Bariatric Surgery and COVID-19: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Bariatric Surgery and COVID-19 (NCT05532891), a registered phase not stated study. The primary outcome is Revisions.

2. Clinical Question

The study asks how elective bariatric surgery in the UK during the COVID-19 pandemic compared with surgery in a pre-pandemic period. The cohorts were identified from the National Bariatric Surgical Registry, a prospectively collected database described as covering patients undergoing elective bariatric surgery in the UK. The comparison concerns patients who underwent surgery in the specified periods; it is not a randomized comparison of assigned treatments.

The record also names completed laparoscopically, length of stay, surgical complications, and presence of a second consultant as secondary outcomes. The question is relevant to service provision during the pandemic, but the comparison alone cannot isolate the effect of pandemic conditions from differences between the cohorts or their care settings.

3. Trial Design

Despite the record’s trial-style registration, the study described is a retrospective cohort analysis of registry data, not a randomized clinical trial. The National Bariatric Surgical Registry is described as prospectively collected, while the study identifies and compares patients who underwent surgery in two calendar periods. The record does not state how patients were allocated to the periods; in this design, period membership follows when surgery occurred rather than random assignment.

The pre-pandemic cohort comprises patients who underwent elective bariatric surgery during one year from 1st September 2018. The pandemic cohort comprises patients who underwent elective bariatric surgery during one year from 1st April 2020. The registry record describes the former as a control period. Calling it a control period does not make the comparison a randomized or necessarily like-for-like control comparison. The record itself flags that the arms are not a like-for-like control comparison.

The study record lists an enrollment of 11545, a start date of 2018-09-01, a primary completion date of 2021-04-01, and a completed status. Those registry entries do not, by themselves, establish how many participants contributed to each outcome or how missing observations were handled. The group counts listed for safety outcomes are specific to those safety entries and should not be treated as outcome denominators for the primary endpoint.

The primary outcome is revisions, described as the number of patients undergoing revisional bariatric surgery, with a timeframe of one year. Four secondary outcomes are named: completed laparoscopically, length of stay, surgical complications, and presence of a second consultant. The record lists six outcome measures, but results for these outcomes and the identities of any further measures are not reported in the registry record. No statistical analyses are listed.

The design can describe patterns among surgeries recorded in the two periods. Because period membership was not randomized and the record does not provide adjusted analyses, it cannot establish that the pandemic itself caused any difference between the cohorts.

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.

Population: Detailed analysis was performed for the NHS patients only. This is why the numbers of participants here differs from in overall study group, where both private and NHS records were included.

6. Primary Endpoint

The registry lists the primary outcome as Revisions, measured over 1 year.

The primary endpoint is revisions: the number of patients undergoing revisional bariatric surgery during a one-year timeframe. A revision is a further bariatric operation following an earlier bariatric procedure. Counting patients with this event gives a clinically interpretable measure of how often revisional surgery occurred in the cohorts, rather than a measure of the amount of time each patient was followed.

For a binary event outcome, the key quantities would ordinarily include the number of patients with the event and the number evaluated in each cohort. A comparison might then express how the event frequency differed between the pandemic and pre-pandemic periods, with an uncertainty interval to show the precision of that estimate. These are general statistical principles, not reported results from this study.

The registry record does not report the revision counts, denominators for this endpoint, an effect estimate, or an uncertainty interval. Therefore, the endpoint’s observed frequency and the precision of any cohort comparison cannot be assessed from the registry record. The enrollment entry is not a substitute for endpoint-specific denominators.

Revisions are only one aspect of bariatric surgery provision. The named secondary outcomes include operative approach, length of stay, surgical complications, and the presence of a second consultant. Findings for those outcomes are also not reported in the registry record. The endpoint can frame a comparison of recorded revisions across periods, but by itself it does not characterize all dimensions of surgical care.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The available information supports a limited account of the methods. It describes a retrospective cohort study using the National Bariatric Surgical Registry, which is characterized as a prospectively collected database for patients undergoing elective bariatric surgery in the UK. The study identifies patients according to when surgery occurred and compares a pre-pandemic period with a pandemic period. The registry record does not report a statistical analysis plan or any completed analyses.

What does a cohort comparison estimate?

In a cohort comparison, outcome experience is described for groups defined by an exposure or characteristic. Here, the defining characteristic is the period in which elective surgery took place. The intended contrast is between revisions among patients operated on in the pre-pandemic period and revisions among patients operated on in the pandemic period. Because the groups were not randomized, the contrast is an association between period and outcome in the recorded cohorts, not automatically a causal effect of pandemic conditions.

For a binary event such as whether a patient underwent revisional surgery, an analysis would generally need a count of patients with the event and a suitable denominator for each group. It could report event frequencies and a measure of contrast, alongside an interval expressing uncertainty. The registry record does not report these primary outcome inputs or results. No recalculated effect can therefore be obtained from the posted information.

