Hemorrhage Risk Prescribed Arixtra: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Hemorrhage Risk Prescribed Arixtra (NCT01064362), a registered phase not stated study. The primary outcome is Number of Participants With the Indicated Types of Haemorrhages During Hospitalization or Follow-up for Major Orthopaedic Surgery of Lower Limbs (MOSLL).

2. Clinical Question

The study addresses a post-marketing safety question in patients having hip fracture surgery and/or hip or knee replacement surgery. It uses linked demographic, drug-dispensing, and hospital morbidity information from PHARMO RLS rather than describing a randomized treatment assignment. The comparison is clinically relevant to the stated concern about haemorrhage during thromboprophylaxis, but the registry record does not provide enough results to quantify or interpret a difference between groups.

3. Trial Design

The registry lists the study as completed and records enrollment of 13442. Its title describes a comparison of haemorrhage risk in patients prescribed Arixtra and LMWH. The study summary describes an observational assessment using patients in the Netherlands PHARMO RLS database. Although the registry record describes treatment groups, it does not state how allocation occurred; the comparison is not randomized and the arms are not a like-for-like control comparison.

The population is described as patients age 18 years and older with hip fracture surgery and/or hip or knee replacement surgery who received fondaparinux or LMWH as their initial in-hospital thromboprophylactic agent. The summary specifies at least three months of PHARMO RLS history before cohort entry. It also says that patients with a history of hospitalization for haemorrhage, renal failure, or liver failure in the past three months were excluded. The primary-outcome population description refers to patients with in-hospital pharmacy data and follow-up between January 2003 and December 2008.

The registry identifies fondaparinux sodium and LMWH as the interventions. It describes a single treatment setting—prophylaxis following major orthopaedic surgery—with both treatments represented. The study aimed to provide additional observational data from a European country for review of haemorrhage and post-marketing safety. The record lists descriptive statistics for characteristics such as age, gender, treatment length, comorbidities, concomitant medications, and other covariates as planned descriptions, but it does not report those summaries or explain how any covariates were used in comparative analysis.

The record gives start and primary completion dates of 2010-01 and 2010-06. These registry dates do not establish the timing of each patient's treatment or follow-up; the outcome description defines follow-up separately. The analysis plan, treatment-group sizes, and participant flow are not reported in the registry record. Enrollment should not be treated as the number of participants included in an outcome analysis because no analyzed population is described.

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

Analysis population for the primary outcome, as defined in the registry: From the PHARMO RLS, all patients \>=18 years of age with in-hospital pharmacy data, a primary discharge diagnosis for hip fracture and/or a hospitalization for MOSLL, and follow-up between January 2003 (introduction of Arixtra) and December 2008.

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 Number of Participants With the Indicated Types of Haemorrhages During Hospitalization or Follow-up for Major Orthopaedic Surgery of Lower Limbs (MOSLL), measured over Follow-up continued until the date of first event, death, end of initial therapy, hospital discharge, end of follow-up in PHARMO RLS, or 60 days after discharge, whichever came first.

The primary endpoint is the number of participants with the indicated types of haemorrhage during hospitalization or follow-up for major orthopaedic surgery of the lower limbs. Haemorrhages were identified using ICD-9-CM codes. The registry description says PHARMO RLS links patient demographics, drug dispensing, and hospital morbidity records, and characterizes it as a population-based, patient-centric system with information on approximately 2.3 million inhabitants in the Netherlands.

The stated follow-up ended at the earliest of the first event, death, the end of initial therapy, hospital discharge, the end of follow-up in PHARMO RLS, or 60 days after discharge. This definition specifies that observation could end for reasons other than a haemorrhage event. The record does not report how many participants reached each stopping point, or whether follow-up completeness differed between treatment groups.

A count of participants with an event is a binary-outcome summary: each participant either meets the specified event definition during the observation period or does not. Such counts can describe the occurrence of coded haemorrhage in the study population, but counts alone do not convey the amount of observation time or account for differences in patient characteristics. The endpoint can be useful for a safety-focused comparison, while interpretation still depends on clear outcome ascertainment, comparable follow-up, and an appropriate analysis of the nonrandomized groups.

The registry names a secondary outcome concerning indicated haemorrhages during hospitalization for major orthopaedic surgery of the lower limbs. It does not report endpoint results or explain the relationship between the primary and secondary definitions in an analysis. Neither outcome can be interpreted as a treatment effect from the endpoint descriptions alone.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes a binary outcome: the number of participants with specified haemorrhages during hospitalization or follow-up. In general, a binary outcome classifies each participant according to whether an event meeting the definition was recorded during the relevant observation period. Group event counts can be descriptive, but a comparison also needs the relevant group denominators and a clearly defined analysis. Neither comparative counts nor denominators are reported in the registry record.

What does a confidence interval add?

