Development of a Combined Risk Assessment Model for Venous Thromboembolism and Bleeding in Hematopoietic Stem Cell Transplantation Patients
Bibliographic record
Abstract
Background: Venous thromboembolism (VTE) and bleeding are common complications amongst allogeneic and autologous hematopoietic stem cell transplant (HSCT) patients. Balancing risk of bleeding and thrombosis in this population is particularly challenging as HSCT patients are concurrently pancytopenic and coagulopathic. However, risk factors for bleeding and thrombosis in HSCT patients are understudied. We aimed to derive a combined bleeding and VTE risk assessment tool to be used in the immediate post-transplant period (90 days). Methods: We conducted a retrospective cohort study of adult patients who underwent allogeneic or autologous HSCT between January 1, 2011 and December 31, 2021 at a tertiary care centre in Canada. The risk models patients were followed from transplant until occurrence of the event of interest with last follow up at Day 90 post-transplant, or death. Primary outcome for VTE was the occurrence of a confirmed thrombosis including proximal upper and lower extremity deep vein thrombosis, pulmonary embolism, or thrombosis of unusual sites, including cerebral and splanchnic. Primary outcome for bleeding was the occurrence of a major or clinically significant non-major bleeding event, not related to anticoagulation, as per the definition of the International Society of Thrombosis and Hemostasis. Group characteristics were compared using chi-square, Fisher's exact, or Student's T-tests as appropriate. Potential predictors for VTE and bleed amongst our patients were evaluated using single variable logistic regression and confirmed with multiple variable logistic regression with forward selection. For continuous variables, optimal cut-off points were estimated using ROC curves. Final VTE and bleed risk scores were derived based on weighted variables and compared using Cox regression with non-parametric bootstrapping used for internal validation. Results: A total of 476 patients (317 autologous, 159 allogeneic) were included. 47 patients (9.8%) suffered from VTE, and 32 patients (6.7%) suffered from bleed unrelated to anticoagulation within 90 days post-HSCT. A VTE risk assessment tool was derived, and internally validated, and it included: second central venous catheter insertion (due to complications from first central line) (2 points), and previous cancer treatment with steroids (1 point). The overall cumulative incidence of VTE was 31.8% in the high-risk group (>2 points) versus 7.6% in the low-risk group (0-1 points). The high-risk group was associated with higher mortality at 90 days (15.9% versus 3.2%, p <0.001; Figure 1). A bleeding risk assessment tool was also derived, and internally validated, and it included: baseline (day 0 of transplant) platelet count of <90 (1 point), and baseline hemoglobin of <96 (1 point). The overall cumulative incidence of bleed was 17.2% in the high-risk group (2 points), versus 4.3% in the low-risk group (0-1 points). The high-risk group was associated with higher mortality at 90 days (12.6% versus 2.6%, p<0.001). Amongst the low-risk bleed group, overall cumulative incidence of VTE was 10.03%, with 31.03% in the high-risk VTE group versus 8.33% in the low-risk VTE group (p<0.001). The positive VTE group was associated with higher mortality at 90 days (10.26% versus 1.72%, p=0.001; Figure 2). Amongst the high-risk bleed group, overall cumulative incidence of VTE was 9.19%, with 33.33% in the high-risk VTE group, versus 4.17% in the low-risk VTE group (p<0.001). The positive VTE group was associated with higher mortality at 90 days (37.5% versus 10.13%, p=0.026). Conclusion: We derived a predictive score, VTE and Bleeding in Marrow Transplant (VBMT) - risk score, for both VTE and bleeding in the immediate post-transplant period (90 days), to be used in conjunction with other criteria for bleeding and thrombosis. Amongst the low-risk bleed patients, we recommend strong consideration of chemical VTE prophylaxis, especially amongst the high-risk VTE group. These risk assessment models will help stratify patients based on bleed and thrombosis risk, and ultimately help guide VTE prophylaxis and surveillance strategies amongst HSCT patients.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.004 | 0.008 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".