Why does randomization matter?

Randomization aims to make comparison groups similar, on average, with respect to both measured and unmeasured characteristics at the time of assignment. This study instead compares patients from different calendar periods, and allocation is not stated as randomized. Differences between the cohorts could reflect changes in the patients receiving surgery, clinical practice, service availability, recording, or other circumstances associated with time. These are general possibilities in a period-based observational comparison, not documented explanations for this study’s results. The registry record does not report adjusted analyses that might account for measured differences.

What does an uncertainty interval tell a reader?

In general, a confidence interval describes a range of estimates compatible with the observed data and a specified statistical procedure. A narrow interval indicates greater precision than a wide interval, while neither width nor position alone resolves bias from an observational design. No confidence interval is reported for the revision endpoint, and the absence of an interval means its precision cannot be evaluated.

What can be said about the sample-size approach?

The registry record describes a planning illustration based on an observed effect, using a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. The illustration is described as targeting a stated fraction of the observed effect. These are planning assumptions, not evidence of the study’s original design or its actual statistical power. The registry record does not report the numerical planning results, and it does not report the observed effect needed to interpret such an illustration. A sample-size illustration is not observed power and cannot replace an analysis of the endpoint data.

The record does not report a sample-size justification for the study itself. It also does not report whether any interim analyses were planned or performed. These items should not be inferred from the planning illustration or from the record’s enrollment entry.

How should event counts be interpreted?

Counts are useful descriptions, but their meaning depends on what was counted, the population at risk, the observation period, and completeness of recording. A count alone is not a rate or a comparison adjusted for differences in group size. The safety entries give at-risk counts and event counts for the adverse-event record, but the record does not specify enough detail to extend those denominators to other outcomes. The primary endpoint’s denominators are not reported in the registry record.

Over-dispersion is a general concern when outcomes are more variable than a simple statistical model assumes; it can make standard errors too small if the model does not account for that extra variation. It is not possible to assess whether over-dispersion is relevant to this endpoint from the registry record, because no model, data distribution, or analysis is reported. Likewise, no model choice or diagnostic assessment is reported.

What remains unknown about the analysis?

The registry record does not report how missing observations were handled, whether analyses followed predefined populations, how covariates were selected, or whether results were adjusted. The record lists six outcome measures but does not provide all measure descriptions or analysis results. No registered analyses are listed. Without those details and the outcome results themselves, readers cannot determine whether a particular inferential method was appropriate or reproduce the primary comparison from the posted record.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Pre-pandemic", "Pandemic")
n      <- c(8615, 2930)
events <- c(746, 316)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Pre-pandemic  8615  746  8.7  8.1  9.3  8.1  9.3  8.1  9.3
ARM  Pandemic  2930  316  10.8  9.7  12.0  9.7  12.0  9.7  12.0

11. Multiplicity, Missing Data and Analysis Populations

The record lists six outcome measures, including revisions as the primary outcome and four named secondary outcomes: completed laparoscopically, length of stay, surgical complications, and presence of a second consultant. The remaining outcome measures are not reported in the registry record. No analysis results are listed, and the record gives no multiplicity strategy. It is therefore not reported in the registry record whether the study adjusted for multiple outcome comparisons, designated any outcomes as exploratory, or prioritized a subset of secondary outcomes.

As a general statistical principle, examining multiple outcomes can increase the chance of obtaining an apparently notable result by chance if each is tested separately without an appropriate plan. This does not mean a particular result is false; it means that interpretation depends on the number of comparisons and the analysis strategy. Since results and a multiplicity plan are not reported, statistical significance across outcomes cannot be assessed.

Missing data procedures are also not reported in the registry record. The record does not explain whether outcome data were complete, how missing measurements or records were handled, or whether missingness differed between periods. It also does not specify the analysis population for each outcome. The enrollment entry and adverse-event at-risk counts do not establish endpoint-specific populations or missing-data handling.

These omissions limit reproducibility and interpretation. Readers cannot determine from the registry record whether all planned outcomes were analyzed, whether analyses followed prespecified rules, or how missing records affected the comparisons.

12. 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.

Most frequent non-serious adverse event terms

The adverse-event record lists zero serious adverse events among 8615 at risk in the pre-pandemic group and zero among 2930 at risk in the pandemic group. It lists deaths as 0 in the pre-pandemic group and 1 in the pandemic group. These are the registry’s descriptive entries; they are not estimates of the primary revisions outcome.

The record also lists other affected counts of 151 for the pre-pandemic group and 38 for the pandemic group. The registry record does not provide event definitions, event-specific observation details, or an analysis comparing these counts. Counts alone do not account for differences in the number at risk or establish that the groups had comparable follow-up or event ascertainment.

Zero recorded serious events does not establish that serious events were impossible or that the true risk was zero. It describes what is entered in the registry record for the listed groups. The record does not list serious-event terms and does not report an uncertainty interval or formal safety comparison. These entries should therefore be read as descriptive safety information, not as proof of equivalent safety between periods.