In general, a confidence interval expresses the uncertainty around an estimated effect under the assumptions of the statistical method. Its width reflects the precision of that estimate; it does not remove bias from nonrandom treatment assignment, incomplete outcome information, or an unsuitable model. The registry record reports no effect estimate or confidence interval, so the precision of any comparison cannot be evaluated.

Why does nonrandom allocation matter?

The study summary describes patients who received fondaparinux or LMWH in routine care, and the registry does not state that allocation was randomized. In general, randomization helps balance measured and unmeasured characteristics across treatment groups in expectation. In an observational comparison, treatment selection can be associated with patient characteristics that are also related to the outcome. A crude difference in event counts can therefore reflect differences in the groups as well as any treatment-associated difference. Statistical adjustment may address measured characteristics, but it cannot automatically resolve unmeasured confounding. The record lists descriptive characteristics but does not report an adjustment method or comparative model.

How should the follow-up rule be understood?

The primary endpoint description defines follow-up as continuing until the first event, death, end of initial therapy, hospital discharge, end of follow-up in PHARMO RLS, or 60 days after discharge, whichever came first. In general, this is a rule for determining when outcome observation stops. Different stopping times can mean that participants contribute different opportunities for an event to be recorded. A count-based summary does not itself describe time at risk. A time-to-event method can account for differing observed times under its assumptions, but the registry record does not report such an analysis or the handling of censoring.

Why can over-dispersion matter?

Over-dispersion is a general issue for models that assume a particular relationship between a mean and variability, especially count models. If variability exceeds what a model assumes, standard errors may be too small and confidence intervals too narrow unless the model accounts for it. The registered primary outcome is described as participants with an event, not as a recurrent-event count for each participant. No count model, dispersion assessment, or model-based estimate is reported, so over-dispersion cannot be evaluated as an analysis feature of this record.

What is known about statistical planning?

The registry record does not report a statistical analysis plan, sample-size justification, interim analyses, or a planned adjustment strategy. A planning illustration described for this material is based on a stated fraction of an observed effect, uses a two-sided significance level of 0.05, equal allocation, and a normal approximation; time-to-event illustrations use the Schoenfeld method and give a total event count, while other illustrations refer to participants per arm. These are planning assumptions, not evidence of the study's original design or observed power. The record gives no comparative effect estimate with which to judge such illustrations, and they cannot replace the missing outcome analysis.

For a binary endpoint, a useful analysis would ordinarily make the event definition, analysis population, denominators, handling of incomplete outcome information, and comparison method explicit. In general, an unadjusted comparison describes the groups as observed, whereas an adjusted comparison attempts to account for specified measured factors. The choice of measure and model should fit the outcome and study design. The registry record does not report which approach was used, whether estimates were adjusted, or whether assumptions were checked. Statistical significance, precision, and clinical interpretation therefore cannot be assessed from the posted results.

10. Multiplicity, Missing Data and Analysis Populations

The registry record identifies a primary haemorrhage endpoint and a secondary outcome concerning haemorrhages during hospitalization for major orthopaedic surgery of the lower limbs. It does not report a multiplicity strategy, explain whether additional outcomes or subgroup comparisons were tested, or state how any multiplicity would be handled. In general, assessing several outcomes or comparisons can increase the chance of a chance finding if the analyses are interpreted without regard to how many were considered. No multiplicity-adjusted results are reported.

Missing-data handling is not reported in the registry record. The description does not state how patients with incomplete demographic, dispensing, hospital morbidity, or follow-up information were handled, or whether any outcome data were missing. It also does not provide participant flow or group-specific analysis populations. It is therefore not possible to determine whether missing information could have affected the observed comparison.

The record names the eligible population and describes exclusions before study inclusion, including recent hospitalization for haemorrhage, renal failure, or liver failure. It does not state whether exclusions or loss of database follow-up were summarized by treatment group or whether any sensitivity analyses examined their possible effects. The endpoint's follow-up stopping rules are specified, but the corresponding censoring or incomplete follow-up analysis is not reported in the registry record.

These gaps do not demonstrate that the study failed to address multiplicity or missing data; they mean that the relevant methods and results are not reported in the registry record. Any interpretation of the endpoint would require those details, alongside the comparative event counts and analysis populations.

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

The registry record establishes that a completed observational study was intended to compare haemorrhage among patients prescribed fondaparinux or LMWH after major orthopaedic surgery. It describes eligibility, the linked database, the coded outcome, and the follow-up stopping rule. It does not report the comparative outcome results needed to say how haemorrhage occurrence differed between the treatment groups.

Accordingly, the posted information does not show that either treatment had a higher or lower haemorrhage count, and it does not establish that the groups were equivalent. Absence of a reported estimate is not evidence of no difference. Without event counts, group denominators, and an effect estimate with its uncertainty, neither the size nor precision of a difference can be assessed.

