ECT Pulse Amplitude and Medial Temporal Lobe Engagement: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for ECT Pulse Amplitude and Medial Temporal Lobe Engagement (NCT02999269), a randomized phase 4 study. The primary outcome is Depression Severity.

2. Clinical Question

The primary posted outcome was the Hamilton Depression Rating Scale with 24 items, measured after ECT at 4 weeks after study initiation. The registry reports a full longitudinal model for this outcome. The posted group summaries give mean scores for the 600 mA, 700 mA, and 800 mA groups, while the reported sponsor analyses give p-values for analyses involving all three groups. The registry does not report enough detail to identify the precise contrast or outcome represented by each of those two p-values.

This is a comparison among three active ECT amplitude settings, not a comparison with no ECT or with a non-ECT control. The clinical question is therefore about differences across the tested settings in the registered outcome, rather than whether ECT itself is better than another treatment.

3. Trial Design

The registry describes NCT02999269 as a completed Phase 4, randomized, double-masked, parallel-group trial sponsored by the University of New Mexico. It lists enrollment of 62 and identifies major depressive disorder as the condition. The study start is listed as 2016-10, and primary completion as 2020-03-23.

Participants were assigned to one of three right unilateral ECT amplitude groups: 600 mA, 700 mA, or 800 mA. Each arm description names the MECTA Spectrum 5000Q Amplitude device. The registry does not report the allocation ratio, the randomization procedure, or who was masked. It also does not specify in the record how masking was maintained across the different amplitude settings.

The primary outcome is depression severity on the Hamilton Depression Rating Scale with 24 items, with a post-ECT timeframe described as 4 weeks after study initiation. Posted outcome summaries include 16 participants in the 600 mA group, 15 in the 700 mA group, and 16 in the 800 mA group. These outcome counts are not the same as the registry’s stated enrollment; how the recorded outcome counts relate to all enrolled participants is not reported in the registry record.

The registry identifies a full longitudinal model with an unstructured repeated-measures covariance matrix for the primary outcome. The record does not report the visit schedule, the number of repeated measurements included in the model, details of the analysis population, or the original sample-size justification. It also does not give the protocol-level statistical analysis plan in enough detail to establish which analyses were prespecified.

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

Analysis population for the primary outcome, as defined in the registry: full longitudinal model with an unstructured repeated measures covariance matrix

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 Depression Severity, measured over post-ECT Hamilton Depression Rating Scale -24 item. The time frame is 4 weeks after study initiation..

The primary endpoint measures depression severity using the Hamilton Depression Rating Scale with 24 items. The registry describes scores from 0 to 76, with higher scores indicating more depression severity. The listed timeframe is post-ECT, 4 weeks after study initiation. The posted results present a mean score for each amplitude group and a standard error for that mean.

A depression-severity scale is used to represent symptom burden in a form that can be compared across groups and over time. In general, a lower score on a scale where higher scores indicate greater severity represents a lower measured severity score at that assessment. This interpretation describes the scale, not a conclusion about the overall clinical value of any treatment setting.

The registry says the primary outcome was analyzed using a full longitudinal model with an unstructured repeated-measures covariance matrix. In general, a longitudinal model uses measurements taken over time and accounts for the fact that observations from the same participant may be related. An unstructured covariance matrix allows the model to estimate different relationships among the repeated measurements rather than imposing a single fixed pattern. The registry does not give the model’s full specification or identify the measurements included in the analysis.

The posted arm means are endpoint summaries. The registry does not provide baseline scores in these results, so they should not be interpreted as within-person changes from baseline. It also does not post secondary outcome results in the record.

7. Results Posted in the Registry

• 700 mA Right Unilateral ECT: mean 12.00 versus 22.50 on control; difference -10.50 (95% CI -16.60 to -4.40), standardized effect d = -1.21; z-test p = <0.001.
• 800 mA Right Unilateral ECT: mean 14.10 versus 22.50 on control; difference -8.40 (95% CI -14.92 to -1.88), standardized effect d = -0.89; z-test p = 0.012.

