Preventing Relapse After Successful Electroconvulsive Therapy for Depression: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Preventing Relapse After Successful Electroconvulsive Therapy for Depression (NCT06751875), a randomized na study. The primary outcome is Relapse Rate.

2. Clinical Question

The central question is about relapse prevention after a successful ECT course, rather than initial treatment response. The registry record describes a randomized, single-masked trial and identifies relapse rate as the primary outcome. It also says that depression severity was assessed using the Inventory of Depressive Symptomatology—Clinician Rated (IDS-C). The project summary describes self-rated mood alongside clinician ratings, but does not give a primary-outcome result for self-ratings in the posted results.

The question is clinically and statistically distinct from asking whether the two arms had different average depression severity or cognitive performance. Those are secondary outcomes, and they measure different aspects of follow-up. The primary comparison concerns participants who relapsed. The registry record gives arm-level event counts and a recalculated unadjusted contrast, as well as a sponsor-reported mixed-model analysis. Those analyses do not agree, so the primary finding requires particular care in interpretation.

3. Trial Design

This was a randomized, single-masked, parallel-group trial sponsored by Universiteit Antwerpen. The registry record lists the study as completed, with enrollment of 96 participants. It concerns major depressive disorder and evaluates lithium as an addition to personalized maintenance ECT after successful ECT. The arm labels are Lithium Arm and Non-Lithium Arm. The registry record does not provide further treatment details in the arm descriptions.

The primary endpoint was relapse rate, described as the percentage of patients in each arm who relapsed. The assessment period began at the start of the maintenance phase, with repeated assessments at fixed points from month 1 through month 6. The outcome description also refers to follow-up through six months. Depression severity was assessed using the IDS-C, a clinician-rated measure. The record describes a total score range of 0-84, with higher scores indicating more severe depressive symptoms. Cognitive performance was also posted as a secondary outcome, assessed at the start of maintenance and at month 1.

The registry record reports 50 participants in the Lithium Arm and 46 in the Non-Lithium Arm for the displayed primary and secondary outcomes. These are the analysis group sizes shown with the results; the registry record does not explain whether these are all randomized participants or how any exclusions or missing observations were handled. The trial's original sample-size justification, allocation procedure, masking details beyond the single-masked designation, and analysis population definitions are not reported in the registry record. The displayed planning illustrations are retrospective calculations from observed effects, not evidence of the original design rationale.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: Ninety-six patients (63.9%) entered the Maintenance Phase of the study and were randomized to either the lithium (n=50) or non-lithium (n=46) group. During the Follow-up Phase, 60 patients were followed, and 50 completed the full trial. Patients were recruited across the four sites: UPC Duffel (n=37, 38.5%), UPC KU Leuven (n=27, 28.1%), AZ Sint-Brugge (n=23, 24%) and Erasmus MC Rotterdam (n=9, 9.4%).

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.

Population: Participant numbers reflect those with available baseline data included in the analysis for each study arm.

6. Primary Endpoint

The registry lists the primary outcome as Relapse Rate, measured over Start of Maintenance Phase with repeated assessments at fixed points for six months (1 month, 2 month, 3 month, 4 month, 5 month, and 6 month) during the maintenance period through 6-month follow-up..

Relapse rate is a binary participant-level outcome: each participant is classified according to whether a relapse occurred during the stated maintenance follow-up. The registry record describes the rate as the percentage in each treatment arm who relapsed. It specifies repeated assessments at fixed monthly points through month 6, while the primary result is also presented as counts of participants with an event.

This endpoint is used because the study question concerns relapse following successful ECT. A count or rate of participants with relapse speaks directly to the occurrence of the event, whereas a symptom-severity score describes the intensity of depressive symptoms on a scale. The two outcomes are related but are not interchangeable: a difference in symptom scores does not, by itself, establish a difference in the number of people classified as having relapsed.

The registry record additionally describes IDS-C as a measure of depression severity. It is a 30-item clinician-rated scale, with one of two appetite or body-weight disturbance items selected for scoring. The resulting 28-item score ranges from 0-84, and higher scores indicate more severe depression. This helps distinguish the primary event outcome from a secondary measure of symptom burden. The record does not provide a detailed operational definition of relapse, such as the precise threshold or adjudication procedure, so those details are not reported in the registry record.

