An analysis of the thromboembolic outcomes of 2472 splenectomized individuals

To the editor:

The diseases associated with the reason for splenectomy play an important role in the rate of infections in splenectomized patients.¹  Our hypothesis was that the same reasoning would apply to the rates of venous thromboembolism (VTE) after splenectomy and that these findings could inform future preventive care.

All patients undergoing a splenectomy in Victoria, Australia, were identified by using linked hospital discharge data with International Classification of Diseases, 10th Revision, Australian Modification (ICD-10-AM) diagnostic codes.²  The sampling frame included patients between July 1, 1998, and December 31, 2006, who were ≥15 years of age. This sampling frame was selected to ensure adequate numbers of patients in each of the splenectomy indication groups for statistical power. Splenectomy indications were divided into 6 mutually exclusive groups, and patients who had multiple indications were ordered hierarchically: (1) trauma, (2) therapeutic malignant (planned; malignant disease such as lymphoma or leukemia), (3) therapeutic hematologic (planned; hematologic disease such as idiopathic thrombocytopenic purpura [ITP] or hemolytic anemia), (4) therapeutic other (local infections or congenital abnormalities of the spleen), (5) iatrogenic malignant (unplanned; unintended accompaniment to surgery for malignant disease), and (6) iatrogenic nonmalignant (unplanned; unintended consequence of surgery for nonmalignant disease).

VTE was divided by using ICD-10-AM codes into lower-extremity acute deep vein thrombosis (DVT), pulmonary embolism (PE), and portal vein thrombosis (PVT). This classification was based on studies using similar coding methodologies.³  The groups were not mutually exclusive. Patients who had VTE that occurred in the first 30 days after splenectomy were excluded to avoid inclusion of complications potentially resulting from the primary surgery or related admission. Incidence rates of first VTE were calculated for sex, age group, and indication for splenectomy. Multivariate Cox proportional hazards regression models were fitted to compute hazard ratios adjusted for age, sex, and indication for splenectomy. Hazards proportionality was assessed by using analysis of scaled Schoenfeld residuals. All reported P values were two-tailed and, for each analysis, P < .05 was considered significant. We used Stata, version 131.0 (STATA, College Station, TX) for analysis. The Victorian Department of Health granted ethics approval for this study.

In all, 2472 patients underwent splenectomy over the 8-year period with 8236 person-years of follow-up (mean follow-up, 3.3 years per person). Of those patients, 1147 (46.4%) were female and 1525 (61.7%) were aged ≥50 years. The indications for splenectomy are described in Table 1.

A total of 142 splenectomized patients (5.74%) had a VTE requiring hospitalization; the incidence of first VTE was 1.30 per 100 person-years (95% confidence interval [CI], 1.10 to 1.53) (Table 1). Sites of VTEs included lower-extremity DVT (95; 66.43%), PE (60; 41.96%), and PVT (11; 7.69%).

Age was the most important risk factor for VTE, with a hazard ratio of 2.76 when splenectomized patients younger than age 50 years were compared with those age 50 years or older.

There was no statistical difference in the rate of VTE according to indication for splenectomy. When patients splenectomized for trauma were compared with patients who had other indications, there was no difference in hazard of VTE (Table 1).

Our VTE cumulative incidence of 5.74% is consistent with that of other large studies of splenectomized patients that used similar methodologies; Boyle et al³  found cumulative incidence of 4.4% for DVT and PE and 1.7% for PVT in 1762 splenectomized patients with ITP. Our VTE rate of 1.30 per 100 person-years is similar to that in a Danish study with a rate of VTE between 1.03 and 1.90 per 100 person-years.⁴ 

The finding of no significant difference in the rate of VTE between those splenectomized for trauma and those splenectomized for therapeutic hematologic reasons (such as ITP) differs from the findings of previous studies.³  Boyle et al³  reported that patients splenectomized for ITP have higher rates of VTE than nonsplenectomized ITP patients. Our findings suggest that splenectomy per se might increase the risk of VTE above that which is posed by the indication for splenectomy.

The weaknesses of this study are the lack of a disease-specific control group, the broad splenectomy indication categories, and the possibility that VTE may have been missed if patients were not hospitalized.

The strength of this study is the large number of splenectomized patients. This study provides new evidence that the risk of VTE following splenectomy does not differ between patients splenectomized for trauma and those splenectomized for hematologic reasons. These findings may have implications for practice regarding targeted administration of thromboembolic prophylaxis.