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies the primary measure as a continuous score and reports arm-level means with standard errors. A mean describes the arithmetic center of the observed scores in a group. A standard error describes the uncertainty in estimating that group mean; it is not the spread of individual scores. In general, individual-level variability is described by a standard deviation, whereas the standard error concerns precision of the estimated mean.

The primary analysis is described as a full longitudinal model with an unstructured repeated-measures covariance matrix. In general, longitudinal analysis is appropriate when the same participants contribute measurements at multiple time points. Measurements from one participant tend to be related, so a model needs to account for that dependence rather than treating every measurement as independent. An unstructured covariance specification lets the model estimate the relationships among repeated measurements without requiring them to follow a simpler preset pattern. The registry does not report the number or timing of measurements included, the model’s covariates, or its full estimation procedure.

The sponsor’s reported method is linear regression. Two superiority analyses are listed for the three groups, with p-values of 0.04 and < 0.001. The record does not provide the corresponding estimates, confidence intervals, or enough information to determine how the two analyses differ. It also does not report whether the analyses were adjusted for other variables, whether they used the longitudinal model description, or which population was included in each sponsor analysis.

The pairwise mean differences and confidence intervals discussed in the interpretation are recalculated, unadjusted comparisons based on the posted group summary statistics. They compare the 700 mA and 800 mA groups separately with the 600 mA group. These calculations do not reproduce the full longitudinal model, and they do not incorporate model adjustments that are not represented in the summary statistics. The registry flags a discrepancy between the sponsor’s analysis and the recalculations but does not explain it. The sponsor’s own analysis takes precedence.

In general, a confidence interval gives a range of estimates compatible with the data and the method used to calculate the interval. A narrower interval indicates greater precision than a wider interval, all else being equal. The registry does not report the confidence level for the recalculated intervals, so no confidence percentage should be inferred from the record. The p-values are also limited without a detailed description of the hypotheses and analysis plan.

Multiplicity arises when several outcomes, time points, contrasts, or analyses are tested. In general, considering many tests can increase the chance of obtaining at least one apparently notable result under a setting where no true effects exist. The record includes two sponsor p-values and three arms, but it does not report whether those analyses were part of a prespecified testing sequence or whether any multiplicity adjustment was used. The existence of two p-values alone does not establish what testing strategy was intended.

Missing-data handling is not reported in the registry record. The posted endpoint counts differ by arm, and the record does not explain whether these counts reflect missing measurements, exclusions, or another analysis-population definition. The primary outcome population is described as a full longitudinal model, but the record does not state whether all randomized participants were included, how partial measurements were treated, or whether any imputation was used. The distinction matters because results can depend on how participants with incomplete follow-up are handled.

In general, randomized assignment supports comparisons among assigned groups when the analysis preserves that assignment and addresses missingness appropriately. The registry does not say whether the sponsor analysis followed an intention-to-treat approach or another population definition. It also does not report a per-protocol analysis or sensitivity analyses. Those details are needed to evaluate how robust the reported findings are to analysis choices, and they should not be inferred from the model label alone.

For interpretation, the sponsor’s reported analysis and the unadjusted recalculations answer related but not necessarily identical statistical questions. The sponsor analysis is the primary reported analysis in the record, but its details are limited. The recalculations provide numerical comparisons from posted means and standard errors, with the important limitation that they do not reproduce the sponsor’s stated longitudinal model.

10. Confidence Intervals and Precision

The recalculated pairwise comparisons have confidence intervals that quantify uncertainty around the estimated mean differences. For 700 mA compared with 600 mA, the estimate is -10.5 and the interval runs from -16.6 to -4.396. For 800 mA compared with 600 mA, the estimate is -8.4 and the interval runs from -14.92 to -1.885. Both intervals are entirely below zero, consistent with lower mean scores in the higher-amplitude groups in these unadjusted comparisons.