7. Results Posted in the Registry

• Non-Lithium Arm: 19.6% (9/46) versus 14.0% (7/50) on control; risk difference 5.6 percentage points (95% CI -9.4 to 20.5), risk ratio 1.40 (0.57–3.45); unadjusted two-proportion z-test p = 0.465.

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.

Depression Severity Differences

• 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:
  • Non-Lithium Arm: mean 41.30 versus 42.70 on control; difference -1.40 (95% CI -4.94 to 2.14), standardized effect d = -0.16; z-test p = 0.439.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Non-Lithium Arm and the control. The interval gives the range of effects that remain compatible with the data.

Cognitive Performance

• 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:
  • Non-Lithium Arm: mean 23.10 versus 23.80 on control; difference -0.70 (95% CI -2.05 to 0.65), standardized effect d = -0.21; z-test p = 0.308.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Non-Lithium Arm and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes cover symptom severity and cognitive performance, so they add context about dimensions other than the primary relapse event. Their posted results do not produce a single, consistent pattern that can be treated as a direct confirmation of the primary analysis. In particular, a symptom score and a cognitive score do not measure relapse itself, and their numerical contrasts cannot resolve the disagreement between the sponsor's primary mixed model and the event-count recalculation.

The tables show unadjusted comparisons based on arm-level summaries, alongside sponsor-reported analyses using mixed models for depression severity and ANOVA for cognitive performance. These analyses are not all testing the same contrast in the same way. For depression severity, the sponsor entries are reported separately by arm rather than as a direct between-arm estimate. For cognitive performance, the sponsor reports a p-value but not an estimate or interval. That limits direct comparison between sponsor analyses and recalculations.

Secondary results generally carry less weight than the primary result when they are not the main question used to organize the trial, especially when several outcomes or analyses are considered. Each additional analysis creates another opportunity for a nominally notable result to occur by chance. The registry record does not report a multiplicity adjustment or a prespecified hierarchy for these outcomes. The displayed p-values should therefore be considered exploratory unless a fuller analysis plan establishes otherwise. This does not make the outcomes uninformative; it means they should be interpreted as supporting context rather than independent confirmation.

Taken together, the secondary outcomes broaden the description of follow-up but do not settle the primary statistical discrepancy. The posted tables provide effect estimates and intervals for the recalculations, while the sponsor's reported results are incomplete for assessing magnitude and precision. Interpretation should focus on what each endpoint measures and on the limitations of its analysis, rather than treating every p-value as an equivalent test of the same clinical question.

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

• Depression Severity Differences
• Cognitive Performance

9. Statistical Interpretation

10. Statistical Methods Explained

Several kinds of statistical results appear in the registry record, and they answer different questions. The primary outcome is binary: whether a participant experienced relapse. The table's arm counts summarize observed events, while a risk difference expresses the contrast between event proportions in the two arms. The posted recalculation is an unadjusted comparison based on those summary counts. It is a simple description of the displayed data, not a substitute for a model that may account for repeated observations or other features of the analysis.

A confidence interval describes a range of estimates compatible with the data and the method used to construct the interval. Its width is informative about precision: a broad interval indicates that the estimated contrast is uncertain. When a risk-difference interval includes zero, the data do not rule out no difference under that interval calculation. This is not equivalent to proving that the arms are the same. The primary recalculated interval extends from -0.094 to 0.205, so it includes zero and also allows for contrasts in either direction.

A p-value summarizes how unusual the observed data, or data more extreme, would be under a specified null model. It does not measure the probability that the null hypothesis is true, the size of an effect, or the importance of a result. A p-value also depends on the analysis method and assumptions. The unadjusted p-value of 0.465 belongs to the recalculation from the displayed event counts. The sponsor-reported mixed-model p-value of 0.014 comes from a different analysis. The registry record does not provide its estimate, interval, or sufficient model details to establish exactly why the two results differ.

General: why a mixed model may be used. A mixed model can represent correlated observations, such as repeated measurements from the same participant, rather than treating each measurement as independent. It can also accommodate specified fixed and random effects. These are general properties of mixed models, not a description of the sponsor's exact specification. The record names a mixed-model analysis but does not report the outcome scale, link function, covariates, repeated-measures structure, random-effects structure, or missing-data assumptions. It would therefore be inappropriate to infer what adjustments were made or how those choices produced the sponsor's result.