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

The study addresses an important statistical question about whether the pattern of elective bariatric surgery, including revisional procedures, differed between a pre-pandemic period and a period during the COVID-19 pandemic. Its use of a prospectively collected national registry offers a way to examine recorded practice across those periods. The registry record identifies the cohorts, the primary outcome, and several secondary outcomes.

However, the posted information does not include the revision counts, denominators, or comparative estimates. It therefore does not establish the direction or magnitude of any difference in the primary endpoint, nor how precisely a difference might be estimated. The listed safety counts are descriptive entries and do not fill that gap.

Even if an outcome contrast were reported, the nonrandomized period-based design would require caution. The two groups may differ in patient characteristics or circumstances beyond the pandemic period. The registry record does not report information sufficient to assess or adjust for such differences. An association between period and outcome would not, on its own, show that the pandemic caused the difference.

The absence of statistical analyses and details about missing data, analysis populations, and multiplicity further limits what can be inferred. The record’s sample-size planning illustration is not observed power and does not indicate whether the study had adequate precision for the primary outcome. Without the outcome results and analysis details, neither a positive nor a null conclusion is supported by the registry record.

The appropriate interpretation is therefore bounded: the record describes a retrospective comparison of two surgery periods and specifies revisions as its primary endpoint, but does not report the results needed to answer the clinical question. It should not be used to claim that care improved, worsened, or remained unchanged across the periods.

14. Why This Trial Matters Statistically

Statistically, this study illustrates the distinction between a registry’s design description and the evidence needed to evaluate an outcome. The record identifies a real-world cohort comparison and a clinically meaningful endpoint, but the absence of endpoint counts, denominators, estimates, and uncertainty intervals prevents an assessment of the primary result. Naming an outcome is not the same as reporting its evidence.

The period-based design also highlights the challenge of interpreting observational comparisons. Patients treated before and during the pandemic were not assigned at random to those periods. A difference, if reported, could reflect a range of changes associated with time rather than the pandemic alone. For a useful interpretation, readers need both transparent outcome reporting and methods that address relevant cohort differences where possible.

The registry record also names several secondary outcomes and lists six measures overall, while providing no results or multiplicity strategy. This emphasizes why complete outcome reporting matters: readers need to know which comparisons were made and how they were interpreted, not just which outcomes were intended.

Finally, sample-size planning illustrations should not be confused with observed power or with evidence that an estimate is precise. Statistical interpretation depends on the observed data and the uncertainty around the estimate. In this record, those elements are not reported for the primary endpoint. The study’s question may be important for understanding surgical provision, but the posted record does not supply the analysis needed to answer it.

15. 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 central limitation for interpretation is that the registry record does not report results for the primary endpoint. There are no revision counts, endpoint-specific denominators, effect estimates, confidence intervals, or p-values. The named secondary outcomes also have no reported findings. As a result, the direction, size, and precision of any outcome differences cannot be assessed.

The study compares surgery in calendar periods rather than randomized groups. The record explicitly notes that the arms are not a like-for-like control comparison. It does not report cohort characteristics, adjustment methods, or analyses that could help assess differences between patients or care conditions across periods. Any contrast would therefore be vulnerable to confounding by factors associated with time.

Several reporting details are also absent. The registry record does not report endpoint-specific analysis populations, missing-data handling, a multiplicity approach, or a statistical analysis plan. It lists six outcome measures but does not identify all of them in the available record. No analyses are listed. These gaps make it difficult to assess completeness, reproducibility, or the appropriateness of any inferential claims.

The safety information is limited to counts and at-risk entries; event definitions and comparative safety analyses are not reported. The enrollment entry should not be interpreted as the denominator for each outcome. Finally, the planning illustration is not a substitute for reported outcome estimates or their uncertainty, and the record does not report the numerical planning results.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Bariatric Surgery and COVID-19 — NCT05532891.
• Registry record last updated 2023-12-01; status last verified 2023-11.
• Results first posted 2023-12-01.
• 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.

19. Record Summary

The registry record describes a completed retrospective cohort study using the National Bariatric Surgical Registry. It compares elective bariatric surgery in a pre-pandemic period of one year from 1st September 2018 with surgery during a pandemic period of one year from 1st April 2020. The record lists enrollment as 11545 and names revisions as the primary outcome, measured as the number of patients undergoing revisional bariatric surgery over one year.

Four secondary outcomes are named: completed laparoscopically, length of stay, surgical complications, and presence of a second consultant. The record lists six outcome measures overall, but outcome results and analyses are not reported in the registry record. Safety entries list zero serious adverse events in both groups and deaths of 0 and 1, respectively. Because this is a nonrandomized period comparison, and the primary outcome results are absent, the registry record does not establish whether revision frequency differed between periods or why.