make at least one apparently unusual result more likely, even when no true effect exists. In general, a prespecified testing hierarchy or an adjustment procedure can address this concern, but the registry record does not report whether either was planned or used.

The record lists two outcome measures overall, including the primary connectivity outcome and one posted secondary outcome on depressive symptoms. It reports an unadjusted p-value for each comparison. No multiplicity correction is described. The record also lists zero reported analyses, so the exact analysis set and any additional tests cannot be assessed. The unadjusted secondary result should consequently be treated as less definitive than a result supported by a reported testing plan.

Missing-data procedures are not reported in the registry record. The outcome arm counts are 10 and 22 for the primary measure, but the record does not explain whether these counts represent all randomized participants, all participants with a measurement at both time points, or another analysis set. It does not report reasons for missing measurements, imputation, sensitivity analyses, or whether participants were analyzed according to their randomized assignment. The listed enrollment alone does not resolve these questions.

The record does not describe an intention-to-treat, per-protocol, or safety analysis population. It also does not report interim analyses or stopping rules. These matters should be treated as unreported rather than inferred from the randomized design or completed status. Without analysis-population and missing-data information, it is difficult to determine how well the reported comparisons preserve the benefits of randomization or how sensitive they may be to incomplete follow-up.

These omissions do not negate the descriptive results, but they constrain how fully they can be audited. A complete interpretation would require details of the planned tests, multiplicity strategy, missingness, analysis populations, and handling of departures from the assigned treatment approach; those details are not reported in the registry record.

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 0% reporting threshold.

The registry reports zero serious adverse events for New Target TMS among 10 participants at risk, and zero for Standard TMS among 22 participants at risk. It also reports zero other adverse events in both groups and zero deaths in both groups. These are the adverse-event figures recorded for the listed groups.

Zero reported events means no events are recorded in these categories for the groups shown; it does not establish that there is no risk. The group counts are limited, and uncommon events may not appear in a study of this size. The record does not provide event descriptions, severity details beyond the serious-versus-other categories, follow-up-specific exposure information, or a fuller safety analysis. The absence of listed serious-event terms is consistent with the zero serious-event counts, but should not be interpreted as evidence about risks outside the registry’s recorded observation.

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

The posted results show a statistically clear difference between the randomized groups in change in the specified resting MRI correlation measure. The mean difference is positive, its confidence interval lies above zero, and the unadjusted p-value is < 0.001. The secondary symptom comparison points in the same direction and also has an unadjusted p-value of < 0.001. These statements describe the posted results, not a general conclusion about every person or setting.

The primary outcome is a brain connectivity measure, not a direct measure of depression symptoms. The registry also records a symptom outcome, but its presence does not make the imaging measure a clinical endpoint or establish that changes in connectivity correspond to meaningful symptom changes. The record does not report a threshold for clinical importance on either scale. It also does not report enough detail about analysis populations, missing data, or multiplicity to fully evaluate the robustness of the results.

The findings do not establish the mechanism through which MRI-guided targeting might relate to connectivity or symptoms. They do not show that a particular participant would experience a specific outcome, nor do they establish that either approach should be chosen in clinical practice. The record’s goals and posted group comparisons are not a substitute for a complete trial report or evidence across other populations.

The completion status and posted outcome summaries indicate that the registry contains results for a completed study. The small outcome-group counts and limited analysis documentation mean the effect estimates should be interpreted as results from this study rather than as definitive estimates that necessarily generalize. The appropriate statistical reading is that the observed group contrasts are separated on the reported measures, while clinical meaning, reproducibility, and broader applicability remain uncertain from the registry record alone.

17. Why This Trial Matters Statistically

This study raises a statistical question that is increasingly relevant when treatment targeting depends on measurements of brain function: whether a change in an imaging marker differs between a targeted approach and a standard approach. The primary outcome is explicitly a change in functional connectivity between two named regions. That makes the result interpretable as a comparison on a defined imaging measure, rather than as a vague claim about brain activity.

The record also illustrates why outcomes need to be kept distinct. Connectivity and depressive symptoms represent different constructs and are measured on different scales. The direction of the posted comparisons is consistent across the primary and secondary outcomes, but that consistency does not make the imaging result a direct measure of symptom benefit or establish a causal pathway between them. Clear separation of endpoint meanings is essential when interpreting multidomain trials.

Another statistical lesson is the importance of uncertainty and analysis transparency. The posted primary confidence interval lies above zero, and the unadjusted p-value is small, yet the outcome groups are limited in size and the registry does not report several important analysis details. An apparently strong contrast can still be difficult to generalize when the record does not explain missingness, analysis populations, multiplicity, or the original design assumptions.

The planning illustrations also show how sample-size requirements depend on the effect selected for planning. Calculations based on the full observed effect can yield very different planning values from calculations based on a smaller fraction. Because planning from an observed effect can be optimistic, these illustrations are not a substitute for a prospectively justified design. They are useful as conditional calculations, not as proof of adequate planning or observed power.

18. Limitations

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

The registry record leaves important study-specific details unavailable. Eligibility criteria are not reported, and the summary’s discussion of people who did not respond to standard TMS alongside people beginning standard TMS does not fully specify how those groups relate to the randomized outcome comparison. The arm descriptions identify MRI-guided New Target TMS and Standard TMS, but do not provide additional treatment-delivery details or explain the exact targeting procedures.

The primary outcome is defined as a post-minus-baseline change in resting MRI correlation, with assessment described before treatment and after completion of TMS treatments. The record does not report the number of observations available at each assessment, how incomplete scans were handled, or whether the same participants contributed to both time points. It also does not report the analysis population or whether the comparison was adjusted for baseline values or other variables.

Although the study is listed as randomized and double-masked, the roles of those masked and the allocation procedure are not reported. The record does not describe a prespecified statistical analysis plan, multiplicity control, interim analyses, or the original sample-size rationale. It lists zero reported analyses. These omissions make it difficult to assess whether the posted comparisons were planned in advance or how they relate to the full set of analyses conducted.

The posted outcome counts are limited, and the registry provides aggregate means and standard deviations rather than participant-level data. Such summaries do not reveal the distribution of individual changes, influential observations, or potential subgroup patterns. The record also provides no clinically meaningful threshold for either the connectivity measure or symptom change. Finally, the safety record lists zero events in the categories shown, but the detail available is limited and cannot establish the absence of risk.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Neuroimaging in Patients Undergoing TMS for Depression — NCT02974296.
• Registry record last updated 2020-10-23; status last verified 2020-10.
• Results first posted 2020-10-23.
• 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.

22. Record Summary

The registry record describes a completed randomized, double-masked, parallel comparison of MRI-guided New Target TMS with Standard TMS for depression. The listed enrollment is 33. The study start date is 2017-04-30, and the primary completion date is 2019-08-20.

The primary outcome is change in resting MRI correlation between the left dorsolateral prefrontal cortex and subgenual anterior cingulate, measured before treatment and after completion of TMS treatments. The posted primary comparison has a positive mean difference, a confidence interval above zero, and an unadjusted p-value of < 0.001. The secondary outcome is change in depressive symptoms measured with the 17-item Hamilton Depression Rating Scale; its posted comparison also has an unadjusted p-value of < 0.001.

The registry lists zero reported analyses. Details of the original design assumptions, analysis populations, missing-data handling, and multiplicity procedures are not reported in the registry record. Adverse-event counts are zero in the categories recorded for both groups.