RE-COVERY DVT/PE: Global Study on Treatment Secondary Prevention of Acute Venous Thromboembolism: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for RE-COVERY DVT/PE: Global Study on Treatment Secondary Prevention of Acute Venous Thromboembolism (NCT02596230), a registered phase not stated study. The primary outcome is Objective 2: Incidence Rate of ISTH (International Society on Thrombosis and Haemostasis) Major Bleeding and CRNMB (Clinically Relevant Non Major Bleeding) Per 100 Patient-years (Pt-yrs).

2. Clinical Question

3. Trial Design

RE-COVERY DVT/PE was a multinational, multicenter observational study based on new data collection. Its two objectives were to characterize patients with acute DVT and/or PE and to compare the safety and effectiveness of dabigatran etexilate regimens with VKA regimens in routine clinical practice. The registry lists an enrollment of 7797 and describes the study as completed.

For population characterization, patients with DVT and/or PE were enrolled for a cross-sectional description of baseline characteristics. For the treatment comparison, patients receiving dabigatran or a VKA were followed for outcome events for up to one year. The primary endpoint was the incidence rate of major bleeding and clinically relevant non-major bleeding per 100 patient-years over the 12 months following acute DVT and/or PE. Secondary outcomes included recurrent DVT and/or PE, VTE-related mortality, and all-cause mortality.

This was not a randomized trial. The registry states that allocation was not stated and that the treatment arms are not a like-for-like control comparison. The primary endpoint description says patients in the two treatment groups were weighted according to their estimated probability of receiving the opposite treatment, conditional on relevant observed baseline variables. Propensity scores were estimated with multivariable logistic regression.

The primary endpoint analysis used a restricted patient set: eligible patients within the region of propensity-score overlap who took the prescribed treatment at least once. A sensitivity analysis in this restricted set calculated incidence rates and cumulative risks without censoring at permanent discontinuation. Further details of treatment selection, follow-up procedures, and statistical planning are not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: All subjects were screened for eligibility to participate in the trial. Subjects attended specialist sites which would then ensure that they met all inclusion/exclusion criteria. The study enrolled and characterized all patients within 30 days after being diagnosed with an acute DVT and/or PE.

Objective 1, One Visit During 14 Days

Screening Period for Objective 2

Reasons for not completing this period:

Objective 2, One Year

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All eligible patients with a DVT and/or PE (regardless of treatment) were enrolled for cross-sectional characterisation of the VTE patient population

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: All Eligible: patients fulfilling all inclusion criteria and no exclusion criteria.

6. Primary Endpoint

The registry lists the primary outcome as Objective 2: Incidence Rate of ISTH (International Society on Thrombosis and Haemostasis) Major Bleeding and CRNMB (Clinically Relevant Non Major Bleeding) Per 100 Patient-years (Pt-yrs), measured over 12 months following acute deep vein thrombosis (DVT) and/or pulmonary embolism (PE)..

The primary endpoint measures the incidence rate of International Society on Thrombosis and Haemostasis (ISTH) major bleeding and clinically relevant non-major bleeding (CRNMB) per 100 patient-years. Its timeframe is 12 months following acute DVT and/or PE. The rate expresses the number of events relative to accumulated time at risk, allowing event occurrence to be described in relation to patients’ follow-up time rather than only as a simple count.

In general, rates per person-time are useful when follow-up duration differs among participants or when observation ends at different times. A confidence interval describes the statistical uncertainty around an estimated rate; a narrower interval generally indicates greater precision, while a wider interval indicates less precision. The registry identifies a 95% confidence interval as the measure of dispersion for this endpoint, but the estimate and interval are not reported in the registry record.

The endpoint combines major bleeding and clinically relevant non-major bleeding. The registry record names the event definitions but does not give further operational detail about how events were adjudicated or classified. The endpoint was analyzed in a restricted population selected by eligibility, propensity-score overlap, and having taken the prescribed treatment at least once. That analysis population is therefore not simply every person enrolled for cross-sectional characterization.

The endpoint’s intended role is to describe bleeding incidence in the two treatment groups while adjusting for measured baseline differences through weighting. It does not, by itself, establish that any difference between groups was caused by treatment. The observational design and the absence of a reported endpoint estimate limit what can be concluded about the comparative results.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary analysis used propensity-score weighting. In general, a propensity score is an estimated probability of receiving a particular treatment given recorded characteristics. Here, the registry says the score was estimated with multivariable logistic regression and that each patient in the two treatment groups was weighted in proportion to the probability of being assigned to the opposite treatment, conditional on relevant observed baseline variables. The goal of this approach is to make the weighted groups more comparable on those measured variables.

What does weighting do?

In general, weighting gives different influence to different patients so that the analyzed groups can resemble one another with respect to measured characteristics. A patient whose observed characteristics make their received treatment less expected can receive greater weight in an inverse-probability approach. The registry describes weights based on the probability of the opposite treatment. It does not report the specific baseline variables used, the score distributions, the resulting weights, or diagnostics showing whether balance was achieved. Those details are not reported in the registry record.