The secondary depression-severity outcome is continuous and reported as a mean with a standard deviation in each arm. Its recalculated mean difference is an unadjusted contrast between arm summaries. The table also gives a confidence interval and a p-value for that recalculation. The sponsor's reported analyses for this outcome instead list separate mixed-model p-values for each arm. Those within-arm entries are not a direct estimate of the between-arm contrast. Their interpretation depends on the model and null hypothesis, which are not described in enough detail in the registry record.

The cognitive-performance outcome is also summarized as a mean and standard deviation, with an unadjusted mean difference calculated from the displayed group summaries. The sponsor reports an ANOVA result for the two arms. ANOVA is a method for comparing means under its model assumptions, but the record does not post an estimate or confidence interval for that sponsor analysis. The p-value alone cannot show the magnitude or precision of the contrast.

General: why over-dispersion can matter. For binary outcomes, a simple model may assume that variability follows the binomial distribution implied by the event probabilities and sample sizes. If observations vary more than that model expects, standard errors can be too small and intervals too narrow unless the analysis accounts for the extra variability. This is a general consideration, not a claim that over-dispersion occurred. The registry record does not report a diagnostic or adjustment for over-dispersion.

The recalculated comparisons are labelled unadjusted and should be kept separate from the sponsor's analyses. They use posted summary statistics and do not reconstruct participant-level data or the sponsor's model. No conclusion about which analysis is more appropriate can be reached from the posted information alone. The registry record does not report the full statistical analysis plan, the analysis population, model diagnostics, or details of how repeated outcome measurements entered the primary analysis.

11. Confidence Intervals and Precision

The primary recalculated risk difference is 0.056, with a 95% confidence interval from -0.094 to 0.205. The interval includes zero, so the unadjusted summary data are compatible with no difference as well as with differences in either direction. Its width conveys substantial uncertainty about the size and direction of the contrast. The event counts are modest, and the interval is more informative about precision than a p-value alone.

The sponsor's primary mixed-model result is reported as p = 0.014, but the registry record gives no effect estimate or confidence interval for that analysis. A p-value cannot indicate how large the estimated contrast was or how precisely it was estimated. The missing estimate and interval make it difficult to assess the practical scale of the sponsor-reported result or to compare it directly with the event-count recalculation.

The secondary recalculations also have intervals that cross zero: the depression-severity mean difference is -1.4, with a 95% confidence interval from -4.942 to 2.142, and the cognitive-performance mean difference is -0.7, with a 95% confidence interval from -2.045 to 0.645. These intervals do not establish equivalence. They indicate that the posted summary comparisons do not narrowly pin down the between-arm differences. The sponsor analyses for these outcomes do not include estimates or intervals in the registry record.

Confidence intervals describe statistical uncertainty under the analysis used; they do not account for every source of uncertainty. Here, differences between the sponsor analysis and recalculation, unreported model details, and unreported missing-data handling are additional limits on interpretation. The posted estimates should therefore be read with attention both to interval width and to how each analysis was produced.

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 sample-size table and accompanying curve are planning illustrations based on observed effects, not the trial's original design calculations. For the relapse outcome, the table gives participants per arm. The continuous secondary-outcome tables also give participants per arm. These are not total-event counts. The registry record's planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; each row uses the stated fraction of the observed effect as the effect to detect.

The 80% power column describes the number of participants per arm estimated to give an 80% chance of detecting the specified effect under those planning assumptions, if that effect were the true effect and the assumptions held. The 90% power column describes the corresponding planning target for a 90% chance. These are design targets, not probabilities that the observed results are correct. This is not observed power.

Each row changes the effect used for planning. Planning for 100% of the observed effect uses the full observed estimate; the other rows use smaller fractions of it. Smaller effects are harder to distinguish from random variation and therefore require more participants per arm for the same target power. That relationship is what the table and curve are intended to make visible. The registry record does not report the trial's own sample-size assumptions or justification, so these retrospective illustrations should not be read as evidence about the investigators' original target or planning process.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.140; treatment proportion p₁ = 0.196; average p̄ = 0.168.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5285 = 1.0359.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5270 = 0.4436.
4. n = (1.0359 + 0.4436)² / (0.0557)² = 706.7, rounded up to 707 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
p0 <- 0.14                          # control proportion
p1 <- 0.195652173913043                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(6751875)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 707 (80% power), 946 (90% power)
 75% of observed effect: n per arm = 1,216 (80% power), 1,627 (90% power)
 50% of observed effect: n per arm = 2,639 (80% power), 3,533 (90% power)
power.prop.test: 706.7 
Simulated power at n = 707 per arm: 0.804 (20000 replicates)

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

14. Analyses Reported by the Sponsor

The registry record reports the sponsor's primary analysis as a mixed-model analysis with p = 0.014, classified as a superiority analysis. The record does not include the sponsor's effect estimate or confidence interval for this result. The recalculation on this page uses the posted event counts and is an unadjusted comparison: it gives a risk difference of 0.056, a 95% confidence interval from -0.094 to 0.205, and p = 0.465. These are not the same analysis, and their results differ.

