Iron Isomaltoside/Ferric Derisomaltose vs Iron Sucrose for the Treatment of Iron Deficiency Anemia (IDA): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Iron Isomaltoside/Ferric Derisomaltose vs Iron Sucrose for the Treatment of Iron Deficiency Anemia (IDA) (NCT02940886), a randomized phase 3 study. The primary outcome is Change in Hemoglobin (Hb) From Baseline to Week 8.

2. Clinical Question

The primary measure is a continuous laboratory outcome: change in hemoglobin concentration, expressed in g/dL. The posted least-squares means were 2.49 g/dL in both groups. Their 95% confidence intervals were 2.41 to 2.56 g/dL for iron isomaltoside/ferric derisomaltose and 2.38 to 2.59 g/dL for iron sucrose. The registry also reports a non-inferiority analysis of the mean difference in final values, with an estimate of 0.00 and a 95% confidence interval from -0.13 to 0.13.

The question is statistical rather than clinical advice: it concerns the estimated average hemoglobin response and the precision of the comparison recorded in the registry. The record does not report a non-inferiority margin, so the reported analysis cannot be independently interpreted against a stated margin from the registry information alone.

3. Trial Design

FERWON-IDA was a Phase III, randomized, open-label, parallel-group trial comparing intravenous iron isomaltoside/ferric derisomaltose with intravenous iron sucrose in people with iron deficiency anemia. The registry lists an enrollment of 1512. The recorded start date is 2016-11-08, the primary completion date is 2018-03-28, and the status is completed.

Randomization assigns participants to treatment groups by chance, which is intended to support a fair comparison by reducing systematic differences in group assignment. The parallel design means participants were assigned to one treatment group rather than receiving both treatments in sequence. The trial was open-label, with masking listed as none. Consequently, treatment assignment was not masked according to the registry. This may matter for outcomes that depend on judgments or reporting, although hemoglobin concentration is a laboratory measure.

The primary outcome population was the intention-to-treat population, described as all randomized subjects. The primary endpoint was change in hemoglobin from baseline to week 8. The posted endpoint means use 1009 participants in the iron isomaltoside/ferric derisomaltose group and 503 in the iron sucrose group. Enrollment and the endpoint group counts describe different stages or analysis details; the registry record does not explain the difference.

The registry lists 20 outcome measures and 32 analyses. Secondary outcomes include cardiovascular adverse events, serum phosphate, hemoglobin response thresholds, and iron-status measures. The specific secondary results discussed in the posted comparisons cover only some of those outcomes. The full protocol, treatment schedules beyond intravenous administration, eligibility details, and sample-size justification are not reported in the registry record. They should not be inferred from the trial title or from the observed group sizes.

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: Subjects who did not have a documented history of intolerance of oral iron for at least 1 month within the last 9 months had a run-in period. During the run-in period, the subject received oral iron for up to 1 month in order to document intolerance or lack of response to oral iron.Subjects were monitored for AEs indicative of iron intolerance.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intention to treat (ITT). All randomised subjects.

5. Baseline Characteristics

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

6. Primary Endpoint

The registry lists the primary outcome as Change in Hemoglobin (Hb) From Baseline to Week 8, measured over Baseline to week 8.

The primary endpoint is the change in hemoglobin concentration from baseline to week 8, measured in g/dL. Hemoglobin is a continuous measure, so the reported group summaries estimate average change rather than simply counting whether each participant crossed a response threshold. The registry describes the endpoint as an assessment of the effect of treatment on hemoglobin in subjects with iron deficiency anemia. It defines response in terms of the ability to increase hemoglobin by week 8.

A continuous endpoint retains information about the size of change across participants. By contrast, a binary response outcome classifies participants according to whether they meet a specified threshold. Such thresholds can be useful for describing a clinically interpretable event, but they discard distinctions among values on the same side of the cutoff. The trial also lists secondary measures based on hemoglobin thresholds; those do not replace the primary continuous measure.

The recorded endpoint parameter is a least-squares mean, with a 95% confidence interval. General explanation: a confidence interval describes uncertainty around an estimated quantity under the assumptions of the analysis. It is not a range guaranteed to contain the true value, and it does not show how individual participants’ hemoglobin changed.

