VTE risk score and bleeding risk score may inform thromboprophylaxis decision post-discharge
Presenter: Scott Woller, MD
In hospitalized medical patients, the risks of 90-day hospital-associated venous thromboembolism (HA-VTE) and hospital-associated major bleeding (HA-MB) can be estimated at the time of discharge through the use of a risk-estimation tool, based on ubiquitous laboratory values, to inform decisions on extended duration thromboprophylaxis, according to a model presented by Scott Woller, MD, from Intermountain Health Care, Murray, UT.
Further, with broad adoption of electronic medical records, risk of these outcomes could be assessed with no additional time or expense, he said.
From 50% to 75% of all HA-VTE events occur after hospital discharge. “Yet evidence suggests that there is a narrow margin that exists between the outcomes of bleeding and thrombosis when post-discharge thromboprophylaxis is administered,” said Dr. Woller. “So the real conundrum is, How do we identify those patients that would best benefit from extended- duration thromboprophylaxis (EDT) without bleeding complications?”
Current VTE risk scores, such as the Modified IMPROVE VTE risk score that incorporates D-dimer, have enduring limitations that include the difficulty of using computerized interrogation to identify VTE clinical risk factors and inclusion of a measurement (D-dimer) that is not routinely obtained in clinical practice.
“What we were especially interested in doing is understanding if we could generate a risk score from common covariates that are ubiquitously available in routine clinical care,” he said. “We found that those that were most predictive of VTE and major bleeding 90 days post-discharge were age, white blood cell count, red cell distribution width, platelet count, blood glucose, sodium, and creatinine.”
The Intermountain risk score (IMRS) is a highly predictive mortality risk estimation tool derived from components of the complete blood cell count and the basic metabolic panel. The research team generated VTE and bleeding risk scores as refinements to the IMRS, reweighing components from the complete blood cell count and basic metabolic panel to be predictive of 90-day outcomes.1
Their dataset included 45,669 medical patients who survived hospitalization in which 1,038 (1.6%) VTE and 611 (1.3%) major bleeding events were recorded. This cohort was split into derivation and validation cohorts. A Cox model was fit in the entire derivation cohort with covariates that included red cell distribution width, blood urea nitrogen, age, glucose, white blood cell count, platelet count, red blood cell count, sodium, and creatinine. The complete blood cell count and basic metabolic panel candidate variables were split into quintiles. Each quintile was weighted based on “predictiveness,” and points were assigned from 0 for the least predictive quintile (referent) to a maximum of 5.
A 90-day VTE rate of 2% was selected as clinically important, and correlated to an IMRS of 7. A 90-day major bleeding rate of 1% was selected as clinically important, and correlated to an IMRS of 8.
In the derivation set, an HA-VTE IMRS of 7 or higher was associated with an area-under-the-receiver-operating-characteristic curve (AUC) of 0.646, and an HA-MB IMRS of at least 8 was associated with an AUC of 0.692. In the validation cohort, an HA-VTE IMRS of 7 or higher generated an AUC of 0.60, and an HA-MB IMRS of 8 or higher generated an AUC of 0.643.
Survival at 90 days free from VTE was 1.6% in patients with an HA-VTE IMRS less than 7 and 2.6% in those with an HA-VTE IMRS of 7 or higher (hazard ratio 1.69; 95% confidence interval 1.35–2.13). Major bleeding in patients HA-MB IMRS of 8 or higher occurred at a rate of 1.9% vs 0.8% in those with an HA-MB IMRS less than 8 (hazard ratio 2.35; 95% confidence interval 1.75–3.16).
“If validated, the strength of this work will be that we will be able to present to physicians, in a ‘just-in-time’ fashion, individual patient risk for thrombosis within 90 days after discharge, but also the risk for major bleeding,” said Dr. Woller. “Because of the narrow margin that exists between thrombosis risk and bleeding risk in discharged hospitalized medical patients, this could help inform optimal care.”
Reference
- Woller D, Stevens S, Snow G, et al. Derivation and validation of the HA-VTE and HA-MB Intermountain risk scores from ubiquitous clinical biomarkers to predict 90-day hospital-associated venous thromboembolism and major bleeding among medical patients. CHEST 2020; 158(4 Suppl]:A2452–A2453. DOI:https://doi.org/10.1016/j.chest.2020.09.034