The interval widths indicate that the estimates are not exact values: a range of differences remains compatible with each calculation. The 700 mA comparison has a more negative point estimate than the 800 mA comparison, but the intervals overlap. This does not itself test whether the 700 mA and 800 mA groups differ from each other; no direct pairwise comparison between those groups is posted. The registry also does not provide a confidence interval for either group mean or for the sponsor’s reported regression analyses.

The confidence level for the recalculated intervals is not reported in the registry record. The posted standard errors for the arm means are 2.3, 2.1, and 2.4. In general, standard errors describe precision of estimated means, while confidence intervals communicate uncertainty for a specified estimate under a calculation method. Neither measure describes the full range of individual participant scores.

These precision statements apply to the unadjusted recalculations from summary statistics, not necessarily to the sponsor’s full longitudinal model. The discrepancy flagged in the registry record is not explained, so the sponsor’s own analysis takes precedence. The available record does not provide enough detail to judge how much precision would change under the sponsor model or under another missing-data or covariate strategy.

11. Planning a Similar Trial: Sample Size

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

The sample-size figures are planning illustrations based on the observed effect, not the original trial’s design justification. The original design assumptions are not reported in the registry record. The illustrations use a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. They consider effects equal to 100%, 75%, and 50% of the observed effect.

The figures are participants per arm for 80% power and participants per arm for 90% power. In plain terms, 80% power means that, under the assumptions used for the illustration and if the stated effect were real, a study of that planned size would have an 80% chance of detecting an effect at the specified significance level. The 90% column represents a higher chance of detection under the same assumptions, and therefore requires a larger planned sample for a given effect size.

The comparisons are pairwise: 700 mA versus 600 mA, and 800 mA versus 600 mA. The figures should be read separately for each comparison rather than as a total sample for all three arms. The planning calculations assume equal allocation between the two arms in each comparison. They do not account for unreported features of the original trial design or establish how the original enrollment target was chosen.

Each row scales the effect to be detected relative to the corresponding observed difference. A smaller effect is harder to detect, so the planning illustration requires more participants per arm for the smaller-effect rows. These are hypothetical design calculations, not observed power; observed power is not a measure of how persuasive the trial result is.

12. How the Sample Size Is Derived

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

Formula

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

Worked calculation for this trial

1. Observed difference Δ = -10.500; pooled standard deviation σ = 8.701.
2. n = 2 × (1.96 + 0.842)² × 8.701² / -10.500² = 10.8, rounded up to 11 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 <- -10.5                      # observed difference in means
sigma <- 8.70136731229341                     # 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(2999269)
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 = 11 (80% power), 15 (90% power)
 75% of observed effect: n per arm = 20 (80% power), 26 (90% power)
 50% of observed effect: n per arm = 44 (80% power), 58 (90% power)
power.t.test: 11.8 
Simulated power at n = 11 per arm: 0.767 (20000 replicates)

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

13. Analyses Reported by the Sponsor

The registry reports two sponsor analyses using linear regression for the three groups, with p-values of 0.04 and < 0.001. It does not report effect estimates, confidence intervals, or enough information to determine which result corresponds to which outcome or analysis specification. The registry also flags a discrepancy between the sponsor’s analysis and the recalculations, without explaining what differs.

The recalculated pairwise results use the posted arm means and standard errors to compare 700 mA with 600 mA and 800 mA with 600 mA. They are simple, unadjusted re-calculations and are not a reproduction of the full longitudinal model described for the primary outcome. The p-values and intervals from those comparisons therefore should not be treated as a substitute for the sponsor’s analysis.

When the record indicates a discrepancy, the sponsor’s own analysis takes precedence. The posted summary does not explain the discrepancy or provide the additional information needed to reconcile the results. Accordingly, the recalculations are best understood as a separate descriptive check on the summary statistics, while the reported sponsor analyses remain the registry’s own inferential results.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists three amplitude groups, two outcome measures, and two sponsor analyses with reported p-values. It does not report whether the analyses correspond to different outcomes, time points, or other hypotheses. No multiplicity adjustment or prespecified testing hierarchy is described in the registry record. In general, when several statistical tests are conducted, the interpretation of individual p-values can depend on whether and how the testing plan controls for multiple comparisons.

