Prevalence of clinical venous thromboembolism in the USA: Current trends and future projections

Abstract

The U.S.-wide prevalence of venous thromboembolism (VTE) is unclear, with reported VTE incidence estimates varying widely. This retrospective analysis of healthcare claims data from patients in the Thomson Reuters national MarketScan® Commercial and Medicare databases (January 2002–December 2006) estimates the U.S. prevalence of VTE, and assesses associated temporal trends. Of 12.7 million study-eligible patients, 200,007 had VTE. The overall prevalence of VTE (cases per 100,000) increased by 33.1% during the study period: from 317 in 2002 to 422 in 2006. VTE was more prevalent in women than men throughout the study. The annual prevalence of VTE increased with age: 1,382 in patients ≥65 years versus 231 in patients <65 (2006 data). The number of U.S. adults with VTE is projected to more than double from 0.95 million in 2006 to 1.82 million in 2050. These data confirm that VTE remains a major healthcare burden in the US, particularly among the elderly, and highlight a continuing increase in prevalence of the disease. Greater efforts are required to improve awareness of VTE and improve standards of VTE prevention in healthcare organizations. The U.S.-wide prevalence of venous thromboembolism (VTE) is unclear, with reported VTE incidence estimates varying widely. This retrospective analysis of healthcare claims data from patients in the Thomson Reuters national MarketScan® Commercial and Medicare databases (January 2002–December 2006) estimates the U.S. prevalence of VTE, and assesses associated temporal trends. Of 12.7 million study-eligible patients, 200,007 had VTE. The overall prevalence of VTE (cases per 100,000) increased by 33.1% during the study period: from 317 in 2002 to 422 in 2006. VTE was more prevalent in women than men throughout the study. The annual prevalence of VTE increased with age: 1,382 in patients ≥65 years versus 231 in patients <65 (2006 data). The number of U.S. adults with VTE is projected to more than double from 0.95 million in 2006 to 1.82 million in 2050. These data confirm that VTE remains a major healthcare burden in the US, particularly among the elderly, and highlight a continuing increase in prevalence of the disease. Greater efforts are required to improve awareness of VTE and improve standards of VTE prevention in healthcare organizations. Venous thromboembolism (VTE), comprised of deep-vein thrombosis (DVT) and pulmonary embolism (PE), is an important healthcare risk in the US. In addition to the serious health consequences of VTE, thromboembolic disease is associated with a significant economic burden [1, 2]. It is important to understand the current and future prevalence of VTE in specific national populations to assess the clinical and economic burden of the disease. Reported incidences for DVT and PE vary widely due to differences in the study populations, diagnostic recording or classification of VTE events, and failure to include cases of outpatient or community-acquired VTE [3-5]. Even in studies of specific patient populations, such as those with cardiorespiratory disease or acute infections, reported estimates of VTE incidence vary widely due to different study methodologies, and in particular differing inclusion and exclusion criteria [6]. The incidence of VTE may also vary within the same type of patients due to different levels of disease-related risk. For example, in patients with cancer, the incidence of VTE depends on the type, location, and stage of the malignancy [7]. To date, few data exist regarding the temporal trends of VTE incidence [3, 8], or trends in VTE incidence according to age or gender [3, 4]. The majority of studies published to date were conducted on a community-wide scale and therefore do not provide information on a national basis covering a range of distinct patient populations [3-5, 8, 9]. This study determined the national VTE prevalence patterns in the U.S. population within a 5-year period from 2002 to 2006 based on real-world data from patients in the MarketScan® Commercial and Medicare databases. The data were also used to identify characteristics of patients with and without VTE, and extrapolated to produce projections of VTE prevalence up to 2050. A total of 12.7 million enrollees in the Commercial and Medicare MarketScan databases were eligible for the study. The mean age of the enrollees in the database was 45.7 years and 53.1% were women. Most enrollees were from the South region (43.5%), followed by North Central (23.8%), West (22.6%), Northeast (9.3%) regions, or unknown (0.8%). Patients were classified as having VTE if they had either a PE or DVT diagnosis on a single inpatient claim or on a single, nondiagnostic, outpatient claim with evidence of anticoagulant therapy. Using this criteria, 200,007 of the 12.7 million enrollees were identified as having had VTE; 139,769 (69.9%) had DVT, 47,424 (23.7%) had PE and 12,814 (6.4%) had both DVT and PE. The mean age of VTE patients was 62.7 years; overall 54.4% were women. The most common types of insurance plans were point-of-service with capitation (35.3%), indemnity (33.3%), and health maintenance organizations (22.7%). During the 5-year study period, the annual prevalence of VTE changed from 317 cases per 100,000 patients in 2002 to 422 cases per 100,000 patients in 2006, representing an overall increase of 33.1% (Fig. 1A). The observed prevalence of VTE was consistently higher in women