Ciprofloxacin Versus Levofloxacin in Stem Cell Transplant: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Ciprofloxacin Versus Levofloxacin in Stem Cell Transplant (NCT03850379), a randomized phase 2 study. The primary outcome is Proportion of Bloodstream Bacterial Infections.

2. Clinical Question

The registry describes both drugs as agents used for prophylaxis in autologous and allogeneic stem cell transplant settings. It notes that some transplant centers had replaced ciprofloxacin with levofloxacin and that small retrospective studies had suggested a higher rate of breakthrough gram-positive bloodstream infections with ciprofloxacin. It also states that the optimum oral antibacterial prophylaxis agent remains uncertain. Those statements provide the rationale recorded for the question; they do not establish which agent is preferable.

The outcome is a comparison of participants with a bloodstream bacterial infection during the specified follow-up period. It does not, on its own, distinguish infection types, describe severity, or indicate whether an infection was caused by a particular organism. Those details are not reported in the registry record. The result addresses the recorded comparison of the two prophylactic regimens, not every possible outcome of transplant care.

3. Trial Design

This was a randomized, parallel-group phase 2 trial. The registry lists 308 enrolled participants and describes two groups: levofloxacin, with 154 participants, and ciprofloxacin, with 154 participants. The record identifies no masking; participants and study personnel were therefore not masked according to the registry entry. Further details about how the random sequence was generated, whether allocation was concealed, or how treatment assignments were implemented are not reported in the registry record.

The levofloxacin regimen was 500 mg once daily, and the ciprofloxacin regimen was 500 mg BID. The registry does not report treatment duration, adherence, treatment changes, or whether participants received other antibacterial agents. It identifies the sponsor as Henry Ford Health System. The recorded start date is 2018-03-14, the primary completion date is 2024-08-30, and the overall status is completed.

The primary outcome was the proportion of participants with bloodstream bacterial infections up to 60 days after stem cell transplant. The registry reports the outcome as participant counts, with 18 in each group. It does not give a more detailed account of eligibility, baseline characteristics, transplant types, or the analysis population. Although the registry labels this a randomized trial, it does not report the randomization procedure or a full statistical analysis plan. The enrollment number describes the trial record; it should not be mistaken for confirmation that every enrolled participant contributed complete outcome data.

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.

6. Primary Endpoint

The registry lists the primary outcome as Proportion of Bloodstream Bacterial Infections, measured over 60 days post stem cell transplant.

The primary endpoint concerns the proportion of participants who experienced a bloodstream bacterial infection by 60 days after stem cell transplant. The registry describes the outcome as a count of participants with an event and gives the follow-up window as 60 days post transplant. The posted results list 18 participants with the outcome in the levofloxacin group and 18 in the ciprofloxacin group; each group is listed with 154 participants.

This endpoint is relevant to the recorded question because the trial rationale concerns breakthrough bloodstream infections during antibacterial prophylaxis. Counting participants with an event gives a direct comparison of how often the defined outcome was recorded in each group over the stated period. It does not indicate how many infections each affected participant experienced, since the result is expressed as participants with an event rather than total infection episodes.

The registry does not provide organism-level results, infection severity, diagnostic criteria, adjudication procedures, or details about handling infections after the follow-up period. It also does not report whether all participants had outcome ascertainment through the full 60-day window. The endpoint therefore speaks only to the recorded occurrence of bloodstream bacterial infection within the specified period, as defined and reported in the registry.

7. Results Posted in the Registry

• Cipro: 11.7% (18/154) versus 11.7% (18/154) on control; risk difference 0.0 percentage points (95% CI -7.2 to 7.2), risk ratio 1.00 (0.54–1.85); unadjusted two-proportion z-test p = 1.000.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary comparison is framed as a difference in risks: the proportion with a bloodstream bacterial infection in the ciprofloxacin group compared with the proportion in the levofloxacin group. A risk difference expresses an absolute contrast on the proportion scale. A value of zero corresponds to equal observed proportions; values on either side of zero indicate a difference in one direction or the other, according to the comparison ordering. The registry reports a risk difference of 0.0, consistent with the equal event counts and equal listed group sizes.

