Skip to content

Prothrombin Fragment 1.2 Predicts Thrombosis in Hospitalized Patients with COVID-19

Key findings

  • This study evaluated prothrombin fragment 1.2 (PF1.2) in 115 patients hospitalized with COVID-19
  • PF1.2 was elevated in most patients, but synchronously measured D-dimer was almost universally elevated
  • Random measurements of PF1.2, but not random measurements of D-dimer, was significantly associated with thrombosis in multivariable analyses
  • The optimal threshold for PF1.2 to identify patients who developed either venous thromboembolism alone or any thrombotic manifestation was greater than or equal to 523 pmol/L

Numerous studies have reported hypercoagulability in critically ill patients with COVID-19, and this complication can be fatal or lead to considerable morbidity. Researchers at Massachusetts General Hospital have reported that D-dimer elevation at admission predicts bleeding, thrombosis, critical illness and death in patients with COVID-19.

Now, expanding on those previous findings, Hanny Al-Samkari, MD, a classical hematologist and clinical investigator, at the Center for Hematology at the Mass General Cancer Center, and Rachel Rosovsky, MD, of the Center for Hematology at the Mass General Cancer Center, and colleagues report that prothrombin fragment 1.2 (PF1.2) was more discriminant than D-dimer in identifying thrombosis in hospitalized patients with COVID-19. Their findings appear in the American Journal of Hematology.

Study Details

The study involved 115 patients hospitalized with COVID-19 at Mass General who had PF1.2 measured at any time during their hospitalization between April 1 and May 6, 2020. The data cutoff date was May 27, 2020. Analyses comparing PF1.2 and D-dimer compared values measured on the same day, usually in the same blood draw.

In addition to venous thromboembolism (VTE), clinically significant non-vessel thrombotic manifestations were evaluated (clotting of a line or hemofiltration circuit).

Incidence of Thrombosis

Over the median follow-up of 29 hospital days, 56 patients (49%) developed VTE or a non-vessel thrombotic manifestation. For further analyses, 37 patients who were on therapeutic anticoagulation were excluded.

PF1.2 and D-dimer

Overall, PF1.2 and D-dimer were moderately positively correlated (r = 0.542, P < .001). At the individual patient level, though, there was a considerable discrepancy between the measurements, with an elevation of D-dimer in 99% and elevation of PF1.2 in 59%.

Other Analyses

  • The optimal threshold for PF1.2 to identify patients who developed either VTE alone or any thrombotic manifestation during hospitalization was >523 pmol/L
  • On multivariable analysis, PF1.2 >500 pmol/L was associated with VTE (OR, 4.26; 95% CI, 1.12–16.21; P =.03) and any thrombotic manifestation (OR, 3.85; 95% CI, 1.39–10.65; P = .01)
  • There was no significant association between D-dimer elevation and thrombosis on multivariable analysis
  • 91% of patients who had PF1.2 within the normal reference range did not develop VTE during their hospitalization

Effect of Anticoagulation

Four patients had multiple PF1.2 measurements obtained close in time to a radiographically confirmed VTE event. In the days leading up to the diagnosis of VTE, PF1.2 values trended upward. After anticoagulation was initiated, PF1.2 values rapidly normalized, whereas D-dimer values declined but did not normalize.

Guidance for Patient Care

The PF1.2 measurement obtained at hospital admission for COVID-19 may be as good or better than D-dimer in predicting VTE and thrombotic manifestations. For patients at high risk of thrombosis who are not receiving therapeutic anticoagulation, it may be useful to watch the trend in PF1.2.

of hospitalized COVID-19 patients exhibited D-dimer elevation

of hospitalized COVID-19 patients exhibited PF1.2 elevation

greater risk of thrombosis in hospitalized COVID-19 patients with PF1.2 >500 pmol/L

of COVID-19 patients with normal PF1.2 did not develop VTE

View all COVID-19 updates

Refer a patient to the Mass General Cancer Center

Related topics


In this Q&A, Rebecca Karp Leaf, MD, and Hanny Al-Samkari, MD, of Massachusetts General Hospital, give us a rundown of what is known to date about the correlation between blood type and COVID-19 transmission/severity, findings from their study and why this connection may or may not exist.


A substantial rate of bleeding in patients with COVID-19 suggests caution with routine escalation of anticoagulation beyond standard doses.