Haemorrhagic complications in COVID-19 patients linked to in-ICU mortality risk

A large global study of patients with severe COVID-19 admitted to the intensive care unit (ICU) has shown a 14 percent prevalence of haemorrhagic, coagulopathic, and thrombotic (HECTOR) complications in these patients, with mortality risk increased in those with haemorrhagic complications. 

This observational study was conducted in 229 ICUs across 32 countries and involved 11,969 individuals aged ≥16 years who were hospitalized for symptomatic, severe COVID-19 between January 2020 and December 2021.

Fourteen percent of the cohort (n=1,732) experienced HECTOR complications. Ten percent of patients (n=1,249) experienced acute thrombosis, with 57 percent (n=712) experiencing pulmonary embolism, 33 percent (n=413) myocardial infarction or cardiac ischaemia, 7.4 percent (n=93) deep vein thrombosis, 3.9 percent (n=49) ischaemic stroke or cerebrovascular accident, and 4.4 percent (n=55) other thromboembolisms. [ISTH 2022, abstract LB 02.2]

Haemorrhagic complications were detected in about 5 percent of patients (n=579), primarily gastrointestinal haemorrhage (48 percent, n=276), haemorrhagic stroke (14 percent, n=83), and pulmonary haemorrhage (13 percent, n=77). Other common haemorrhagic complications included skin and soft tissue haemorrhage (13 percent, n=74) and haemorrhage at extracorporeal membrane oxygenation (ECMO) cannula site (12 percent, n=68). Eleven patients experienced disseminated intravascular coagulation.

While there was overlap between the different complications, most patients only experienced one type of complication, said study author Dr Jonathon Fanning from Johns Hopkins Medicine, Baltimore, Maryland, US, at ISTH 2022.

Patients with HECTOR complications had significantly higher BMI and were more likely to have cardiac or renal disease, hypertension, and diabetes, and a history of smoking than those without complications. Patients aged <60 years were also at greater risk of haemorrhagic complications than those aged ≥60 years (6.2 percent vs 3.8 percent; p=0.041). However, sex or ethnicity did not appear to influence the risk of HECTOR complications. 

Compared with those without complications, patients with HECTOR complications had greater disease severity upon admission (median SOFA* score 6.0 vs 4.0; median APACHE* II score 17.0 vs 14.0; p<0.0001 for both). They were also more likely to require mechanical ventilation (82 percent vs 73 percent; p<0.0001) for a longer duration (median 18.5 vs 13 days; p<0.0001), ECMO (25 percent vs 7.5 percent), or transfusion of any blood product (54 percent vs 43 percent).

Thirty-seven percent of patients (n=4,425) died in the ICU. Overall mortality in the ICU was higher in patients with vs without HECTOR complications (44 percent vs 36 percent; p<0.0001), specifically 28-day mortality (25 percent vs 13 percent; p<0.001) and 90-day mortality (32 percent vs 15 percent; p<0.0001). In-ICU mortality rates due to haemorrhagic and thrombotic complications were 57** and 53 percent, respectively. Multiorgan failure was the most common cause of death among patients with HECTOR complications (36 percent), and respiratory failure the most common among those without complications (33 percent).

Mortality risk was not significantly increased in patients with vs without HECTOR complications (adjusted hazard ratio [adjHR], 1.01, 95 percent confidence interval [CI], 0.92–1.12; p=0.784). This was because the elevated mortality risk among patients with haemorrhagic complications (adjHR, 1.26, 95 percent CI, 1.09–1.45; p=0.002) was offset by the lower risk among those with thrombotic complications (adjHR, 0.88, 95 percent CI, 0.79–0.99; p=0.030).

Among survivors, duration of ICU stay was longer among those with vs without complications (median 18 vs 12 days).

Subanalysis by ECMO exposure

HECTOR complications occurred more often in patients receiving vs not receiving ECMO (36.9 percent vs 12.5 percent; p<0.001), specifically haemorrhagic events (24.6 percent vs 2.8 percent; p<0.001).

Among non-recipients, mortality risk from HECTOR complications was increased (adjHR, 1.12; p=0.029), due to haemorrhagic (adjHR, 1.57; p<0.001) but not thrombotic complications (adjHR, 1.03; p=0.869). Among ECMO recipients, overall HECTOR complication-related mortality risk was not increased (adjHR, 1.18; p=0.070), though there was an increased risk of death due to haemorrhagic complications (adjHR, 1.42; p<0.001) and a trend toward a lower risk of death due to thrombotic complications (adjHR, 0.78; p=0.049).

Take-home message

“Serious COVID-19 infection has been associated with thrombotic and haemorrhagic complications,” said Fanning. However, current data is limited due to small sample sizes from a few sites, primarily concerns thrombotic complications, is not specific to the ICU setting, and has limited information on risk factors.

“[This study showed that] HECTOR events are frequent complications of severe COVID-19 in ICU patients. Haemorrhagic, but not thrombotic complications, are associated with increased ICU-mortality,” he continued.

He acknowledged several limitations including the retrospective analysis and heterogeneity and missing data. “The degree to which HECTOR complications are a marker of high disease severity rather than direct contributor to ICU-mortality requires further investigation,” he added.