What does the risk difference describe?

For a binary endpoint, each participant is classified according to whether the event occurred during the defined period. The risk in a group is the proportion classified as having the event. Comparing those risks makes the absolute gap visible. Unlike a relative measure, an absolute difference is stated in the same scale as the proportions being compared. It describes the observed contrast in this trial; it does not by itself establish causation or show how the result would translate to another population.

The registry also lists relative measures, including a risk ratio and odds ratio with confidence limits. Such measures express the contrast in different ways from a risk difference. They can be useful for describing relative association, but their meaning depends on the chosen measure and the frequency of the outcome. The registry comparison does not report a method explanation or an adjusted estimate. The primary interpretation therefore centers on the stated absolute risk difference and its uncertainty, rather than treating any one effect scale as a complete account.

What does the p-value mean?

The posted unadjusted p-value is 1.0. In general, a p-value measures how compatible the data are with a specified null model, assuming that model and the testing procedure apply. It is not the probability that the null hypothesis is true, and it is not a measure of clinical importance. A high p-value does not establish equivalence or prove that the groups have the same underlying risk. The registry record does not identify the test or provide enough detail to reconstruct its calculation independently.

The term unadjusted indicates that the posted p-value is not presented as adjusted for other variables or for multiple comparisons. The record does not report covariate adjustment, stratification, or a prespecified analysis hierarchy. It also does not state whether any analyses were planned beyond the recorded primary endpoint. The registry lists zero analysis records and no secondary outcomes. This does not explain why no analysis entries are present; the reason is not reported in the registry record.

Why read the confidence interval with the estimate?

A confidence interval gives a range of effect values compatible with the observed data under the interval method and its assumptions. It conveys uncertainty that a single point estimate cannot. Here, the interval for the risk difference crosses zero, meaning it includes no difference and also includes possible differences in either direction. The exact interval method is not reported in the registry record. General statistical interpretation does not make the interval a guarantee that the true effect lies within its endpoints.

Because this is a binary outcome, the precision of an estimated proportion or difference depends on the amount of information in the event and non-event counts, along with the statistical method. A low number of observed events can leave a wide range of plausible effects even when the point estimates match. This is a general property of binary-outcome estimation, not evidence about why this particular interval has its reported width. The registry does not state a sample-size justification or describe an interim analysis plan.

What can planning calculations say?

Sample-size planning calculations are prospective illustrations under stated assumptions; they are not observed power and do not replace the trial's results. The registry's planning description refers to a two-sided significance level of 0.05, equal allocation, and a normal approximation, with the effect to detect defined as a fraction of the observed effect. It states that non-time-to-event calculations are expressed as participants per arm. These are planning assumptions, not a report of the trial's own design rationale or proof that the study was adequately powered. The registry record does not report its original sample-size justification or target effect.

Randomization is intended, in general, to support a fair comparison by distributing prognostic factors without systematic allocation. The registry identifies the study as randomized but does not report the sequence-generation method, concealment procedures, or any stratification factors. The absence of masking is recorded, while the implications for outcome assessment depend on how infections were identified and recorded; those procedures are not reported. These design details matter when judging the protections against bias, but they cannot be inferred from the trial label alone.

10. Confidence Intervals and Precision

The point estimate for the risk difference is 0.0, and the reported confidence interval extends from a negative value to a positive value. Because the interval includes zero, the data do not isolate a difference in one direction. The interval also allows nonzero differences in either direction, so the equal observed counts do not establish that the underlying risks are identical.

A confidence interval is a measure of statistical uncertainty around an estimate, interpreted under the method used to construct it. In repeated use under its assumptions, a confidence-interval procedure has a coverage property; for a particular trial, the interval is not a probability statement about the fixed true value. The registry does not report the interval method or its assumptions in detail. Its endpoint values are not repeated; the essential interpretive feature is that the interval spans zero.