Weighting does not recreate random assignment. Its ability to reduce confounding depends on the measured variables and on how well the propensity model represents treatment assignment. In general, unmeasured differences between groups can remain after adjustment. Extreme weights can also make estimates unstable, and limited overlap can make some comparisons difficult. The restricted population was defined partly to address overlap: it included eligible patients within the region of propensity-score overlap who took the prescribed treatment at least once. The record does not report the size or characteristics of that analysis population.

How should an incidence rate be understood?

The endpoint is a rate of events per 100 patient-years. In general, a person-time rate relates event counts to the total time participants contribute while under observation and at risk. This can be useful when participants have different follow-up durations. It is not the same quantity as the proportion of participants who experience an event, because the denominator is accumulated observation time rather than simply the number of participants. No observed primary rate or patient-time denominator is reported in the registry record.

The registry specifies a 95% confidence interval as the dispersion measure. In general, a confidence interval conveys uncertainty around an estimate under the assumptions of the analysis. It does not by itself address bias from confounding, measurement, selection, or model misspecification. The primary estimate and its confidence interval are not reported, so neither the observed rate nor its precision can be interpreted.

What is the restricted-set sensitivity analysis?

The primary endpoint population was restricted to eligible patients in the propensity-score overlap region who took the prescribed treatment at least once. The described sensitivity analysis calculated incidence rates and cumulative risks without censoring at permanent discontinuation. In general, censoring ends a participant’s contribution to a particular analysis at a defined point. Changing whether follow-up is censored at discontinuation can alter the time included and the interpretation of the result. The registry record gives the sensitivity-analysis rule but does not report its estimate or how it compared with the main analysis.

Incidence rates and cumulative risks are different summaries. In general, an incidence rate describes events relative to person-time, whereas a cumulative risk describes the probability of an event over a specified period under the analysis definition. Their interpretation depends on follow-up and censoring choices. The registry names both calculations in the sensitivity analysis, but does not report the values or further details of the risk calculation.

What can be inferred from the reported method?

The method describes an adjusted observational comparison, not a randomized treatment effect. The registry record does not report the primary estimate, the confidence interval, the restricted-set size, model diagnostics, missing-data procedures, or enough detail to evaluate the implementation of weighting. It also does not report a protocol-based sample-size justification or interim-analysis plan. Without these elements, the statistical approach can be described, but its results and robustness cannot be evaluated.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 9 outcome measures and 5 analyses. It identifies the primary endpoint as the incidence rate of ISTH major bleeding and CRNMB and names recurrent DVT and/or PE, VTE-related mortality, and all-cause mortality as secondary outcomes. The registry record does not report a multiplicity adjustment, a testing hierarchy, or a prespecified strategy for interpreting multiple outcomes and analyses. It is therefore not possible to determine how the study accounted for the increased chance of chance findings when examining multiple outcomes.

Missing-data handling is not reported in the registry record. This includes the approach to missing baseline characteristics, incomplete outcome information, or other unavailable measurements. The record describes weighting based on observed baseline variables, but does not state how missing covariate values were handled before propensity-score estimation. It also does not describe imputation, complete-case analysis, or another missing-data method.

The primary endpoint population is described as a restricted set of eligible patients within the propensity-score overlap region who took the prescribed treatment at least once. The registry does not report the number included in this set, the number excluded, or the reasons for exclusion. It also does not report how many patients contributed data to each secondary outcome analysis.

The sensitivity analysis in the restricted set calculated incidence rates and cumulative risks without censoring at permanent discontinuation. This is a specified alternative to the censoring approach described for the primary analysis, but the results of either analysis are not reported. Other sensitivity analyses and the handling of loss to follow-up are not reported in the registry record.

11. 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: death (0.9%), cardiac arrest (0.5%), cardio-respiratory arrest (0.4%), respiratory failure (0.3%), upper gastrointestinal haemorrhage (0.3%).

The serious adverse event figures are presented by registry group. For “Not Assigned to Dabigatran Etexilate or Vitamin K Antagonist,” 3,714 were at risk and 60 were affected; the deaths count is 54. For “Dabigatran Etexilate (1),” 1,006 were at risk and 22 were affected, with 13 deaths. For “Dabigatran Etexilate (2),” 910 were at risk and 47 were affected, with 42 deaths. For “Vitamin K Antagonist (1),” 1,375 were at risk and 34 were affected, with 25 deaths. For “Vitamin K Antagonist (2),” 792 were at risk and 41 were affected, with 30 deaths. The other-event affected count is 0 in each listed group.

The labels “(1)” and “(2)” are not explained in the registry record, so their meaning should not be inferred. The counts are not adjusted comparative estimates and do not give event rates per patient-year. The record does not explain the denominator used for the listed serious-event terms, or how those terms were collected and classified. Listed terms include Death (0.9%), Cardiac arrest (0.5%), Cardio-respiratory arrest (0.4%), Respiratory failure (0.3%), Upper gastrointestinal haemorrhage (0.3%), and Pneumonia (0.3%). These safety summaries do not substitute for the primary endpoint analysis, whose estimate is not reported in the registry record.