The primary outcome is described as a full longitudinal model with an unstructured repeated-measures covariance matrix. However, the record does not specify which visits or repeated measurements were included, whether any baseline adjustment was used, or how the model was parameterized. It also does not report whether the two sponsor analyses used the same population and model. The p-values therefore cannot be mapped confidently to particular contrasts or outcomes from the record alone.

Missing-data procedures are not reported in the registry record. The outcome summaries list 16 participants in the 600 mA group, 15 in the 700 mA group, and 16 in the 800 mA group, while overall enrollment is listed as 62. The record does not explain the relationship between enrollment and the outcome counts, the reasons for any incomplete data, or how incomplete longitudinal observations were handled.

The record does not state whether the analysis included all randomized participants, defined an evaluable population, or used another analysis set. It does not report imputation, sensitivity analyses, or a method for assessing how missing observations might affect the results. The phrase “full longitudinal model” alone does not establish how missing data were handled or which participants contributed to the reported estimates.

These omissions limit assessment of the sponsor findings and their robustness. They do not prove that the analyses were inappropriate; they mean that important details needed to evaluate multiplicity, missingness, and analysis populations 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 5% reporting threshold.

Most frequent non-serious adverse event terms

The registry lists serious adverse events for 20 participants at risk in each of the three groups. The number affected by serious adverse events is 0 in the 600 mA group, 0 in the 700 mA group, and 0 in the 800 mA group. Deaths are listed as 0 for each group. No serious adverse-event terms are listed.

Other adverse events are listed for 5 participants in the 600 mA group, 6 in the 700 mA group, and 6 in the 800 mA group, with 20 participants at risk in each group. These are distinct from the serious adverse-event counts. The registry does not provide further details about the other adverse events in the information, so their nature, timing, severity, and relationship to treatment cannot be characterized from these figures.

“At risk” refers to the denominator reported for the adverse-event results. It differs from the endpoint counts reported for depression severity, and the record does not explain the different analysis populations. The safety counts should therefore be read as the registry’s adverse-event reporting, not assumed to describe every enrolled participant or every event experienced during the trial. Zero reported serious adverse events does not establish the absence of all possible harms.

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

The posted primary outcome summaries show lower mean depression-severity scores at the stated post-ECT assessment in the 700 mA and 800 mA groups than in the 600 mA group. The unadjusted recalculations quantify those differences relative to 600 mA and include confidence intervals. The sponsor’s own reported analyses are linear-regression superiority analyses across the three groups, with p-values of 0.04 and < 0.001, but the record does not provide their effect estimates or explain which analyses those values represent.

The discrepancy flagged in the record is not described. The sponsor’s analysis therefore takes precedence over the recalculations, which are based only on posted summary statistics. Neither the sponsor p-values nor the recalculated comparisons, considered alone, establish the clinical importance of a score difference. The registry does not provide a threshold for interpreting the magnitude of score changes, and the posted endpoint summaries are not baseline-to-follow-up changes.

The trial compares three active ECT amplitude settings. It does not provide a comparison with no ECT or another treatment. The results do not establish that any amplitude setting is optimal, nor do they determine how findings apply to people outside this trial. The registry’s brief summary describes aims related to brain changes, clinical outcomes, and memory side effects, but the posted primary outcome is depression severity. The record does not post secondary outcome results, so the primary score summaries cannot answer those broader questions.

Adverse-event reporting lists no serious adverse events and no deaths among the participants at risk reported for each arm. Other adverse events are listed, but their details are not available in the record. These counts describe the registry entries; they do not establish that no other harms occurred.

Overall, the results support a description of the posted group scores and the reported analyses, with important limits from missing model detail, absent sponsor estimates, and the unexplained discrepancy. They should not be turned into a treatment recommendation or a claim about outcomes the registry did not report.