The registry’s endpoint description gives the clinical context for seeking a hemoglobin increase, including ineffective or unusable oral iron and investigator judgment that rapid iron-store repletion was needed. However, the record does not report further details of the endpoint model or adjustment variables. The primary result therefore supports a reading of the posted group estimates and their uncertainty, not a reconstruction of unreported analytical choices.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

General: a randomized parallel-group trial compares outcomes among groups assigned to different treatments. Random assignment is intended to make the groups comparable on average at the start, although it does not ensure that every characteristic is balanced. The registry identifies randomization and a parallel design, but details of the allocation procedure are not reported in the registry record. Open-label status means masking was listed as none; the record does not describe procedures to reduce the influence of treatment knowledge on outcome assessment.

What does the primary estimate represent?

The primary outcome is continuous, and the registry reports least-squares means. In general, a least-squares mean is a model-based group mean; it can account for terms included in the statistical model. The model terms and adjustment variables are not reported in the registry record, so the posted means should not be reverse-engineered into an assumed model. The estimates are 2.49 g/dL in both groups, with group-specific 95% confidence intervals of 2.41 to 2.56 and 2.38 to 2.59.

General: a confidence interval communicates the uncertainty of an estimate under the model and analysis assumptions. Its width is affected by factors such as variability and sample size. It does not describe the distribution of individual outcomes, and it is not a probability statement about a fixed parameter after the interval has been calculated. Overlap between separate group confidence intervals is not, by itself, the formal test of a between-group difference. The registry separately reports the mean-difference analysis.

How should the non-inferiority result be read?

The registry classifies its primary comparison as non-inferiority and reports a mean difference in final values of 0.00, with a 95% confidence interval from -0.13 to 0.13. General: a non-inferiority analysis asks whether an estimate rules out a disadvantage larger than a prespecified margin. The margin is essential to that interpretation. It is not reported in the registry record, nor are the analysis model and decision details. Therefore, the posted estimate and interval can be described, but the registry information alone does not permit an independent check of the non-inferiority criterion.

The endpoint is described as change from baseline to week 8, whereas the reported non-inferiority parameter is labeled as a mean difference in final values. The record does not explain this wording or give further model details. It is important not to assume that the labels refer to identical estimands without that explanation.

What do recalculated secondary comparisons add?

For selected binary outcomes, the posted recalculations compare event proportions using unadjusted summary statistics. A risk difference expresses the difference between two event proportions; its confidence interval describes uncertainty in that difference under the calculation method. General: when an interval includes zero, the observed data do not establish a difference at conventional two-sided testing thresholds on that scale. That statement does not establish equivalence or prove that there is no effect.

The registry also reports logistic-regression odds ratios for some secondary outcomes. An odds ratio compares event odds, not event probabilities, and the interpretation depends on the model and outcome. For the hemoglobin threshold outcomes, the registry’s regression analyses and the recalculated risk-difference comparisons use different effect measures and may use different analytical approaches. They should not be expected to produce identical numerical estimates. The registry does not report the full regression specifications.

General: over-dispersion matters when count or binary outcome variability is greater than a model assumes. If it is ignored, uncertainty estimates may be too narrow. The registry entries do not establish whether over-dispersion was assessed or relevant to these analyses; no such assessment is reported in the registry record. The recalculations are simple unadjusted comparisons, not replacements for a sponsor’s model-based analysis.

How should planning illustrations be understood?

The sample-size illustrations are based on observed effects and specified planning assumptions: a two-sided significance level of 0.05, equal allocation between groups, and a normal approximation. They show participants per arm for 80% power or participants per arm for 90% power for the stated fraction of an observed effect. These are planning illustrations, not the trial’s original sample-size calculation and not observed power. They do not establish that the trial was adequately or inadequately powered for its primary question. The trial’s own sample-size assumptions and justification are not reported in the registry record.

General: power calculations describe the probability of detecting an assumed effect under stated design and model assumptions. Once outcomes have been observed, a retrospective “observed power” calculation adds little beyond the estimate and its confidence interval. The planning illustrations should instead be read as conditional examples of the sample sizes that would correspond to the stated assumptions and effect fractions.

10. Analyses Reported by the Sponsor

The registry’s own primary analysis reports a non-inferiority comparison of mean final values: the mean difference was 0.00, and its 95% confidence interval ranged from -0.13 to 0.13. The group-level least-squares mean changes were 2.49 g/dL in each treatment group. These are the sponsor-reported results recorded for the primary outcome and are the central evidence for describing the trial’s primary comparison.

The registry record does not report a separate recalculated primary comparison from summary statistics. Accordingly, there is no recalculated primary estimate to set against the sponsor’s analysis. The secondary recalculations use simple, unadjusted comparisons of posted counts and are not a substitute for the trial’s own model-based results. For secondary outcomes where the registry reports logistic regression, the odds ratio and the unadjusted risk difference answer related but not identical statistical questions.