The study's observational design is also important. In general, a comparison of patients who received different treatments may reflect differences in who was treated with each option, as well as any treatment-associated differences. The registry record does not report how this issue was addressed in analysis. Even if an adjusted comparison were available, interpretation would depend on the variables measured and the assumptions used.

The outcome description concerns haemorrhages identified through ICD-9-CM codes during a defined period of hospitalization and follow-up. It does not provide detail on event severity, validation, or adjudication. The record's safety objective and post-marketing context describe why the question was studied, but they do not supply evidence of a comparative result.

The recorded enrollment of 13442 is not an outcome finding and does not reveal the size of either analyzed group. The registry status of completed likewise does not communicate the result. The available record can therefore support a summary of the study question and design, but not a conclusion about relative haemorrhage risk or a clinical recommendation.

12. Why This Trial Matters Statistically

This study is statistically relevant because it frames a post-marketing safety question using data from routine clinical records rather than a stated randomized allocation. Linked demographic, drug-dispensing, and hospital morbidity information can describe treatment and outcomes in a defined healthcare database. At the same time, observational comparisons require careful attention to differences between treatment groups and to how outcome and follow-up information are recorded.

The endpoint also illustrates an important distinction between defining an outcome and reporting an estimate. The registry specifies haemorrhage identified by ICD-9-CM codes and gives a follow-up rule, but provides no group event counts or comparative analysis. Without those results, readers cannot assess the magnitude or precision of an association. A detailed endpoint definition is useful, but it cannot stand in for an analysis.

Follow-up may end for several reasons under the stated rule, including hospital discharge, the end of database follow-up, or the specified post-discharge limit. In general, differing observation periods can affect event detection and should be described when comparing groups. The registry record does not report follow-up summaries or explain how differing observation times were handled.

The record also shows why registry status, enrollment, and results answer different questions. The study is marked completed and lists enrollment of 13442, but these registry entries do not reveal how many people contributed to the primary analysis or what the outcome comparison found. For statistical interpretation, the missing denominators, effect estimates, uncertainty measures, and analytic details are central.

As a result, this record is informative about the intended safety question and the data source, while remaining insufficient for judging the observed comparative haemorrhage results. That distinction is important when reading registry entries as evidence: a study's objective and design can be described even when its statistical findings are not reported.

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.

A central limitation of the registry record is the absence of comparative results. It gives no primary outcome event counts by group, denominators, effect estimates, confidence intervals, or p values. Although enrollment is recorded as 13442, the analyzed population and group sizes are not reported. Statistical precision and the direction or magnitude of any difference cannot be evaluated.

The study is not described as randomized, and allocation is not stated. The registry record also says the arms are not a like-for-like control comparison. The summary lists descriptive characteristics, but does not report baseline group comparisons or explain whether measured covariates were used in adjustment. Possible confounding from differences between patients receiving each treatment therefore cannot be assessed from the record.

Haemorrhage is identified through ICD-9-CM codes in linked records. The record does not report code validation, clinical adjudication, or sensitivity analyses using alternative outcome definitions. It also does not report group-specific follow-up duration, the distribution of follow-up stopping reasons, or methods for handling incomplete follow-up. Those omissions limit assessment of outcome ascertainment and opportunity for event detection.

The registry record does not report how many patients were excluded at each stage, whether missing data occurred, or how missing information was handled. It does not specify the analysis populations, the statistical analysis plan, a covariate-adjustment strategy, or a multiplicity approach. No interim analyses or sample-size justification are reported in the registry record. These limitations describe what is absent from the record; they do not establish what analyses may have been conducted outside it.

14. Related Tutorials

15. Related Statistical Calculators

16. Sources

• ClinicalTrials.gov: Hemorrhage Risk Prescribed Arixtra — NCT01064362.
• Registry record last updated 2015-05-12; status last verified 2011-06.
• Results first posted 2011-02-11.
• 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.

17. Record Summary

NCT01064362 is a completed registry study titled “Hemorrhage Risk Prescribed Arixtra.” It records enrollment of 13442 and lists GlaxoSmithKline as sponsor. The study summary describes an observational comparison of fondaparinux sodium and LMWH for thromboprophylaxis following hip fracture surgery and/or hip or knee replacement surgery, using the Netherlands PHARMO RLS database.

The primary outcome is the number of participants with indicated haemorrhages during hospitalization or follow-up, identified using ICD-9-CM codes. Follow-up ends at the earliest of the first event, death, end of initial therapy, hospital discharge, end of PHARMO RLS follow-up, or 60 days after discharge.

The registry record reports no comparative outcome estimates or statistical analyses. Group event counts, analysis populations, confidence intervals, p values, and details of adjustment and missing-data handling are not reported in the registry record. The design and endpoint can be described, but comparative haemorrhage findings cannot be determined from this record.