Early Diuresis Following Colorectal Surgery: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Early Diuresis Following Colorectal Surgery (NCT02351934), a randomized phase 4 study. The primary outcome is Length of Hospital Stay.

2. Clinical Question

The study asked whether adding furosemide to an Enhanced Recovery After Surgery (ERAS) program after colon surgery could safely shorten the length of hospital stay. The registry describes a prospective, randomized study in participants with colorectal disorders and identifies the primary analysis population as intent to treat.

This is a question about the added intervention within the ERAS care described in the record, not a comparison of ERAS with no enhanced recovery care. The registered objective refers to both safety and shortening hospital stay. The available outcome figures can describe the observed hospital-stay distributions and selected recorded events, but they do not by themselves settle every aspect of safety or identify which parts of the care pathway account for the observed outcomes.

3. Trial Design

The registry identifies this as a prospective, randomized, parallel-group Phase 4 study with no masking. It is listed as completed. The sponsor was Mayo Clinic. The recorded start was 2015-02, and primary completion was 2016-04. Enrollment was 123.

The two groups were furosemide plus ERAS, with 62 participants, and ERAS alone, with 61 participants. The furosemide regimen was 10 mg intravenously on post-operative day #1 and/or 2. The ERAS description includes celecoxib and gabapentin before surgery, single-injection intrathecal analgesia immediately before induction of general anesthesia, and scheduled acetaminophen and a nonsteroidal anti-inflammatory drug after surgery. Other details of the ERAS regimen are not reported in the registry record.

Randomization supports a between-group comparison, but the record does not state the sequence-generation method, allocation concealment procedures, or whether outcome assessors were masked. Since masking is listed as none, participants and care teams were not masked according to the registry entry; any separate masking of outcome assessment is not reported in the registry record. The record does not describe a sample-size justification or the planned statistical analysis in enough detail to reconstruct the protocol.

The primary outcome was length of hospital stay in hours, summarized by medians and inter-quartile ranges. Its timeframe was up to 7 days, and the stated analysis population was intent to treat. Five secondary outcomes are named: readmission within 30 days, nasogastric tube placement, time to stool output, acute kidney injury, and hypokalemia. Numerical results are available in the record for only some of these outcomes; results for time to stool output and hypokalemia 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.

Allocation

Reasons for not completing this period:

Follow-Up

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent to treat analysis

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: Intention to treat analysis

6. Primary Endpoint

The registry lists the primary outcome as Length of Hospital Stay, measured over Up to 7 days.

Length of hospital stay measures the time from the relevant admission or operation period until discharge, expressed in hours in this trial. The registry describes follow-up through the hospital stay, up to 7 days, and states an expected average duration of 2-7 days. The reported summary is the median with an inter-quartile range rather than a mean and standard deviation.

The median is a useful description when an outcome may be asymmetric or contain unusually long stays: it identifies the central point of the observed distribution without being pulled as strongly by extreme values as a mean can be. The inter-quartile range describes the spread of the middle portion of observations. These summaries help readers understand both a typical observed stay and the variability around it, but they do not show every participant's experience.

Length of stay is relevant to the registered objective because the study sought to determine whether adding a diuretic after colon surgery would shorten hospitalization. It is also influenced by the course of recovery and discharge decisions, so it is not a direct measure of one biological mechanism. The registry does not provide a confidence interval for the between-group difference in length of stay. The displayed medians and ranges therefore describe each group, while the reported Wilcoxon (Mann-Whitney) p-value addresses a group comparison without giving a direct estimate and precision interval for the difference in hours.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a time-like measure reported in hours, with a median and inter-quartile range for each randomized group. The sponsor's registered comparison uses the Wilcoxon (Mann-Whitney) test. This is a rank-based method: rather than requiring the outcome to follow a particular bell-shaped distribution, it orders observations and evaluates whether values in one group tend to be higher or lower than values in the other. Its interpretation concerns the distributions of observed values. It is not automatically a test of medians alone unless additional assumptions about the shapes of the distributions are appropriate.

