This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Clinical Outcomes of Methicillin Resistant Staphylococcus Aureus (MRSA) Hospital-Based Pneumonia (NCT01819935), a registered phase not stated study. The primary outcome is Time to 30-day Mortality.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT01819935 |
| Phase | Phase not stated |
| Condition | Pneumonia |
| Allocation | Not Stated |
| Design model | Not Stated |
| Masking | Not Stated |
| Enrollment | 5271 |
| Arms in the primary outcome | n/a |
| Lead sponsor | Pfizer |
| Results posted | Yes |
2. Clinical Question
Population
The study population was drawn from patients admitted to Veterans Affairs hospitals between 1 January 2001 and 30 September 2010 with an ICD-9 code for MRSA and pneumonia. The primary outcome population is described as the full analysis set. The registry lists an enrollment of 5271, but does not report how participants were distributed between treatment groups or provide further details about eligibility.
Intervention
The treatment of interest was linezolid, also identified as Zyvox. This was an analysis of existing data, rather than a description of treatment assigned by a randomized study.
Comparator
The comparator was vancomycin. The registry does not describe the treatment-group allocation process, and the record characterizes the arms as not being a like-for-like control comparison.
Primary question
The primary outcome was time to 30-day mortality, measured in days. The study aimed to evaluate clinical outcomes related to MRSA hospital-based pneumonia by treatment and among subpopulations. The registry record does not report the treatment-group estimates for this outcome, so it does not establish the size or precision of a difference between treatments.
Other outcomes listed include time to therapy change, discharge from hospital, transfer out of the intensive care unit, intubation, 30-day readmission, 30-day MRSA reinfection, and clinical success. These outcomes describe a broader set of clinical events and care transitions, but their estimates are not reported in the registry record. The clinical question is therefore observational: how outcomes in the existing data were associated with treatment received. It is not a randomized comparison of assigned treatments, and treatment-group differences may reflect factors other than treatment.
3. Trial Design
This is a completed, non-interventional, retrospective study of existing data. Its stated purpose was to evaluate outcomes related to MRSA hospital-based pneumonia according to treatment and among subpopulations. The interventions named in the registry are linezolid (Zyvox) and vancomycin; the record does not describe treatment assignment or provide arm-specific descriptions.
The population for the primary outcome is described as patients admitted to Veterans Affairs hospitals from 1 January 2001 through 30 September 2010 with an ICD-9 code for MRSA and pneumonia. The registry gives an enrollment of 5271. It does not report the number analyzed in each treatment group, details of subgroup definitions, or how missing or incomplete records were handled. The study start is listed as 2012-12, primary completion as 2013-01, and status as completed. The outcome population’s retrospective data window is distinct from those listed study dates.
The primary outcome was time to 30-day mortality. The registry describes all-cause death occurring within 30 days of treatment initiation and says mortality was assessed from admission vital status databases. The recorded outcome parameter is the median, with full range as the dispersion measure. Additional outcomes concern treatment changes, discharge and ICU transfer, intubation, readmission, reinfection, and clinical success.
The registry does not identify the allocation process; it explicitly characterizes the treatment arms as not a like-for-like control comparison. Consequently, the design supports an observational association analysis, not a causal conclusion from randomized assignment. The registry record does not report the protocol, adjustment strategy, eligibility details beyond the stated population, or a statistical analysis plan.
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
| Stage | Linezolid | Vancomycin |
|---|---|---|
| STARTED | 328 | 4943 |
| COMPLETED | 328 | 4943 |
| NOT COMPLETED | 0 | 0 |
Analysis population for the primary outcome, as defined in the registry: FAS included all participants who were admitted to veterans affairs hospitals between 1 January 2001 and 30 September 2010 (3559 days) with an ICD-9 code for MRSA and pneumonia.
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: Full analysis set (FAS) included all participants who were admitted to veterans affairs hospitals between 1 January 2001 and 30 September 2010 (3559 days) with an ICD-9 code for MRSA and pneumonia.
| Characteristic | Unit | Linezolid | Vancomycin | Total |
|---|---|---|---|---|
| Participants analysed | 328 | 4943 | 5271 | |
| Age, Continuous | years | 69.1 (SD 12.5) | 69.1 (SD 12.7) | 69.1 (SD 12.7) |
| Sex: Female, Male: Female | Participants | 5 | 99 | 104 |
| Sex: Female, Male: Male | Participants | 323 | 4844 | 5167 |
6. Primary Endpoint
The registry lists the primary outcome as Time to 30-day Mortality, measured over Baseline (1 January 2001) up to 3559 Days (30 September 2010).
| Element | Registry entry |
|---|---|
| Result type | Median |
| Measure of spread | Full Range |
| Unit | days |
| Endpoint classification | Time-to-Event Analysis |
The primary endpoint is time to 30-day mortality, expressed in days. It concerns time from treatment initiation to all-cause death occurring within 30 days. The registry says mortality was assessed using admission vital status databases. For this endpoint, the outcome parameter is the median and the stated dispersion measure is the full range.