than in men throughout the study, although this difference between the genders was reduced by 2006 (Fig. 1A). The prevalence of VTE in the population increased with age (Fig. 1B). There was little change in prevalence of VTE for the 18–29 years age group, but all other age groups showed increasing prevalence during the study period. The highest increase in prevalence and the greatest overall rise in prevalence over the study period were observed in the oldest age group (>85 years). Prevalence of VTE overall and according to (A) gender and (B) age in the U.S. population between 2002 and 2006. Patients with VTE were subsequently matched to controls without VTE based on age, gender, region, and months of preperiod enrollment. Patients with VTE were significantly more likely to have comorbid conditions such as heart failure, pulmonary disease or infection, or active cancer compared to patients without VTE (P < 0.001 for all comparisons) (Table I). Patients with VTE had a threefold increase in the risk of heart failure compared to matched patients without VTE (adjusted odds ratio [OR]: 3.23, 95% confidence interval [CI]: 3.15–3.32). Similarly, patients with VTE were six times more likely to be diagnosed with pulmonary disease or infection (adjusted OR: 6.40; 95% CI: 6.29–6.50). The annual population growth rate in the USA is estimated to be 0.975% [10], and will be accompanied by marked aging of the population. We applied the observed population rates of VTE by age and gender to the projected U.S. census population at various time points. We estimate that the total number of adults with VTE in the US will more than double in the next 40 years, increasing from 0.95 million in 2006 to 1.82 million in 2050 (Fig. 2). Prevalence and projected estimates of VTE prevalence in the U.S. population (2002–2050). This study suggests that the incidence of VTE, coupled with an aging U.S. population, increasingly presents a major challenge in the US to patients, providers, and payers. Despite a large number of clinical studies on VTE prevention and treatment, only few studies have reported the incidence or prevalence of the disease, and in these studies the VTE rates vary widely [3-5]. Incidences of between 43.7 and 145.0 cases per 100,000 for DVT [4, 5] and from 20.8 to 65.8 per 100,000 for PE [4, 5, 11, 12] have been previously suggested. The higher prevalence rate reported in our analysis may be partly accounted for by the differences in the way that VTE events are identified from patient records (clinical diagnostic records compared to the analysis of healthcare service codes in this study). The positive and negative predictive value of using ICD-9 codes from administrative data to identify VTE events has been examined, with mixed results. Of particular concern are the findings of White et al. [13-15] in a 2009 study which found a high level of false-positives. The finding that the prevalence of VTE increases with age has been consistently reported [3-5]. In this study, patients with VTE were significantly more likely to be diagnosed with comorbidities such as heart failure and pulmonary disease compared with patients without VTE, and the comorbid conditions selected for inclusion in this analysis were all well-known predictors of VTE [16, 17]. However, the VTE risk associated with these comorbid conditions may also relate to hospitalization, as these patients are more frequently hospitalized than patients without these conditions. Our large patient population was derived from healthcare claims data in a national database that included a variety of payers and plan benefit designs. As such the current study represents an advance in published literature which until now comprised smaller studies that were conducted at a local level [3-5]. Moreover, the reported age/gender specific prevalence data and projections are based on a larger population sample size than in previously published studies. Another important aspect is that the VTEs recorded in this analysis are all clinical events (i.e. symptomatic, diagnosed and confirmed) that required healthcare interventions. Clinical trials, by contrast, often report asymptomatic VTE as an outcome measure. Recording and reporting symptomatic events increases the clinical and economic relevance of prevalence data, as the DVT and PE events included in this study all required treatment interventions. This study also has limitations because it involved a national population sample derived from a healthcare claims database. Although the study population was large and included a diverse range of patients, it was not a random population sample. The patient sample was derived from the data of patients with employer-funded private insurance, which may not be representative of the whole U.S. population. The study population did not include patients without health insurance, with Medicaid, or those with military-based health insurance programs and therefore the rates of VTE may not generalize. Furthermore, as this study was derived from a real-world population, there may have been differences in the risk profiles of patients between different time periods, which could have influenced the prevalence rates observed. Another methodological limitation was no differentiation between new and recurrent cases of VTE in the sample population. However, the main aim of the study was to assess prevalence and trends in prevalence of VTE within the general population. Prevalence may be defined as the proportion of individuals