What does the reported p-value tell readers?

A p-value measures how unusual a test statistic at least as extreme as the observed one would be under a specified null model and the assumptions of the test. It is not the probability that the null hypothesis is true, and it does not measure the probability that the result will replicate. The primary p-value is 0.564. This result does not provide strong evidence against the null comparison, but absence of a small p-value does not establish equivalence or rule out effects that the study may estimate imprecisely.

For the primary outcome, the registry reports group medians and inter-quartile ranges but not a confidence interval for a direct between-group effect. A confidence interval, in general, gives a range of effect values compatible with the data and the specified method, reflecting statistical uncertainty. Without such an interval for length of stay, readers cannot use the registry entry to judge the precision of a direct estimate in hours. The overlap of group ranges is descriptive; it is not itself a formal test of the difference.

How should binary outcome comparisons be read?

Readmission, nasogastric tube placement, and acute kidney injury are presented as counts of participants. A risk difference expresses the difference between the proportions experiencing an outcome in the compared groups. The recalculated comparisons include confidence intervals for these differences. In general, an interval spanning no difference indicates that the data are compatible with no difference as well as with effects on either side, subject to the method used. These are simple unadjusted recalculations from the posted summary counts, not replacements for the sponsor's own analyses.

The recalculations also include planning illustrations based on stated fractions of the observed effect. These are not the number of participants enrolled, and they are not observed power. The planning assumptions use a two-sided significance level of 0.05, equal allocation, and a normal approximation. The illustrations show how required participants per arm can change with the effect size chosen for planning. Since the effect is taken from observed trial data, it is uncertain and may not be a reliable target for a future design. The registry's own sample-size justification is not reported in the registry record.

Why can methods yield different p-values?

The sponsor reports Wilcoxon (Mann-Whitney) tests for the primary and listed secondary comparisons. The secondary recalculations use binary summary counts to obtain unadjusted comparisons of risks. These approaches do not necessarily test the same statistical quantity or use the same assumptions, and their p-values need not match. The recalculated results should therefore be treated as an independent, unadjusted description rather than a correction to the sponsor's analysis. The record does not give enough detail to reproduce the sponsor's exact analysis procedures beyond naming the test.

Over-dispersion is a general concern when count outcomes vary more than a model's assumed distribution allows. It can make standard errors too small if the model fails to account for the extra variability. The primary endpoint is reported as a continuous duration with a rank-based comparison, and the posted binary secondary recalculations are based on participant counts; the registry does not report a count regression model or an over-dispersion assessment. This general issue should not be attributed to an analysis that the record does not describe.

What is and is not documented?

The primary analysis population is labeled intent to treat, which generally means participants are analyzed according to assigned group, regardless of subsequent adherence. The registry does not describe how missing primary outcomes, withdrawals, protocol deviations, or incomplete follow-up were handled. It also does not report whether an interim analysis was planned or conducted, how the randomization sequence was generated, or whether multiplicity adjustments were used. No particular procedures should be assumed in their absence.

For outcomes with sparse counts, estimates can be unstable and confidence intervals broad. A small observed count does not imply precise knowledge of the underlying risk. For binary comparisons, the interval is often more informative about uncertainty than a threshold-based reading of a p-value. The registry's limited detail on the sponsor's methods and missing-data handling means the reported results can be described, but not fully independently reconstructed.

10. Analyses Reported by the Sponsor

The sponsor's registered primary analysis compares length of hospital stay using a Wilcoxon (Mann-Whitney) test and reports p = 0.564. The registry gives medians and inter-quartile ranges for each arm, but does not report a sponsor-estimated difference in hours or a confidence interval for that difference. The unadjusted recalculation framework for binary outcomes does not generate a replacement primary analysis, because the primary endpoint is not binary.