The width of an interval matters because it indicates how precisely the effect has been estimated. A point estimate can be close to no difference while uncertainty remains substantial. In this trial, the interval permits effects in both directions, and the registry does not report a prespecified threshold for what would count as an important difference. It is therefore not possible to use the interval to claim equivalence or rule out a clinically consequential contrast.

Precision should also be considered in relation to the endpoint's event count and the analysis population. The registry lists 18 participants with the outcome in each arm, with 154 listed in each group, but does not explain missing-data handling or confirm the completeness of follow-up. It does not give a sample-size justification or a target precision. These gaps prevent a fuller assessment of whether the precision matched the intended question.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome measure and zero analysis records, with no secondary outcomes posted. It does not report a multiplicity plan describing how the chance of false-positive findings would be controlled across multiple endpoints, time points, subgroups, or analyses. With no stated analysis hierarchy, it is not possible to determine whether the posted comparison was the only planned analysis or how any additional analyses were intended to be interpreted.

Missing-data procedures are not reported in the registry record. There is no description of how participants without an observed outcome by the end of the 60-day window were handled, whether follow-up was complete, or whether any outcome values were imputed. The outcome listing gives 154 participants for each group, but that alone does not establish that every participant had a known outcome or that all enrolled participants were included in the same analysis.

The analysis population is also not specified. The record does not say whether the comparison followed an intention-to-treat principle, included only participants who received assigned medication, or used another definition. Treatment adherence, discontinuation, crossover, and protocol deviations are not described. Without these details, readers cannot assess how closely the comparison reflects assignment to treatment versus treatment actually received.

The registry does not report a statistical analysis plan, subgroup analysis plan, interim analyses, or covariate-adjusted results. These elements should not be presumed absent from the conduct of the trial; their status is not reported in the registry record. Their absence from the record makes it difficult to evaluate prespecification, selective reporting, and the influence of analysis choices on the posted result.

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 5% reporting threshold.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: arrythmias (10.4%), fluid overload (1.9%), encephalopathy (1.3%), seizure (0.6%), bowel obstruction (0.6%).

The safety record lists serious adverse events affecting 23 participants in the levofloxacin group and 20 in the ciprofloxacin group; each group is listed with 154 at risk. It also lists zero participants with other adverse events in each group. These are registry counts, and the record does not explain the definitions, collection procedures, or time window used for these entries.

The registry lists arrythmias at 10.4%, fluid overload at 1.9%, encephalopathy at 1.3%, seizure at 0.6%, bowel obstruction at 0.6%, and pulmonary embolism at 0.6% among the top serious adverse-event terms. The denominator and whether these percentages apply across groups or to a particular group are not reported in the registry record. The terms should therefore be read as listed event categories, not as a basis for comparing treatment groups.

The death counts are 42 for levofloxacin and 54 for ciprofloxacin. The registry does not give the timing, causes, attribution, or analysis of these deaths in the safety entry. These counts cannot establish that deaths were caused by study treatment or support a causal comparison. Serious adverse-event and death entries require context about ascertainment, follow-up, and attribution, and that context is not reported.

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

The posted primary result establishes that the registry recorded the same number of participants with a bloodstream bacterial infection in each group: 18 in the levofloxacin group and 18 in the ciprofloxacin group. The listed group size is 154 for each regimen, and the registry comparison gives a risk difference of 0.0. This is a description of the observed result for the defined endpoint and follow-up window.

The result does not establish that the drugs have identical effects. The reported confidence interval spans zero and includes possible differences in either direction. The unadjusted p-value of 1.0 does not prove equality, and the registry does not report an equivalence or non-inferiority design. A conclusion that two treatments are equivalent requires an appropriate design and a prespecified margin; no such margin is reported in the registry record.

Nor does the result establish which regimen is preferable for an individual or for all transplant settings. The record does not provide participant characteristics, transplant-type-specific results, infection organisms, or details of treatment adherence and missing outcomes. Its findings cannot be generalized automatically to other populations, prophylactic regimens, follow-up periods, or clinical outcomes.