17. Why This Trial Matters Statistically

This trial raises a statistical question about comparing several active settings of a treatment parameter rather than comparing treatment with no treatment. That makes the choice of contrast important: the posted recalculations compare each higher amplitude with 600 mA, while the sponsor reports analyses involving all three groups. Those are related but not necessarily identical questions.

The record also illustrates why an estimate and its precision are useful alongside a p-value. The recalculated mean differences and confidence intervals communicate the direction and uncertainty of two specific comparisons. The sponsor p-values indicate evidence under the sponsor analyses but, without corresponding estimates and model details, provide limited information about magnitude. The discrepancy noted in the record makes it especially important to distinguish sponsor results from unadjusted recalculations.

Repeated measurements call for attention to within-participant dependence. The registry’s use of a longitudinal model and an unstructured covariance matrix recognizes that repeated scores may be related, but the lack of details about visits, missing data, and analysis populations prevents a full appraisal of that model. Transparent reporting of those choices is important for interpreting both the point estimates and their uncertainty.

The sample-size illustrations also demonstrate that planning depends on the effect a study is intended to detect: smaller effects require more participants for the same target power under the stated assumptions. Those calculations are not evidence about the power achieved by this trial and do not reveal the original design rationale.

Finally, this record separates clinical aims from the outcome evidence actually posted. The brief summary discusses brain changes and memory-related side effects, while the primary results reported are depression-severity scores. Keeping the question, endpoint, analysis, and available safety information distinct helps avoid extending a statistical finding beyond the data recorded.

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 reports enrollment of 62, while the posted primary outcome summaries list 16 participants in the 600 mA arm, 15 in the 700 mA arm, and 16 in the 800 mA arm. The record does not explain how these endpoint counts relate to the enrolled population, why counts differ, or whether the counts represent all participants included in the longitudinal analysis.

The sponsor’s two reported linear-regression analyses have p-values but no posted estimates or confidence intervals. The distinction between those analyses is not reported. The registry flags a discrepancy with the unadjusted recalculations but does not describe its nature. As a result, the sponsor’s findings cannot be fully reconciled with the summary-based pairwise calculations; the sponsor’s own analysis takes precedence.

The registry does not report the original sample-size justification, the allocation ratio, the randomization details, or who was masked. It does not provide the full statistical analysis plan, the model’s covariates, the visit schedule, the handling of missing data, or the analysis-population definition. Multiplicity procedures and sensitivity analyses are also not reported.

The primary result concerns a depression-severity score at a post-ECT assessment. Baseline scores and change-from-baseline results are not posted, and no secondary outcome results are posted. Although the brief summary describes research aims involving brain changes and memory side effects, the available outcome results do not provide results on those measures.

Safety reporting lists serious adverse events and deaths, as well as counts for other adverse events, but does not provide details about the other events. The adverse-event denominators also differ from the primary endpoint counts, and the relationship between those populations is not explained. These limits constrain interpretation of the registry record and should not be replaced by assumptions about the study conduct.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: ECT Pulse Amplitude and Medial Temporal Lobe Engagement — NCT02999269.
• Registry record last updated 2023-07-24; status last verified 2023-06.
• Results first posted 2023-07-24.
• 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

NCT02999269 is a completed Phase 4 randomized, double-masked, parallel-group trial sponsored by the University of New Mexico. The registry lists enrollment of 62 and three right unilateral ECT groups using the MECTA Spectrum 5000Q Amplitude device: 600 mA, 700 mA, and 800 mA.

The primary outcome is depression severity on the Hamilton Depression Rating Scale with 24 items, assessed post-ECT at 4 weeks after study initiation. Posted group means are 22.5, 12.0, and 14.1, respectively. The sponsor reports two linear-regression superiority analyses with p-values of 0.04 and < 0.001. Unadjusted recalculations compare the 700 mA and 800 mA groups with 600 mA, but the record flags an unexplained discrepancy; the sponsor’s own analysis takes precedence.

The registry lists no serious adverse events and no deaths among the reported at-risk participants in any group. The original sample-size justification, detailed analysis plan, missing-data procedures, multiplicity handling, and secondary outcome results are not reported in the registry record.