In general, a time-to-event endpoint captures both whether an event occurs and how long it takes to occur. This can be useful when timing matters, rather than treating every event as if it occurred at the same point in follow-up. A median describes the midpoint of observed event times under the analysis definition; a range describes the span between the smallest and largest reported values. Neither measure alone communicates the uncertainty of a treatment-group contrast.
Time-to-event analyses also need to account for participants whose event time is not observed within the relevant follow-up. In general, censoring methods can use information up to the last known follow-up without treating a person with no recorded event as if an event had occurred. The registry record does not report how censoring was defined or handled for this endpoint, nor does it give the median values, ranges, hazard ratios, or confidence intervals. The record therefore identifies what was measured, but does not permit assessment of the magnitude or precision of the comparison.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry describes a comparison of outcomes among patients treated with linezolid or vancomycin in retrospective Veterans Affairs data. It does not report the primary outcome estimates or sponsor-reported hazard ratios. As a result, no direction, magnitude, or statistical precision of the treatment-group association can be stated from the registry record.
The outcome is time to 30-day all-cause mortality after treatment initiation. In general, a hazard ratio compares the event rates over follow-up between groups, conditional on the time-to-event analysis. It is not itself a difference in median days or a direct measure of the proportion who died. Interpreting a hazard ratio requires knowing its estimate and uncertainty interval, and understanding how the analysis handled follow-up and censoring. Those values and details are not reported in the registry record.
The study is non-interventional and retrospective. Treatment was not described as randomly assigned, and the record states that the arms are not a like-for-like control comparison. Differences in the patients who received each treatment, their clinical circumstances, or their care could therefore contribute to observed differences in outcomes. Statistical adjustment can sometimes address measured differences, but it cannot automatically remove bias from unmeasured factors. The registry record does not report which variables were adjusted for or how subpopulations were analyzed.
The registry identifies a full analysis set of patients admitted to Veterans Affairs hospitals during the stated historical period with an ICD-9 code for MRSA and pneumonia. It does not report arm-specific analysis counts or outcome data. Enrollment alone does not establish how much information contributed to the primary comparison, particularly for a time-to-event endpoint where event occurrence and duration of follow-up both matter.
Other outcomes listed include therapy change, discharge, ICU transfer, intubation, readmission, reinfection, and clinical success. Their presence in the record indicates a broad outcome scope, but their estimates are not reported there. No claims about those outcomes’ direction or precision can be drawn. Likewise, no subgroup result can be interpreted because the subgroup definitions and estimates are not reported.
The record includes planning illustrations based on an observed effect and stated assumptions. Such calculations describe hypothetical planning requirements under those assumptions; they are not observed power and do not establish that the retrospective study was adequately powered. Without the outcome estimates, uncertainty intervals, and analysis details, the posted registry information is insufficient to quantify the treatment association or evaluate its robustness.
9. Statistical Methods Explained
The statistical interpretation depends on the endpoint, the comparison being made, and the information actually reported. For this study, the primary endpoint is time to 30-day mortality, and the registry indicates that the analysis is time-to-event. It also states that only sponsor-reported hazard ratios are used. However, hazard-ratio estimates and uncertainty intervals are not reported in the registry record, so the statistical result cannot be described quantitatively.
What does a time-to-event analysis describe?
General: Time-to-event methods retain information about when an event occurs, rather than reducing follow-up to an event/no-event status alone. The event in this study is all-cause death within 30 days of treatment initiation. Participants whose event is not observed during the analysis period may have incomplete event-time information. In general, such observations can be censored: the analysis uses the time during which the participant is known to be followed, without asserting that an event occurred after that point.
The validity of a time-to-event analysis depends on how the event, time origin, follow-up window, and censoring are defined. The registry identifies treatment initiation as the reference for the 30-day mortality outcome and says vital status was assessed from admission databases. It does not report the detailed censoring rules, the handling of incomplete vital-status information, or the precise time-to-event model specification. These details matter because different rules can change which observations contribute to risk estimates and how follow-up is represented.
How should a hazard ratio be read?
