Hematology · Internal Medicine · 3 h ago
Western VTE risk scores show moderate performance in Chinese inpatients
A prospective multicenter study evaluated five hospital-associated venous thromboembolism (HA-VTE) risk models in 2,948 adult inpatients at 17 Chinese hospitals. All showed moderate discrimination over 90 days, with different trade-offs between sensitivity and specificity.
- Five scores had moderate discrimination in Chinese inpatients.
- Caprini was more sensitive but classified many as high risk.
- Performance declined in the prophylaxis subgroup.
- Authors call for local recalibration and simpler scores.
Researchers prospectively assessed the Caprini, Padua, IMPROVE, Kucher and Intermountain HA-VTE risk scores in 2,948 adult inpatients at 17 hospitals in China. Symptomatic VTE events were recorded through three months of follow-up; the overall 90-day incidence was 2.54%. The study also examined performance in subgroups, including surgical patients and those receiving prophylaxis.
Across the full cohort, area under the receiver operating characteristic curve values ranged from 0.726 to 0.759, and calibration was generally acceptable. Caprini had the highest sensitivity (0.83), but classified 42.6% of patients as high risk. Padua and Intermountain had greater specificity (0.75–0.81) and lower sensitivity (0.64–0.67). In surgical inpatients, Caprini had an AUC of 0.804, sensitivity of 0.90 and negative predictive value of 1.00, but requires more than 40 inputs. Performance was weaker among patients receiving prophylaxis; Caprini’s AUC fell to 0.682. The findings support caution in applying these scores uniformly: model choice involves trade-offs, and local recalibration or simpler population-specific tools may be needed. The study does not establish that using one score improves patient outcomes.
In practice
For adult inpatients in this Chinese cohort, all five evaluated scores showed moderate ability to discriminate 90-day symptomatic HA-VTE risk, but their sensitivity-specificity trade-offs differed. Caprini was more sensitive overall and performed best among surgical inpatients, though it classified many patients as high risk and requires more than 40 inputs. Performance was weaker in the prophylaxis subgroup. These prospective validation findings do not show that score-guided care improves outcomes or establish a universally preferred model; they support caution about transferring scores across populations without local evaluation.
AI-generated clinical interpretation · Not editor-reviewed
Is this summary clinically accurate?
Help fellow clinicians: your rating sends inaccurate summaries straight to our editors.
Sign in to rate this summary →Source
Blood Advances: Performance of Western HA-VTE risk models in East Asian populations: a prospective multicenter study ↗This is an automated AI-condensed summary that has not yet been reviewed by an editor. Always consult the full item at the original source.