Baseline Hemoglobin Values and Clinical Outcomes in Acute Venous Thromboembolism: Insights From the RIETE Registry

Venous thromboembolism (VTE) is a major global health concern and a leading cause of morbidity and mortality. While anticoagulation effectively reduces the risk of recurrent VTE, it is associated with an inherent risk of bleeding. Identifying patient characteristics that influence these risks is critical for personalized management. Baseline hemoglobin (Hb) values have emerged as a potential prognostic marker [1-5], reflecting both underlying comorbidities and hemostatic alterations. However, the impact of Hb values on both thrombotic and bleeding complications in anticoagulated VTE patients remains insufficiently characterized.

We analyzed data from 111,646 patients with acute VTE included in the RIETE registry, a large multinational prospective cohort (ClinicalTrials.gov identifier: NCT02832245) [6]. Patients were stratified into Hb quintiles: < 11.4, 11.4–12.6, 12.7–13.6 (reference), 13.7–14.7, and > 14.7 g/dL. We assessed 90-day rates of recurrent VTE, major bleeding, arterial ischemic events (myocardial infarction, ischemic stroke, or limb amputation), and all-cause mortality. Multivariable logistic regression models were adjusted for demographic and clinical characteristics, including VTE risk factors, initial VTE presentation, comorbidities, renal function, leukocyte and platelet counts, anticoagulant strategy, and concomitant medications.

Patients with lower Hb values were more likely to be female (64%), older, and have a lower body mass index, while those in the highest Hb quintile were predominantly male (81%) and younger. The initial presentation of VTE varied, with presentation as PE becoming more frequent with increasing Hb values, whereas upper-limb DVT was more prevalent in the lowest Hb group (< 11.4 g/dL). Comorbidities such as active cancer, chronic heart failure, renal failure, and prior ischemic stroke were more common in patients with lower Hb values compared to those with higher values. Conversely, unprovoked VTE was more frequent in higher Hb quintiles.

The risk of adverse events declined progressively with increasing Hb values. In the lowest Hb quintile, 90-day rates of recurrent VTE, major bleeding, arterial ischemic events, and all-cause mortality were 2.2%, 4.1%, 0.5%, and 16%, respectively. In contrast, corresponding rates in the highest Hb quintile were 1.3%, 0.9%, 0.2%, and 3.2% (Table 1). Major bleeding and mortality rates were approximately five times higher in the lowest Hb quintile than in the highest, whereas VTE recurrence and arterial ischemic events were nearly twice as frequent. The major bleeding-to-VTE recurrence ratio was highest in the lowest quintile (1.86) and lowest in the highest quintile (0.49), indicating that bleeding risk outweighed thrombotic risk in anemic patients, whereas the opposite trend was observed in those with elevated Hb values.

TABLE 1. Outcomes during the first 3 months from VTE diagnosis, according to Hb values at baseline.

	< 11.4	11.4–12.6	12.7–13.6	13.7–14.7	> 14.7
Patients, N	23 387	21 614	21 681	22 467	22 497
Recurrent VTE	518 (2.2%)***	340 (1.6%)	304 (1.4%)	297 (1.3%)	296 (1.3%)
Recurrent DVT	260 (1.1%)***	174 (0.8%)	160 (0.7%)	149 (0.7%)	152 (0.7%)
Recurrent PE	258 (1.1%)***	166 (0.8%)	144 (0.7%)	148 (0.7%)	144 (0.6%)
Major bleeding	965 (4.1%)***	482 (2.2%)***	336 (1.5%)	260 (1.2%)***	202 (0.9%)***
Gastrointestinal	403 (1.7%)***	146 (0.7%)***	89 (0.4%)	58 (0.3%)**	50 (0.2%)***
Hematoma	167 (0.7%)***	128 (0.6%)**	85 (0.4%)	89 (0.4%)	55 (0.2%)**
Intracranial	94 (0.4%)*	71 (0.3%)	62 (0.3%)	53 (0.2%)	37 (0.2%)**
Retroperitoneal	64 (0.3%)***	51 (0.2%)**	24 (0.1%)	18 (0.1%)	22 (0.1%)
Other sites	237 (1.0%)***	86 (0.4%)	76 (0.4%)	42 (0.2%)**	38 (0.2%)***
Arterial ischemic events	127 (0.5%)***	85 (0.4%)**	51 (0.2%)	50 (0.2%)	47 (0.2%)
Ischemic stroke	82 (0.4%)***	53 (0.3%)	35 (0.2%)	34 (0.2%)	30 (0.1%)
Myocardial infarction	33 (0.1%)**	26 (0.1%)*	11 (0.05%)	11 (0.05%)	13 (0.06%)
Leg amputation	12 (0.05%)	7 (0.03%)	5 (0.02%)	5 (0.02%)	4 (0.02%)
All-cause death,	3668 (16%)***	1687 (7.8%)***	1090 (5.0%)	810 (3.6%)***	713 (3.2%)***
Pulmonary embolism	289 (1.2%)***	173 (0.8%)	142 (0.7%)	114 (0.5%)*	143 (0.6%)
Fatal initial PE	226 (1.0%)***	135 (0.6%)	116 (0.5%)	95 (0.4%)	116 (0.5%)
Fatal recurrent PE	63 (0.3%)***	38 (0.2%)	26 (0.1%)	19 (0.1%)	27 (0.1%)
Bleeding	183 (0.8%)***	78 (0.4%)	61 (0.3%)	37 (0.2%)*	33 (0.2%)**

