Thread by @MattMeiz

1/ Happy to share our study just out in American Journal of Hematology tinyurl.com/yxa9wots examining in-hospital mortality in COVID-19 patients treated with aspirin or intermediate-dose anticoagulation. A great collaboration between @YaleMed & @dfcidatascience

2/ One of the striking early observations in COVID-19 was that patients suffer from surprisingly high rates of thrombosis, including venous thromboembolism (e.g. PE), arterial thrombosis (e.g. stroke), and microvascular thrombosis in the lungs that may contribute to hypoxemia

3/ This occurred despite standard prophylactic anticoagulation, so many hospitals began implementing strategies to combat this disease process. Yale’s algorithm recommended intermediate-dose anticoagulation for patients with elevated D-dimer and later aspirin for all patients.

4/ We do not yet have evidence for these strategies, so we took this opportunity to retrospectively examine patient outcomes after treatment. It was essential to control for patients’ disease severity in this analysis, which we did in multiple ways.

5/ First, we established which patient characteristics predicted in-hospital mortality and should be accounted for in our analyses. We discovered that the Rothman Index is a strong clinical predictor, in addition to known factors including age, sex, obesity, and D-dimer levels.

6/ Within our overall cohort of 2785 hospitalized COVID-19 patients, we examined two smaller, nested cohorts that would allow us, as much as possible, to isolate the effect of the treatments of interest:

7/ 1) an “anticoagulation cohort” who received a maximum of intermediate- or prophylactic-dose anticoagulation (N = 1624), and 2) an “aspirin cohort” who were not on home antiplatelet therapy and received either in-hospital aspirin or no antiplatelet therapy (N = 1956)

8/ Within these cohorts, we performed propensity score matching to minimize the effect of patients’ disease severity and other confounders that affect mortality and may affect patients’ propensity to receive different treatments

9/ Once we had well-matched patient groups, we further performed multivariable regression analysis to model the effect of the treatments and other patient characteristics on the cumulative incidence of in-hospital death, accounting for the competing risk of hospital discharge

10/ We found that the use of intermediate-dose anticoagulation, compared to prophylactic-dose anticoagulation, was associated with a significantly lower risk of in-hospital death (HR 0.518 [0.308-0.872]). We can also visualize this with cumulative incidence curves.

11/ Separately, we found that the use of aspirin was associated with a lower risk of in-hospital death (HR 0.522 [0.336-0.812]). This was also true among patients admitted after aspirin was recommended for all patients (visualized here by cumulative incidence curves).

12/ Summary: In this retrospective study, we used propensity score matching and multivariable regression analysis to determine that two escalated-intensity antithrombotic therapies, aspirin and intermediate-dose anticoagulation, are associated with reduced in-hospital mortality

13/ These data are especially interesting in light of yesterday’s @remap_cap @ATTACC_COVID @ACTIV4a news that therapeutic-dose anticoagulation improved outcomes in moderately ill COVID-19 patients (after showing futility in critically ill patients): tinyurl.com/y4eoarz3

14/ So far, we have no clinical trial data on aspirin or intermediate-dose anticoagulation. More answers should come from the RCTs above and RECOVERY. Until then, we hope this study can help guide clinical decision making, as we continue to be embroiled in this pandemic.

15/ This was the product of extensive collaboration, led by @GeorgeGoshuaMD YiwenLiu @fine_rebecca @tigernole13 DonnaNeuberg #AlfredIanLee, with critical contributions from many others including KejalAmin @ConnPharmIntern @day_mac3 @StephenWangMD @MDChaar @hyungjchun @BenignHeme

16/ @RinderHenry & JonathanSiner. Thanks also to @KaminskiMed @DanaFarber @YaleHemOnc @YalePCCSM @YaleIMed. Here’s hoping for more progress on COVID-19 therapies – and less need for it – in 2021.