Table 3

Multivariate analysis for the primary and secondary endpoints according to the presenting features in 300 patients with low-risk ET followed with observation or treated with antiplatelet therapy

Total thrombosis
Arterial thrombosis
Venous thrombosis
Major bleeding
IRR (95% CI)PIRR (95% CI)PIRR (95% CI)PIRR (95% CI)P
Age older than 40 y 1.7 (0.8-3.8) .2 1.7 (0.6-4.4) .3 1.7 (0.4-6.9) .4 1.4 (0.4-4.9) .6 
Male sex 1.9 (0.9-3.9) .09 2.8 (1.5-7.0) .02 0.85 (0.2-3.4) .8 1.1 (0.3-3.4) .9 
Cardiovascular risk factors 1.7 (0.8-3.7) .2 2.6 (0.98-6.9) .056 0.7 (0.2-2.9) .7 1.0 (0.3-3.3) .9 
Hemoglobin level greater than 140 g/L 1.1 (0.5-2.3) .8 0.6 (0.3-1.6) .3 3.0 (0.8-11.4) .1 0.6 (0.2-2.0) .4 
Leukocyte count greater than 8.7 × 109/L* 0.8 (0.4-1.8) .6 0.95 (0.4-2.3) .9 0.7 (0.2-3.1) .6 3.8 (0.8-17.8) .09 
Platelet count greater than 800 × 109/L* 1.5 (0.7-3.2) .3 1.6 (0.65-4.0) .3 1.0 (0.3-4.1) .9 10.6 (1.6-69.3) .01 
Absence of antiplatelet therapy 0.9 (0.4-1.8) .7 0.6 (0.2-1.6) .3 1.7 (0.5-6.2) .4 — — 
Antiplatelet therapy — — — — — — 1.9 (0.6-6.0) .3 
Total thrombosis
Arterial thrombosis
Venous thrombosis
Major bleeding
IRR (95% CI)PIRR (95% CI)PIRR (95% CI)PIRR (95% CI)P
Age older than 40 y 1.7 (0.8-3.8) .2 1.7 (0.6-4.4) .3 1.7 (0.4-6.9) .4 1.4 (0.4-4.9) .6 
Male sex 1.9 (0.9-3.9) .09 2.8 (1.5-7.0) .02 0.85 (0.2-3.4) .8 1.1 (0.3-3.4) .9 
Cardiovascular risk factors 1.7 (0.8-3.7) .2 2.6 (0.98-6.9) .056 0.7 (0.2-2.9) .7 1.0 (0.3-3.3) .9 
Hemoglobin level greater than 140 g/L 1.1 (0.5-2.3) .8 0.6 (0.3-1.6) .3 3.0 (0.8-11.4) .1 0.6 (0.2-2.0) .4 
Leukocyte count greater than 8.7 × 109/L* 0.8 (0.4-1.8) .6 0.95 (0.4-2.3) .9 0.7 (0.2-3.1) .6 3.8 (0.8-17.8) .09 
Platelet count greater than 800 × 109/L* 1.5 (0.7-3.2) .3 1.6 (0.65-4.0) .3 1.0 (0.3-4.1) .9 10.6 (1.6-69.3) .01 
Absence of antiplatelet therapy 0.9 (0.4-1.8) .7 0.6 (0.2-1.6) .3 1.7 (0.5-6.2) .4 — — 
Antiplatelet therapy — — — — — — 1.9 (0.6-6.0) .3 

ET indicates essential thrombocythemia; and IRR, incidence rate ratio.

*

Cut points for leukocyte count greater than 10 × 109/L and platelet count greater than 1000 × 109/L were also explored by multivariate analysis but yielded no relevant change in the incidence rate ratios (data not shown).