• Note: Differences between patients with Hb levels 12.7–13.6 g/dL (reference group) versus other groups.
• Abbreviations: DVT, deep vein thrombosis; Hb, hemoglobin; PE, pulmonary embolism; VTE, venous thromboembolism.
• * p < 0.05.
• ** p < 0.01.
• *** p < 0.001.

Multivariable analysis confirmed that patients in the lowest Hb quintile had significantly increased risks of major bleeding (adjusted hazard ratio [aHR]: 1.73; 95% CI: 1.52–1.97), arterial ischemic events (aHR: 1.48; 95% CI: 1.06–2.07), and all-cause mortality (aHR: 1.51; 95% CI: 1.37–1.66) compared to the reference group, with no significant differences in recurrent VTE. Conversely, patients in the highest Hb quintile had significantly lower risk of major bleeding (aHR: 0.68; 95% CI: 0.57–0.81), with no significant differences in VTE recurrence or arterial ischemic events.

Subgroup analyses (Multipanel-Figure 1a–d) revealed important sex- and cancer-specific trends. In men, all outcomes declined steadily across increasing Hb levels. In women, mortality decreased across the first four Hb quintiles but plateaued in the highest Hb quintile, suggesting a leveling-off rather than a true U-shaped curve. We did not observe a significant interaction between sex and Hb values in multivariable models, though the visual trend warrants further investigation. Among patients with active cancer, bleeding and mortality rates were high across all quintiles, with mortality decreasing from 29% to 15% between the lowest and highest Hb groups. In contrast, in patients without cancer, bleeding, ischemic, and thrombotic risks declined in a more linear pattern.

FIGURE 1

Open in figure viewerPowerPoint

(a) Ninety-day outcomes according to hemoglobin values in 55 538 male patients with acute venous thromboembolism. (b) Ninety-day outcomes according to hemoglobin values in 56 108 female patients with acute venous thromboembolism. (c) Ninety-day outcomes according to hemoglobin values in 19 660 cancer patients with acute venous thromboembolism. (d) Ninety-day outcomes according to hemoglobin values in 91 986 noncancer patients with acute venous thromboembolism.

Although this analysis did not include hematocrit, mean corpuscular volume, or mean corpuscular hemoglobin (MCH), future RIETE analyses are planned to assess whether these hematologic parameters (particularly MCH, which may reflect iron-restricted erythropoiesis or anemia of chronic disease) modulate the observed mortality gradient, especially in the cancer population.

These findings carry several clinical implications. First, baseline Hb values provide independent prognostic information for bleeding, thrombotic, ischemic, and mortality outcomes in anticoagulated patients with VTE. Second, the shifting balance between bleeding and thrombotic risks across the Hb spectrum suggests that anticoagulant strategies may need to be tailored by baseline Hb. Anemic patients may benefit from bleeding-prevention strategies, whereas patients with elevated Hb values may require close monitoring for thrombotic complications. Finally, current guidelines provide tailored recommendations for patients with cancer, thrombocytopenia, or renal failure, but do not yet address Hb values. Our data suggest Hb deserves similar consideration.

In conclusion, baseline Hb values are strongly associated with 90-day clinical outcomes in acute VTE. Patients with low Hb values bear a disproportionate burden of adverse events. Incorporating Hb values into clinical risk models and future guidelines may help personalize anticoagulation and improve outcomes in this large patient population.