General: A hazard ratio compares the instantaneous event rates between groups over the period considered by the model. It is a relative measure of event rate, not a direct difference in survival time, a risk difference, or a count of deaths. A hazard ratio is typically interpreted alongside a confidence interval, which describes the range of estimates compatible with the data and model under the method’s assumptions. The confidence interval also conveys precision: a wider interval indicates greater uncertainty than a narrower one.
The registry record does not report the sponsor’s hazard-ratio values or confidence intervals for the primary outcome. It therefore does not support a statement about which group had a higher or lower estimated hazard, how large any association was, or how precisely it was estimated. The median and full range are identified as the summary measures for the outcome, but their treatment-specific values are also not reported in the record.
Why does the study design matter?
General: In a randomized comparison, random assignment is intended to make treatment groups comparable in expectation at the start of follow-up. In an observational comparison, treatment is received in routine care or otherwise determined outside randomized assignment. The patients receiving different treatments may differ in measured and unmeasured ways that also relate to outcomes. Those differences can confound a treatment-outcome association.
This registry describes a retrospective, non-interventional study and says allocation was not stated; it also warns that the arms are not a like-for-like control comparison. The record does not report an adjustment model or the covariates used. Consequently, it is not possible to assess from the registry whether measured differences were addressed, or how much residual confounding may remain. Even a detailed adjusted estimate would remain an observational association unless its assumptions were justified.
What do the planning illustrations mean?
The registry describes planning calculations based on the observed effect, with a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. The effect to detect is specified as a fraction of the observed effect. The record states that time-to-event planning uses a total number of events, using the Schoenfeld approach, while other planning results are expressed as participants per arm.
General: A sample-size calculation asks how much information a future study might need to detect a chosen effect under stated assumptions. Its answer depends on the effect targeted and the analysis assumptions. When the assumed effect is derived from an observed estimate, uncertainty in that estimate can make the planning result unstable. These calculations are planning illustrations, not observed power; they do not tell how likely the completed retrospective analysis was to detect an effect, and they do not replace reporting the observed estimate and its uncertainty.
The registry record does not report the protocol, sample-size justification for the retrospective study, interim analyses, or a full statistical analysis plan. It is therefore not possible to verify whether the planning assumptions match the actual analysis or whether other modeling choices were prespecified. The absent outcome estimates and analysis details limit both interpretation and independent appraisal.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists multiple outcomes: the primary endpoint of time to 30-day mortality and secondary outcomes involving therapy change, discharge, ICU transfer, intubation, readmission, reinfection, and clinical success. It also states that outcomes were considered by treatment and among subpopulations. When many outcomes or subgroup comparisons are examined, the chance of at least one apparently notable result can increase. General methods for addressing multiplicity include prespecifying a primary comparison, adjusting decision thresholds, or treating secondary findings as exploratory. The registry record does not report a multiplicity strategy or identify which subgroup analyses were planned.
The record does not report how missing data were handled, how participants lost to follow-up were treated, or whether outcome ascertainment differed between treatment groups. These issues are particularly relevant to time-to-event outcomes, where incomplete follow-up may be represented through censoring and depends on clear analysis rules. The primary endpoint description says mortality was assessed from admission vital status databases, but does not explain the completeness of those data or procedures for resolving unknown vital status.
The primary population is called the full analysis set and is described as patients admitted to Veterans Affairs hospitals in the stated period with an ICD-9 code for MRSA and pneumonia. The registry does not provide treatment-specific analysis counts, exclusions after cohort identification, or the handling of records lacking required information. It also does not provide the protocol or statistical analysis plan. These omissions prevent assessment of whether outcome selection, subgroup reporting, missing-data decisions, and analysis populations followed prespecified rules.
11. What the Result Does, and Does Not, Mean
The registry record establishes that a retrospective, non-interventional study examined clinical outcomes associated with linezolid and vancomycin among patients with MRSA hospital-based pneumonia in Veterans Affairs data. It identifies time to 30-day all-cause mortality as the primary outcome and lists several secondary outcomes. It does not report the estimated treatment-group results for the primary or secondary outcomes.
Accordingly, the record does not show the size or direction of any treatment-group association, the uncertainty around an estimate, or whether an estimate is compatible with a range of materially different effects. The sponsor-reported hazard ratios referenced by the analysis description are not reported in the registry record. Neither the listed median and full-range summaries nor confidence intervals are available there. No conclusion about treatment benefit, lack of benefit, or equivalence can be drawn from these omissions.
The design also constrains what an association could mean. Since the comparison is retrospective and non-interventional, and the record states the arms are not a like-for-like control comparison, treatment-group differences could reflect patient characteristics or care circumstances as well as treatment. The registry does not report the adjustment strategy, so the role of measured confounding cannot be assessed. Statistical adjustment, if used, would not by itself rule out unmeasured confounding.