No discrepancy is flagged between the primary results and a recalculation. More generally, when a discrepancy is present, the sponsor’s own analysis takes precedence over a simple recalculation from posted summary statistics; that situation is not indicated for this record. The registry does not provide enough detail to reproduce the primary model, verify its assumptions, or assess the non-inferiority margin. Thus, the reported estimate and interval can be accurately described, while full independent validation of the sponsor’s analysis is not possible from the registry record alone.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists multiple outcomes and analyses: 20 outcome measures and 32 analyses. It also identifies 18 posted secondary outcomes. With many outcomes or comparisons, the chance of at least one apparently unusual result can increase if each is tested separately. This is the general multiplicity problem. Adjustment procedures or a prespecified testing hierarchy can help control how confirmatory claims are made across multiple tests, but neither an adjustment plan nor a hierarchy is reported in the registry record.

Consequently, secondary p-values and comparisons should be read as individual results rather than as a family-wide confirmatory conclusion. The recalculated secondary comparisons are explicitly unadjusted. They do not account for the number of outcomes examined. A nominally small p-value, if present, would not by itself resolve this issue; the interpretation would depend on whether the outcome and analysis were prespecified and how multiplicity was handled. Those details are not reported in the registry record.

The registry identifies the primary population as intention to treat, meaning all randomized subjects. It does not report how missing week-8 measurements were handled, whether any imputation was used, or what sensitivity analyses were conducted. The group sizes for the primary means are 1009 and 503, while secondary and safety outcomes have their own recorded denominators. The record does not explain all differences among these analysis counts. Missing-data assumptions matter because estimates can be affected if outcome availability is related to participant characteristics or treatment response.

The registry also does not report whether interim analyses were planned or conducted, what stopping rules may have applied, or detailed analysis-set definitions beyond the primary intention-to-treat description. No sample-size justification or non-inferiority margin is reported. These omissions do not invalidate the posted values, but they limit assessment of how the analyses were planned, how robust they are to missing information, and how strongly secondary findings support confirmatory claims.

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 3% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: syncope (0.4%), anaphylactic reaction (0.4%), pancreatitis acute (0.2%), peptic ulcer (0.2%), urinary tract infection (0.2%).

The registry reports serious adverse events among 989 participants at risk in the iron isomaltoside/ferric derisomaltose group and 494 at risk in the iron sucrose group. Serious adverse events affected 21 and 13 participants, respectively. Other adverse events affected 31 and 16 participants in those same at-risk groups. The record lists one death in the iron isomaltoside/ferric derisomaltose group and none in the iron sucrose group.

These figures are descriptive counts, not proof that a treatment caused an event or that a difference between groups is statistically established. The registry’s top serious-event terms include syncope and anaphylactic reaction, each listed at 0.4%, and pancreatitis acute, peptic ulcer, urinary tract infection, and pyelonephritis, each listed at 0.2%. The record does not provide comparative estimates or confidence intervals for these individual terms in the registry record.

General: adverse-event counts depend on the definition, observation period, participant exposure, and completeness of reporting. A difference in counts cannot be interpreted in isolation as a causal effect. The registry record does not report further event-timing details or a full account of how adverse events were adjudicated. These figures describe the recorded safety experience and should be read alongside, not as a substitute for, the trial’s other results.

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

The primary result describes average hemoglobin change from baseline to week 8 among randomized participants included in the intention-to-treat endpoint analysis. Both groups had a posted least-squares mean change of 2.49 g/dL. The registry’s non-inferiority comparison was centered at 0.00, with a 95% confidence interval from -0.13 to 0.13. These estimates indicate that the recorded average responses were close on the reported scale, with uncertainty represented by the interval.

This does not mean the treatments are identical. A point estimate of 0.00 is an estimate, not proof of no difference, and a confidence interval allows for a range of values under the analysis assumptions. Nor does the registry information establish clinical equivalence: an equivalence or non-inferiority conclusion depends on a prespecified margin and analysis details that are not reported in the registry record. The result should not be extended beyond the population, endpoint, and follow-up described by the trial.

The primary endpoint is a group average. It does not show how many individuals experienced a particular amount of change, how consistently individuals responded, or whether every participant improved. Secondary threshold outcomes offer different summaries but do not answer those individual-level questions completely. The open-label design and incomplete reporting of analytical details are additional reasons to keep interpretation tied to what the registry actually records.