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

The registry record establishes that RE-COVERY DVT/PE was an observational study of patients with acute VTE, with objectives to characterize the patient population and compare outcomes among patients treated with dabigatran or a VKA. It defines a primary bleeding-rate endpoint, specifies propensity-score weighting, and describes a restricted analysis population and a sensitivity analysis. These details explain what the investigators intended to estimate and how the primary comparison was approached.

The record does not report the primary bleeding rates, their 95% confidence intervals, or the comparative effect estimate. It also does not report results for the named secondary outcomes. Therefore, the available record does not establish whether the groups differed in bleeding, recurrent VTE, or mortality, nor does it establish the size or precision of any such difference. The serious adverse event counts provide descriptive information for registry groups, but are not a substitute for the adjusted endpoint estimates.

Even if comparative estimates were available, an observational comparison would require cautious interpretation. Propensity-score weighting can adjust for measured baseline differences included in the model, but cannot guarantee that groups are comparable on unmeasured factors. The stated restriction to the overlap region also means that the primary estimate would pertain to the eligible patients represented in that region, rather than automatically to all people with acute VTE.

In general, an estimate describes the data under the chosen analysis and its assumptions; it does not on its own prove a causal effect. Here, the absence of reported endpoint results prevents assessment even of the adjusted association. The record supports conclusions about study design and analysis intent, but not a conclusion about whether either treatment had better or worse outcomes.

13. Why This Trial Matters Statistically

This study is statistically notable because it addresses treatment comparisons using data collected in routine clinical practice rather than randomized assignment. Such studies can describe treatment use and outcomes in a broad clinical setting. At the same time, treatment choice can be related to patient characteristics, so a crude comparison may mix treatment differences with pre-existing differences between groups.

The registry describes propensity-score weighting as an adjustment strategy and restricts the primary analysis to eligible patients in a region of score overlap who took the prescribed treatment at least once. These choices highlight two central issues in observational analysis: adjustment for measured confounding and whether the groups have sufficient overlap for a meaningful comparison. In general, weighting can improve measured comparability, while restriction can focus inference on a population with more comparable treatment probabilities. Neither approach resolves unmeasured confounding by itself.

The endpoint is an incidence rate per 100 patient-years, which relates event occurrence to time under observation. That rate-based formulation is relevant to a study with follow-up described as up to one year, since follow-up time may vary. The registry also describes an alternative calculation without censoring at permanent discontinuation, making follow-up rules an important part of interpretation.

However, statistical importance depends on the actual estimates and their uncertainty. Since the primary and secondary results are not reported in the registry record, the comparative findings cannot be assessed. The record is informative about the intended design and methods, but it does not support a quantitative evaluation of treatment-group differences.

14. 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 main limitation for interpreting comparative outcomes is that the endpoint estimates are not reported in the registry record. The primary incidence rates, 95% confidence intervals, and treatment contrast are unavailable, as are estimates for the named secondary outcomes. This prevents assessment of effect direction, magnitude, and precision.

The study was observational, and the registry states that allocation was not stated and that the arms are not a like-for-like control comparison. Propensity-score weighting addresses measured baseline variables included in the model, but the record does not identify those variables or provide balance diagnostics. Residual confounding from unmeasured or inadequately modeled characteristics therefore cannot be assessed.

The primary endpoint analysis was restricted to patients in the propensity-score overlap region who took the prescribed treatment at least once. The registry does not report the number of people in that restricted set or how it differed from the full enrolled population. The generalizability of an estimate from that set would depend on its composition.

Additional analytic details are not reported in the registry record, including missing-data methods, outcome adjudication procedures, follow-up completeness, and the handling of loss to follow-up. The registry describes a sensitivity analysis without censoring at permanent discontinuation, but does not give its result. It also does not explain the “(1)” and “(2)” labels in the adverse-event groups or provide the denominators for the listed serious-event terms. These omissions limit evaluation of both the comparative analysis and the safety summaries.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: RE-COVERY DVT/PE: Global Study on Treatment Secondary Prevention of Acute Venous Thromboembolism — NCT02596230.
• Registry record last updated 2020-04-29; status last verified 2020-04.
• Results first posted 2020-04-29.
• 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.

18. Record Summary

RE-COVERY DVT/PE (NCT02596230) was a completed, multinational, multicenter observational study based on new data collection, with an enrollment of 7797. It aimed to characterize patients with acute DVT and/or PE and compare dabigatran etexilate with VKA treatment in routine clinical practice. Treatment groups were followed for up to one year.

The primary endpoint was the incidence rate of ISTH major bleeding and CRNMB per 100 patient-years during the 12 months following acute DVT and/or PE. The stated analysis used propensity-score weighting and a restricted population defined by eligibility, overlap, and taking the prescribed treatment at least once. The registry also lists secondary mortality and recurrent VTE outcomes.

The primary and secondary outcome estimates are not reported in the registry record. Serious adverse-event counts are listed by registry group, but do not provide adjusted comparative rates for the primary endpoint. The available information supports a description of study design and methods, not a conclusion about comparative treatment outcomes.