Background:
The prevalence of venous thromboembolism (VTE) varies by race/ethnicity, with African-Americans displaying a significantly higher rate compared to Caucasians, and with Asian/Pacific Islanders and Hispanics experiencing the lowest rates. Genetic polymorphisms associated with thrombosis and molecular mutations in clotting factors do not explain the differential risk. There is a paucity of data to explain the reasons that underlie disparities in VTE prevalence among race/ethnic subgroups.
The MEGA study established that major illness is associated with increased risk of VTE. Socioeconomic factors such as income and healthcare access may also contribute to VTE prevalence. In this study, we examined whether medical or socioeconomic factors experienced by hospitalized patients contributed to the variation in VTE prevalence across race/ethnic subgroups. Such an analysis explores the extent to which race itself is a risk factor versus race serving as a surrogate for differential comorbidity or socioeconomic profiles that affect VTE prevalence.
Methods
We conducted a serial cross-sectional study using 2002-2014 data from the National Inpatient Sample, the largest all-payer inpatient database in the US, to examine the prevalence of VTE (both deep vein thrombosis (DVT) and pulmonary embolism (PE)) among non-pregnant, adult (18+ years) African Americans, Caucasians, Hispanics and non-Hispanic others (primarily Asian Americans). Patients hospitalized with a principal diagnosis of lower extremity DVT and/or PE, as determined by ICD-9-CM codes, were included in the cohort. Codes for superficial venous thrombi and upper extremity DVT were not included. We used the Elixhauser comorbidity algorithm to assess the contribution of comorbid conditions. Socioeconomic factors investigated included zip code-level household income quartile and insurance type.
VTE prevalence was defined as the number of cases per 1000 hospitalized patients. Prevalence of comorbid illness is reported as a percentage of total cases affected. To compare the role of comorbid conditions contributing to VTE with other thrombotic complications, we performed the same analysis in patients admitted with principal diagnosis of acute myocardial infarction (AMI).
Results
VTE prevalence was higher in patients aged 40 to 79 than those in the youngest (18-39) or oldest (80+) age groups (Figure 1). The prevalence of VTE was highest among African Americans, followed by Caucasians. Hispanics exhibited the lowest prevalence of VTE. Age did not affect this race/ethnic pattern.
Diagnosis of chronic anemia, hypertension, renal failure, heart failure and obesity were more frequent in the African American population (Table 1). Subjects with a diagnosis of AMI also had a majority of African Americans with these diagnoses yet the prevalence of AMI was higher in Caucasians and non-Hispanics (Figure 2), suggesting this chronic illness milieu is not the underlying driver for the increase in VTE.
Neither type of insurance (private, Medicare, Medicaid or underinsured) nor community-level household income level explain the differential prevalence of VTE across race/ethnic groups, suggesting that lack of resources does not skew the prevalence of VTE among hospitalized patients with regards to race.
Conclusions
Our nationally-representative study of hospitalized patients in the US confirms previously reported findings that African Americans exhibit a higher prevalence of VTE. In our study cohort, Hispanics displayed lowest prevalence. Number of comorbidities and type of comorbid conditions did not shed any light on the differences in VTE observed among different ethnic groups. Hemoglobinopathy status is thought to influence VTE risk particularly in the African American population. We excluded patients who had diagnoses of sickle cell disease and did not observe any differences in the reported prevalence pattern. We were unable to explore the contribution of sickle cell trait as this condition is rarely reported as a contributing admission diagnosis.
Insurance status and income level did not influence the likelihood of being admitted with VTE diagnosis.
Our findings do not suggest that race is acting as a surrogate for chronic illness or socioeconomic factors; however, this study is observational and largely descriptive in nature. To fully exclude this possibility, more sophisticated, inferential analyses are needed.
No relevant conflicts of interest to declare.
