番茄社区

Preemptive therapeutic-dose anticoagulation for patients hospitalized with COVID-19 wasn't linked to better outcomes in a small observational study.

In fact, relative risk of in-hospital mortality with this strategy was 2.3 times greater (P=0.04) than seen with standard prophylactic dosing on multivariate analysis, with 38.7 vs 14.4 deaths per 100 patients.

"When we focused on patients with greater severity of disease [C-reactive protein ≥200], there was still no clinical improvement in outcomes of preemptive therapeutic anticoagulation," reported Jishu Kaul Motta, MD, of Danbury Hospital in Danbury, Connecticut, and colleagues in a , which was not yet peer reviewed.

Those findings countered those of a larger observational study that suggested better survival with therapeutic anticoagulation in hospitalized COVID-19 patients, although others have suggested .

It's possible that stronger thrombosis prevention is effective but just can't overcome competing risk of death from other disease processes in COVID-19, Motta's group suggested. "Regardless it does not seem from our analyses that therapeutic dosing of anticoagulation prevented overall disease progression."

However, "any interpretation other than acknowledging the need to study this prospectively in a randomized, controlled trial would be invalid," said Jason Katz, MD, director of cardiovascular critical care at the Duke University Health System in Durham, North Carolina, who was not involved with the study.

Aside from the small size, confounding and bias were also likely due to the "huge differences in patient characteristics between those who got prophylactic-dose anticoagulation and those that got full-dose anticoagulation," Katz pointed out.

The retrospective cohort study included 374 adults positive for SARS-CoV-2 treated at two large, acute care hospitals in western Connecticut in April 2020, with follow-up through June 12. Among them, 75 (20.1%) received therapeutic anticoagulation. Enoxaparin was given to nearly all (93.5%) at some time during their admission, while 14.8% took heparin.

The study excluded patients who received therapeutic-dose anticoagulation specifically for a thrombotic indication. Variables in the full logistic model included anticoagulant dosage, age, ethnicity, diabetes, history of cancer or heart disease, hyperlipidemia, peak C-reactive protein, intensive care, mechanical ventilation, and antibiotic use.

Another limitation was the lack of data on bleeding events, cautioned Behnood Bikdeli, MD, of Brigham and Women's Hospital and Harvard in Boston.

None of the studies done so far are able to really inform practice, he said. "What I do take out from this is the heterogeneity, the variability in the signals, and the effect sizes we keep seeing from these studies."

His group consensus, supported by the International Society on Thrombosis and Haemostasis and four other professional societies, recommended standard-dose prophylactic anticoagulation in most cases, despite noting a lack of data. The World Health Organization and NIH guidelines also suggest only using a higher-dose anticoagulant if there's a strong suspicion of thrombosis or in the context of a prospective study.

More than nine randomized controlled trials are underway to address questions of type and dose of antithrombotics for prophylaxis in COVID-19, Bikdeli noted.

The NIH has announced the to evaluate safety and effectiveness of varying types of antithrombotics for adults diagnosed with COVID-19.

Well-designed studies that include enrollment of patients in areas with higher case loads should be able to deliver some answers in 2 to 3 months, Bikdeli predicted. "Because if you practice in the ICU, these questions come up for almost every single patient with COVID-19. We see abnormal coagulation parameters and we ask ourselves, 'Do we increase the dose, do we not increase the dose?'"

But with the real risk of clinically important, even fatal, bleeding, "it's not really something that you can do just by clinical gestalt," he said. "Once we have the data, yes, clinical experience would be important to make patient-by-patient decisions. But without that prospective, ideally randomized, data, we're just going blindly."

Comment