For the secondary outcomes with numerical comparisons, the sponsor reports p-values of 0.675 for readmission, 0.125 for nasogastric tube placement, and 0.595 for acute kidney injury. The simple recalculations from the posted binary counts yield unadjusted p-values of 0.515, 0.054, and 0.987, respectively. These results differ because the recalculations and sponsor analyses do not necessarily use the same statistical method or target the same quantity. The registry identifies the sponsor method as Wilcoxon (Mann-Whitney), while the recalculations compare binary risks from summary counts.

No discrepancy is flagged for the record. The sponsor's analysis remains the registered analysis, and recalculations are supplementary views of the posted summary data, not evidence that the sponsor's analysis should be replaced. The registry does not provide enough methodological detail to reproduce the sponsor's computations fully, so the comparison should be read as a difference in reported approaches rather than a demonstrated error.

11. Multiplicity, Missing Data and Analysis Populations

The record lists one primary endpoint and several secondary outcomes. It does not report whether statistical testing of secondary outcomes was controlled for multiplicity, whether the outcomes followed a prespecified testing hierarchy, or whether any outcomes were exploratory. When multiple hypotheses are tested, the chance of obtaining at least one apparently unusual result by chance can increase. Multiplicity procedures can help manage that issue, but no specific correction or testing sequence should be assumed for this study.

The primary population is described as intent to treat. The record does not explain how that population was operationalized, whether all randomized participants contributed to every outcome, or how missing measurements were handled. It also does not describe imputation, exclusions, censoring, or sensitivity analyses for missing data. Those details matter because incomplete outcome information can affect the groups' comparability and the uncertainty of estimates. They are not reported in the registry record.

The posted denominators for the displayed outcomes are 62 and 61, corresponding to the randomized group sizes, but that alone does not establish that every participant had complete follow-up for each endpoint. The specific missingness pattern is not reported in the registry record. Readmission is defined within 30 days of release from hospital, while other listed outcomes have their own stated timeframes; details about ascertainment and incomplete follow-up beyond those descriptions are not reported.

The registry also does not report interim analyses, stopping rules, protocol deviations, or the method used to generate and conceal the random allocation. These omissions do not establish that such procedures were absent; they mean the record does not document them. Without the full statistical plan and missing-data details, interpretation should remain focused on the posted summaries and their stated analyses, without claims about unreported procedures.

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 registry reports zero serious adverse events among 62 participants in the furosemide-plus-ERAS arm and zero among 61 participants in the ERAS arm. It also reports zero deaths in each group. No serious adverse event terms are listed. These figures mean that no serious events or deaths are recorded in those categories in the posted results; they do not establish that the risk of a serious event is absent.

Other adverse events are listed for 8 participants in the furosemide-plus-ERAS arm and 3 participants in the ERAS arm. The registry record does not identify those events in the registry record for this analysis, so their nature, severity, timing, and relation to treatment cannot be assessed. Counts alone do not show whether a difference is due to chance, and they do not describe event rates beyond the observed participants.

Safety conclusions are constrained by the size of the study and the level of detail in the registry record. In particular, zero observed serious events does not rule out uncommon harms, and the record does not provide a detailed adverse-event narrative or a formal comparative safety estimate. These are descriptive figures, not a complete evaluation of safety.

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

The results describe what was observed among participants assigned to furosemide plus ERAS and ERAS alone. The furosemide arm had a median hospital stay of 99.6 hours, compared with 80.6 hours in the ERAS arm. The sponsor's primary Wilcoxon (Mann-Whitney) comparison reported p = 0.564. Together, the summaries do not provide strong statistical evidence of a group difference in length of stay, and the observed medians do not indicate a shorter stay in the furosemide arm.

This does not show that the two approaches have identical effects. A non-small p-value is not an equivalence test, and the registry does not report a confidence interval for a direct difference in hours. Without that interval, the precision of the estimated group contrast cannot be judged from the record. The summary also cannot identify whether particular participant characteristics or aspects of postoperative care influenced hospital discharge.