The record does not contain enough information to reconcile the discrepancy. It does not state the mixed model's specification, the exact estimand, or how the repeated assessments were used in the primary result. The simple recalculation cannot reproduce a model based on participant-level or repeated-measures information. Conversely, the sponsor's p-value without an estimate or interval does not show the magnitude or precision of its result.

Because the discrepancy is flagged, the sponsor's own analysis takes precedence over the recalculation. This means the sponsor result should be reported as the trial's analysis, while the recalculation should be presented only as a check based on posted summary counts. The discrepancy remains important context: the registry record does not provide the details needed to understand the difference or independently assess the sponsor model.

15. Multiplicity, Missing Data and Analysis Populations

The registry record lists a primary endpoint and two secondary outcomes, and it includes multiple reported analyses. When several outcomes, time points, or statistical tests are considered, the chance of obtaining at least one apparently unusual result can increase. A multiplicity strategy may address this by specifying which comparisons are primary, adjusting significance thresholds, or defining a hierarchy among tests. The record does not report a multiplicity adjustment or a testing hierarchy, so the interpretation of the reported p-values should not assume that such a procedure was used.

Missing-data handling is also not reported in the registry record. The outcome descriptions give assessment schedules, but the posted record does not explain how incomplete follow-up, missed assessments, or withdrawals were handled. This matters particularly for repeated measures, because the number and pattern of available observations can affect a model-based estimate. The record does not state whether the sponsor analysis used all available observations, imputed missing values, or relied on another approach. No particular method should be inferred.

The analysis populations are not reported in detail. Although the displayed arm sizes are 50 and 46, the registry record does not specify whether the analyses followed an intention-to-treat principle, used a per-protocol population, or applied other exclusions. It also does not explain how participants were assigned to the displayed outcome groups if follow-up data were incomplete.

These omissions do not show that the analyses were conducted improperly. They limit what can be independently assessed from the registry record. The full statistical analysis plan, analysis-population definitions, multiplicity approach, model specification, and missing-data procedures are not reported in the registry record. Those details would be needed to understand the sponsor-reported analyses and their relationship to the unadjusted recalculations.

16. 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 serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: cardial arrythmia, needing hospitalisation (2.2%), suicide attempt during study follow-up (2.2%).

The registry record lists serious adverse events for each arm and reports no deaths. It records zero participants with a serious adverse event in the Lithium Arm and 2 in the Non-Lithium Arm. The record lists other adverse events for 14 participants in the Lithium Arm and 13 in the Non-Lithium Arm. These are descriptive counts as posted; they do not establish that an event was caused by a treatment.

The record's top serious-event terms are “cardial arrythmia, needing hospitalisation” and “suicide attempt during study follow-up,” each shown as 2.2%. The registry entry does not provide enough detail to assign those terms to a particular arm, describe their timing, or assess relatedness. The event-term percentages should be read as the registry reports them, not as comparative risk estimates.

Safety counts depend on the way events were collected, defined, and classified, and the record does not describe those procedures in detail. With the available information, the figures summarize reported events during the study, but they do not support a broader comparison of safety profiles or causal conclusions.

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

The posted results describe a randomized comparison of lithium added to maintenance ECT versus an arm without lithium, among people who had responded to ECT for major depressive disorder. The primary endpoint concerns relapse during follow-up. The arm-level counts and the simple unadjusted recalculation are one way to summarize the observed data, while the registry also reports a separate sponsor mixed-model result.

The unadjusted risk-difference interval includes zero and is broad. The sponsor reports p = 0.014 for a mixed-model superiority analysis but does not post the estimate or confidence interval. Since these results differ and the model details are not available in the registry record, neither the recalculation alone nor the sponsor p-value alone gives a complete account of the primary evidence. The sponsor analysis takes precedence over the recalculation, but the posted information leaves uncertainty about the magnitude and precision of the sponsor's finding.

