Perceived relative clinical significance of the risks of venous thromboembolism and bleeding from pharmacologic thromboprophylaxis in children: an international survey of pediatric providers: communication from the ISTH SSC Subcommittee on Pediatric and Neonatal Thrombosis and Hemostasis.

BACKGROUND

In the absence of robust data, the decision to provide primary pharmacologic thromboprophylaxis in hospitalized children depends on the perceived risks of hospital-acquired venous thromboembolism (HA-VTE), balanced against the acceptable bleeding risk from thromboprophylaxis.

OBJECTIVES

We aimed (1) to determine the perceived relative clinical significance of HA-VTE and bleeding from thromboprophylaxis among pediatric providers, and (2) to determine patient and provider characteristics that modify this perception.

METHODS

We conducted a cross-sectional survey among pediatric providers. Using case vignettes and Likert scales, we quantified the likelihood of providing thromboprophylaxis to prevent 4 types of HA-VTE, varying in severity, given equal risks of developing 6 bleeding characteristics, ranging from minor to fatal. We evaluated 9 patient characteristics, including patient age, and 7 provider characteristics that may modify this decision.

RESULTS

A median of 145 (IQR, 139-152) respondents answered each vignette. For all bleeding characteristics other than minor bleeding, respondents were less likely to provide thromboprophylaxis to prevent nonsevere HA-VTE, except for pulmonary embolism with hemodynamic compromise. HA-VTE type, bleeding, and patient characteristics, particularly infant age, were associated with a change in scores in the likelihood of providing thromboprophylaxis. However, changes were minimal, with only infant age associated with a change of ≥1 point.

CONCLUSION

Among pediatric providers, the acceptable bleeding risk for primary pharmacologic thromboprophylaxis varies with perceived HA-VTE severity. Specific features, particularly infant age, modified the respondents' decisions despite the stated risks. Our findings can provide a robust, semiquantitative approach to address these competing risks clinically and in future research.