The secondary results and adverse-event counts provide additional context, but they do not prove benefit or harm. Some registered secondary outcomes lack numerical results in the registry record, and methods for handling multiplicity and missing information are not reported. Zero serious adverse events in the posted figures should not be interpreted as proof that serious events cannot occur.

The study's scope is the addition of furosemide to the ERAS care described in the registry. It does not compare the ERAS program with no ERAS care. The findings are an account of this study's enrolled participants and reported analyses, not a universal prediction for every colorectal surgery setting. The data do not support medical advice or a definitive claim that furosemide does or does not work.

14. Why This Trial Matters Statistically

This trial illustrates why an observed group difference, a test result, and the precision of an effect estimate are distinct pieces of evidence. The medians describe the centers of the two hospital-stay distributions; the inter-quartile ranges describe their spread; and the Wilcoxon (Mann-Whitney) p-value summarizes a rank-based comparison. None of these alone supplies a direct estimate with a confidence interval in hours. Keeping those quantities separate prevents a p-value from being mistaken for the size or certainty of a clinical effect.

It also illustrates the importance of matching the method to the endpoint. Length of stay is reported as a duration and analyzed with a rank-based test, while secondary outcomes include counts of participants with events. Recalculating binary comparisons from posted counts can provide a different view, but it is not a reproduction of a rank-based sponsor analysis. Differences between methods should be explained rather than treated automatically as errors.

The study also highlights limitations of registry-level interpretation. The record identifies randomization, the arms, and the principal outcome, yet does not fully document allocation procedures, missing-data handling, multiplicity, or a sample-size justification. Readers can interpret the posted data while recognizing what cannot be established from the registry alone. Planning illustrations based on observed effects are useful for exploring design assumptions, but are not evidence that the completed trial had a particular power or that a future study will reproduce those effects.

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 primary outcome is summarized by group medians and inter-quartile ranges, but the registry does not report a confidence interval for the between-group effect in hours. This limits assessment of the estimate's precision. The registered p-value gives a test result but does not replace an effect estimate or show which differences remain compatible with the data.

The registry labels the primary analysis intent to treat, but does not explain missing-data handling, exclusions, or outcome-specific analysis populations. It does not report a detailed protocol, statistical analysis plan, sample-size justification, interim analyses, or multiplicity procedures. The randomization method and allocation concealment are not described. Masking is listed as none, and any independent masking of outcome assessors is not reported in the registry record.

The summaries available for secondary outcomes are incomplete: numerical results for time to stool output and hypokalemia are not reported in the registry record. For the other posted secondary comparisons, sponsor tests and unadjusted recalculations use different approaches. Their estimates and p-values should not be treated as if they were generated by a single prespecified analysis.

Safety information is also limited. The record lists serious-event and death counts but does not give serious-event terms or details of the other adverse events in the registry record for this analysis. The study enrolled 123 participants, which limits what can be inferred about uncommon events. Lastly, the observed results pertain to the two care strategies described in this study; the registry record does not establish how they generalize to other settings or variations in postoperative care.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Early Diuresis Following Colorectal Surgery — NCT02351934.
• Registry record last updated 2018-05-07; status last verified 2018-04.
• Results first posted 2017-06-27.
• 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

NCT02351934, titled “Early Diuresis Following Colorectal Surgery,” is recorded as a completed, prospective, randomized, parallel-group Phase 4 study with no masking. Mayo Clinic is listed as sponsor. Enrollment was 123, with 62 participants assigned to furosemide plus ERAS and 61 to ERAS alone.

The primary endpoint was length of hospital stay in hours, summarized by median and inter-quartile range, with follow-up up to 7 days. The sponsor reported a Wilcoxon (Mann-Whitney) p-value of 0.564. The registry also includes selected secondary outcomes and safety counts, while some registered secondary results and important methodological details are not reported in the registry record.