within the population at any point in time who suffer from the disease, and should be distinguished from incidence, which describes the number of new diagnoses (of VTE) with a specific period of time. VTE is often a clinically silent disease, and the initial presentation of fatal PE may be sudden death. This may contribute to an underestimation of the prevalence of PE because in such cases autopsy is required to confirm the cause of death, yet autopsy rates are low in the US [4]. This study suggests that VTE will be an increasing burden on the U.S. healthcare system over the next 40 years, due in part to increased incidence of the disease but more notably due to the aging of the population. The data presented are the first estimates of VTE prevalence in the US based on a national population sample. The increasing prevalence of VTE will translate into a growing clinical and economic burden. National hospital VTE performance measures and the use of optimal thromboprophylaxis regimens [18, 19] should reduce the number of cases of preventable VTE. Greater awareness is required of the risk of VTE and of the preventative measures that can be taken. The study population was drawn from the MarketScan® Commercial Claims and Encounters (CCAE) and Medicare Supplemental and Coordination of Benefits databases from Thomson Reuters between January 1, 2002 and December 31, 2006. The Commercial database contains individual healthcare data for patients whose healthcare is provided under fee-for-service and managed care health plans. The Medicare database contains information on individuals who have supplemental Medicare insurance paid for by employers. By utilizing these two databases, which have been widely used in other studies [20-22], the study was able to profile a larger geographically diverse cross-section of insured US residents. Enrollees were included in the study if they were aged 18 years or older, and enrolled in the Marketscan database at any point during the study window (2002–2006). Patients were identified as having VTE if they had a single VTE diagnosis (ICD-9-CM codes 415.1x, 451.0x, 451.1x, 451.2x, 451.81, 451.9x, 453.4x, 453.8x, or 453.9x), or at least one outpatient nondiagnostic claim with a VTE diagnosis as stated above and subsequent evidence of an anticoagulant prescription within 90 days. Patient records were assessed each calendar year to determine whether a diagnosis of VTE was recorded. The first date of VTE diagnosis was counted, but if the outpatient diagnosis claim fell in one calendar year and there was evidence of anticoagulant use in the next, then the patient was counted in both years. Age- and sex-specific VTE prevalence rates were calculated by dividing the number of VTE cases identified in the Commercial and Medicare databases between January 2006 and December 2006 (age 18 and older) by the number of individuals in the underlying MarketScan populations during that time. These rates were then multiplied by 100,000 to obtain a VTE prevalence rate per 100,000. Given changes in the age gender distribution of the database over time, rates were standardized across age and gender using U.S. Census bureau estimates for that year. No differentiation was made between cases of new and recurrent VTE. The resulting prevalence rates were applied to the corresponding age and gender population estimates from the U.S. Census Bureau to project the total number of individuals in the US expected to have VTE in 2006 and 2008 and thereafter in 5-year increments between 2010 and 2050. The presence of selected conditions known to be associated with VTE or conditions that put patients at increased risk for VTE were recorded based on ICD-9-CM coding, both for patients with VTE and those without VTE. For this analysis control patients aged 18 years or older, with no evidence of a VTE diagnosis but at least one nondiagnostic service claim during the study period, were matched (2:1) to VTE patients according to age, sex, region, and months of preperiod continuous enrolment to allow identification of clinical characteristics associated with presence of VTE. These included heart failure, coronary artery disease, acute coronary syndromes (including acute myocardial infarction), stroke, respiratory disease and infection, diabetes, active cancer, inflammatory bowel disease, nephrotic syndrome, varicose veins, recent surgical procedures (within 30 days), trauma, and limb ischemia [18]. Key demographic and clinical characteristics were measured at the index date, and differences in clinical characteristics between patients with and without VTE were compared using t-tests for continuous variables and chi-square tests for categorical values. Logistic regression models were used to assess the relative probability of patients with VTE having the recorded risk factors, compared with patients without VTE. This study was funded by Sanofi-Aventis US, Inc. The authors received editorial and writing support in the preparation of this manuscript, which was funded by sanofi-aventis, Bridgewater, NJ. Dan Bridges, PhD, provided the editorial/writing support. The lead author was fully responsible for the content and editorial decisions relating to this manuscript and received no form of compensation related to the development of the manuscript. S. B. Deitelzweig*, B. H. Johnson†, J. Lin‡, K. L. Schulman†, * Ochsner Clinic Foundation, New Orleans, Louisiana, † Healthcare, Thomson Reuters, Cambridge, Massachusetts, ‡ Sanofi-Aventis, Bridgewater, New Jersey.