The secondary outcomes describe symptom severity and cognitive performance, not the same event as relapse. Their results add context, but they do not settle the primary discrepancy. The registry's limited reporting of model specification, missing-data handling, analysis populations, and multiplicity further constrains what can be concluded from the posted results.

These data do not establish that the treatment strategy works or does not work. They show what the registry reports and how precisely the available summaries estimate the contrasts. The results also do not determine what would happen for an individual participant, or establish the balance of benefits and harms beyond the events recorded. Interpretation should remain tied to this trial, its reported outcomes, and the limits of the registry record.

18. Why This Trial Matters Statistically

This trial raises a statistical question that is easy to miss when attention focuses only on a p-value: how should readers interpret different analyses of the same primary outcome? The registry record gives a sponsor-reported mixed-model p-value and enough event-count information for a simple unadjusted recalculation. The results differ, and the sponsor's analysis has no posted estimate or confidence interval. This makes transparency about estimands, models, and analysis populations central to evaluating the evidence.

The design also involves repeated assessments during maintenance follow-up. Repeated data can be useful, but the choice of model and the way incomplete observations are handled can affect the answer. The registry record names a mixed-model analysis without describing its specification. That gap illustrates why reporting the model's effect estimate and interval, alongside a clear account of its assumptions, matters as much as reporting a p-value.

The secondary outcomes add another statistical issue: a trial may report several endpoints and analyses, but each result does not carry equal evidential weight. Clear designation of primary and secondary outcomes, and a stated multiplicity plan, help readers understand how to interpret multiple tests. Here, those details are not reported in the registry record.

Finally, the planning illustrations make an important distinction between the original design and retrospective planning based on observed effects. Their figures describe hypothetical participant requirements under stated assumptions; they do not reveal the trial's original rationale. Together, these features make the record useful for learning how effect estimates, intervals, model-based analyses, and planning calculations answer different questions.

19. Limitations

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

Several limitations are specific to what is posted for this trial. Most importantly, the primary sponsor analysis and the unadjusted recalculation do not agree. The registry record does not report enough detail about the mixed model to explain the difference, and it gives no sponsor effect estimate or confidence interval. The recalculation uses arm-level summary counts and cannot reproduce an analysis based on participant-level repeated observations.

The relapse definition is not fully described in the posted outcome text. It identifies relapse as the event of interest but does not specify a threshold, adjudication process, or detailed rule for classifying a participant. The record also does not explain the analysis population or how missing assessments and incomplete follow-up were handled. These omissions make it difficult to assess the impact of outcome classification and missing observations.

The secondary outcomes have their own reporting limitations. The sponsor's depression-severity results are presented as separate arm-level mixed-model p-values rather than a direct between-arm estimate, and the cognitive-performance analysis has no sponsor estimate or interval. The registry record does not report a multiplicity strategy, so the role of multiple tests in interpreting secondary findings is unclear.

Finally, the arm descriptions do not detail the non-lithium regimen or other treatment procedures, and the original sample-size justification is not reported. The planning table is based on the observed effects and specified planning assumptions; it does not fill in the trial's original design rationale. These limits are about the completeness of the registry record and should not be mistaken for evidence that a particular analysis or procedure was absent.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Preventing Relapse After Successful Electroconvulsive Therapy for Depression — NCT06751875.
• Registry record last updated 2026-05-05; status last verified 2026-04.
• Results first posted 2026-05-05.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

The registry record describes a completed, randomized, single-masked, parallel-group trial sponsored by Universiteit Antwerpen, with enrollment of 96 participants. It evaluates lithium added to personalized maintenance ECT after successful ECT for major depressive disorder. The primary outcome is relapse rate during maintenance follow-up through six months; depression severity and cognitive performance are posted as secondary outcomes.

The primary event counts are 7 in the Lithium Arm and 9 in the Non-Lithium Arm. The registry also reports a mixed-model primary analysis with p = 0.014, while the unadjusted recalculation from the posted counts gives p = 0.465 and a broad interval. The discrepancy is important: the sponsor's own analysis takes precedence, but its estimate and interval are not reported. The record also provides secondary outcome summaries and adverse-event counts, with no deaths listed. Details of the sponsor model, multiplicity approach, missing-data handling, analysis populations, and original sample-size justification are not reported in the registry record.