The reported safety entries also need cautious interpretation. Serious adverse-event counts and death counts appear in the registry, but timing, cause, attribution, and methods of collection are not reported. They should not be interpreted as proof of a treatment effect on safety. Taken together, the record describes a randomized comparison with equal observed primary-outcome counts but leaves uncertainty about the size and direction of the underlying difference and about important design and reporting details.

14. Why This Trial Matters Statistically

This trial addresses a practical statistical problem: two active prophylactic regimens can have identical observed event counts while the true difference remains uncertain. The point estimate is useful because it summarizes the observed comparison, but the confidence interval is necessary to understand how much the data constrain plausible effects. Equal counts are not the same as a precise demonstration of equivalence.

The registry also illustrates why the outcome definition and follow-up period matter. The primary endpoint counts participants with bloodstream bacterial infection by 60 days after transplant. That is a focused question, not a summary of all infection outcomes, all safety outcomes, or longer-term transplant outcomes. Clear limits on interpretation help prevent a narrow result from being extended beyond what was measured.

From a biostatistical perspective, the record highlights the value of reporting analysis populations, missing-data methods, interval methods, outcome ascertainment, and a prespecified analysis plan. Those details allow readers to evaluate how randomization and measurement support the estimate, and whether the uncertainty and analysis match the research question. They are not reported in the registry record, which constrains assessment even though the basic group counts and comparison are available.

The trial's randomized, parallel-group design and posted comparison offer a direct contrast of the assigned regimens. At the same time, the result emphasizes that a null-centered estimate and a non-significant comparison cannot by themselves rule out meaningful differences. The registry record is most informative when read as an estimate with uncertainty and acknowledged reporting gaps, rather than as a definitive ranking of the two regimens.

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 does not describe eligibility criteria, baseline participant characteristics, or the distribution of autologous and allogeneic transplants. As a result, the applicability of the recorded comparison to particular transplant populations cannot be assessed from this record. It also does not report the infection definitions, diagnostic criteria, organism-level findings, or procedures for confirming and adjudicating bloodstream infections.

Important trial-conduct details are not reported: the randomization method, allocation concealment, treatment duration, adherence, treatment changes, and use of additional antibacterial agents. The study is listed as unmasked, but the registry does not explain who assessed outcomes or whether outcome assessment was blinded. The potential influence of these factors on the recorded comparison is therefore unclear.

The registry does not report missing-outcome handling, an analysis-population definition, or whether all participants completed follow-up through the endpoint window. It also gives no original sample-size justification, statistical analysis plan, or prespecified threshold for an important difference. The confidence interval crosses zero, but the absence of a stated equivalence framework prevents interpreting the result as evidence of equivalence.

Safety interpretation is limited by missing context about adverse-event definitions, follow-up, attribution, and the denominator for the listed serious-event percentages. The timing and causes of deaths are not reported. These limitations are specific to the detail available in the registry record and do not establish that the relevant procedures or information were absent from trial conduct; their status is not reported.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Ciprofloxacin Versus Levofloxacin in Stem Cell Transplant — NCT03850379.
• Registry record last updated 2025-10-20; status last verified 2025-10.
• Results first posted 2025-10-20.
• 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, randomized, unmasked, parallel-group phase 2 trial sponsored by Henry Ford Health System. It lists 308 enrolled participants and compares levofloxacin 500 mg once daily with ciprofloxacin 500 mg BID. The primary endpoint is the proportion of participants with bloodstream bacterial infections up to 60 days after stem cell transplant.

The posted outcome lists 18 participants with the event in each group, with 154 participants listed per group. The registry reports a risk difference of 0.0 and an unadjusted p-value of 1.0; the confidence interval spans zero. The result is consistent with equal observed counts, but does not establish equivalence or exclude differences in either direction. Details of the analysis plan, missing-data handling, and analysis population are not reported in the registry record.