The planning calculations described in the record are illustrations based on an observed effect and stated assumptions. They are not results about observed power and do not fill the gap left by unreported outcome estimates. The study’s enrollment figure does not, by itself, reveal the amount of information available for any specific endpoint or treatment comparison.
In general, interpreting a clinical comparison requires an estimate, a measure of uncertainty, a clear outcome definition, and enough design and analysis detail to judge bias and applicability. The registry provides some design and outcome definitions, but not the estimates and analysis details needed for a full statistical interpretation.
12. Why This Trial Matters Statistically
This study illustrates a common statistical challenge in clinical research: an observational treatment comparison can address outcomes in an existing care population, while remaining vulnerable to differences between the people who received each treatment. A large registry enrollment figure does not resolve that design issue. Comparability, outcome ascertainment, follow-up, and analysis choices all affect what a treatment association can support.
The primary endpoint is time to 30-day mortality, so interpretation requires more than a simple group count. General: Time-to-event analysis relies on event timing and follow-up information, and its results are best understood with both an effect estimate and an uncertainty interval. The registry names a median and full range as summary measures and refers to sponsor-reported hazard ratios, but does not report the corresponding values. This makes the record a useful example of why endpoint definitions and numerical results must be considered together.
The study also lists several secondary outcomes and subpopulation analyses. Such breadth can be clinically informative, but it raises statistical questions about prespecification, multiplicity, and selective reporting. The registry record does not state how those questions were handled. It likewise does not report the missing-data rules, adjustment variables, or detailed analysis plan.
Finally, planning illustrations based on an observed effect should be distinguished from evidence about the completed analysis. They express hypothetical information needs under stated assumptions and are not observed power. This distinction matters whenever sample-size planning is presented alongside retrospective results, especially when the treatment-effect estimates themselves are unavailable in the registry record.
13. 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 most direct limitation for statistical interpretation is that the registry record does not report the primary outcome estimates, sponsor-reported hazard ratios, or their uncertainty intervals. It identifies the outcome and summary measures but does not provide the values needed to assess magnitude, direction, or precision. Secondary outcome estimates are also not reported.
The study is retrospective and non-interventional. The registry says allocation was not stated and the arms are not a like-for-like control comparison. Treatment-group differences may therefore reflect differences in patient characteristics or care, as well as treatment. The record does not report whether statistical adjustment was used, which variables it included, or how subgroup comparisons were defined.
The primary outcome population is defined using hospital admission, an ICD-9 code for MRSA and pneumonia, and the stated historical data window. The record does not report further eligibility criteria, treatment-group counts, exclusions, or details about how records were linked to admission vital status databases. It also does not explain the handling of incomplete outcome information, censoring, or loss to follow-up.
Several outcomes and subpopulation analyses are listed, but the registry does not report a multiplicity plan or identify which analyses were prespecified. The protocol, sample-size justification, interim analyses, and statistical analysis plan are not reported in the registry record. These omissions limit evaluation of analysis choices and reporting completeness.
The planning assumptions are described, but planning illustrations do not establish observed power or remedy unreported results. The registry’s enrollment figure also does not indicate how much endpoint-specific information was available for each treatment comparison. These limitations mean the record supports a description of the study question and design, but not a quantitative conclusion about its findings.
14. Related Tutorials
15. Related Statistical Calculators
16. Sources
- ClinicalTrials.gov: Clinical Outcomes of Methicillin Resistant Staphylococcus Aureus (MRSA) Hospital-Based Pneumonia — NCT01819935.
- Registry record last updated 2014-01-30; status last verified 2013-12.
- Results first posted 2014-01-30.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
17. Record Summary
The registry describes a completed, non-interventional retrospective study of existing data on MRSA hospital-based pneumonia. It compares outcomes associated with linezolid and vancomycin among patients admitted to Veterans Affairs hospitals during the stated historical period. Enrollment is listed as 5271.
The primary outcome is time to 30-day all-cause mortality, measured in days, with median and full range specified. Secondary outcomes include therapy change, discharge, ICU transfer, intubation, readmission, reinfection, and clinical success. The study is described as evaluating outcomes by treatment and among subpopulations.
The record states that the treatment arms are not a like-for-like control comparison and does not report allocation. Primary and secondary outcome estimates, sponsor-reported hazard ratios, uncertainty intervals, missing-data procedures, and detailed analysis methods are not reported in the registry record. The listed planning calculations are illustrations, not observed power.