Statistical evidence is also distinct from a treatment recommendation. The reported results do not, by themselves, establish which option is appropriate for an individual. They describe the trial’s estimated hemoglobin outcomes and safety counts. Broader decisions require evidence and considerations beyond this single registry record. The most defensible conclusion is therefore narrow: the posted primary estimates are similar, and the registry reports a non-inferiority analysis, but the absent margin and methodological details prevent independent verification of the non-inferiority decision.

14. Why This Trial Matters Statistically

This trial illustrates why the effect estimate, its uncertainty, and the analysis question need to be read together. The two posted group means are both 2.49 g/dL, while the registry’s comparison reports a mean difference of 0.00 with a 95% confidence interval from -0.13 to 0.13. The interval gives more information than the point estimate alone: it shows the range of differences compatible with the analysis at the stated confidence level, subject to its assumptions.

It also demonstrates that a non-inferiority label cannot be fully evaluated without the margin. The statistical question is not simply whether a difference is statistically distinguishable from zero. Non-inferiority requires judging whether the possible disadvantage is acceptably small according to a prespecified margin. Because the margin is not reported in the registry record, the estimate can be explained but the decision criterion cannot be independently checked.

The trial further shows why secondary endpoints need disciplined interpretation. Hemoglobin thresholds, phosphate outcomes, safety events, and iron-status measures address different questions, and many outcomes were recorded. Without reported multiplicity handling, secondary signals are less confirmatory. Unadjusted recalculations can add transparent comparisons from posted counts, but they do not replace model-based sponsor analyses or remove the uncertainty created by multiple testing.

Finally, the planning illustrations are not evidence of the trial’s original design assumptions or its observed power. Their meaning depends on explicit hypothetical assumptions and effect sizes. Keeping planning calculations separate from trial results prevents a common interpretive error: treating a retrospective sample-size illustration as proof that the study could or could not detect a particular effect.

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 registry record leaves several trial-specific details unavailable. It does not report the non-inferiority margin, the full primary statistical model, adjustment variables, or the reasoning behind the sample size. Without the margin, the non-inferiority conclusion cannot be independently checked against its decision threshold. Without model specifications, the least-squares means and reported comparison cannot be fully reproduced from the registry record.

The trial was open-label, with masking listed as none. This does not automatically invalidate a laboratory endpoint, but knowledge of treatment assignment can influence participant experience, reporting, co-interventions, or other aspects of trial conduct. The record does not describe safeguards related to those possibilities. It also does not report detailed treatment schedules beyond intravenous administration in both groups.

The primary endpoint group counts are 1009 and 503, while the registry records different denominators for some secondary and safety outcomes. The reason for these differences is not reported in the registry record. Handling of missing primary outcome data, imputation, sensitivity analyses, and detailed analysis-set definitions are also not reported. These omissions limit evaluation of whether missingness or analysis-population choices could affect the estimates.

Many outcomes and analyses were recorded, but a multiplicity plan or hierarchy is not reported. As a result, secondary comparisons should be viewed cautiously and not treated as independent confirmatory tests. The registry also does not report interim analyses or stopping rules. Finally, the available results do not provide individual participant distributions, detailed subgroup results, or a full account of event timing. Those limits constrain conclusions to the posted group summaries, comparisons, and safety counts.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Iron Isomaltoside/Ferric Derisomaltose vs Iron Sucrose for the Treatment of Iron Deficiency Anemia (IDA) — NCT02940886.
• Registry record last updated 2020-10-06; status last verified 2020-01.
• Results first posted 2020-03-10.
• 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 identifies NCT02940886 as a completed Phase III randomized, open-label, parallel-group trial of intravenous iron isomaltoside/ferric derisomaltose versus intravenous iron sucrose for iron deficiency anemia. Enrollment was 1512. The recorded start date was 2016-11-08, and primary completion was 2018-03-28.

The primary endpoint was change in hemoglobin from baseline to week 8 in the intention-to-treat population. Both groups had a posted least-squares mean change of 2.49 g/dL. The registry reports a non-inferiority mean difference of 0.00, with a 95% confidence interval from -0.13 to 0.13. The non-inferiority margin and detailed model specifications are not reported in the registry record.

The record lists 20 outcome measures and 32 analyses, including secondary hemoglobin, phosphate, cardiovascular, and iron-status outcomes. Interpretation of secondary findings is limited by unreported multiplicity handling and other analysis details. Safety results include serious and other adverse-event counts and one recorded death in the iron isomaltoside/ferric derisomaltose group.