Patients hospitalized for medical reasons who are at risk of developing venous thromboembolism (VTE) rarely receive either pharmacologic or nonpharmacologic prophylactic treatment to prevent the condition. These findings were published online Jan. 8 in Circulation: Cardiovascular Quality and Outcomes.
According to the study, hospitalized patients are at far greater risk of developing VTE than the general population (961 cases per 100,000 person years compared to 7.1 cases per 100,000 in the community). VTE is a major cause of death and disability. Although surgical patients are at higher risk than patients hospitalized for medical reasons, many more patients are hospitalized for medical reasons each year than are admitted for surgery; therefore, preventing VTE among medically ill patients could have a significant impact on the number of VTE cases annually.
The American College of Chest Physicians (ACCP) has long recommended VTE prophylaxis for patients with a range of medical conditions for whom the benefits of VTE prevention outweigh the risk of anticoagulant therapy, and issued guidelines in 2008 that recommended prophylactic anticoagulant therapy for patients admitted to hospital with congestive heart failure, severe respiratory disease, or who are confined to bed and have at least one additional risk factor for developing VTE. The guidelines also advised prophylactic treatment for most patients in critical care units and patients with ischemic stroke.
Senior researcher Gerry Oster, PhD, of Policy Analysis in Brookline, Mass., and colleagues sought to determine the frequency with which patients at risk of developing VTE receive prophylactic treatment. They used the Thomson Reuters Drug Database of hospital admissions extracted from the decision-support and cost-accounting systems of 150 U.S. hospitals and linked it to information from the Thomas Medstat MarketScan Database, a database of retail pharmacy claims from several private U.S. health plans insuring more than 10 million people.
From the admission database the researchers selected records of patients aged 40 and older who were admitted to the hospital between Jan. 1, 2003, and Dec. 31, 2008. They excluded patients who were admitted with VTE, pulmonary embolism and traumatic injury, as well as those who underwent major surgery. Oster et al then searched for eligible patients in the claims database, and linked information for those patients who were represented in both databases. After linkage, they excluded patients with incomplete data available, who had been hospitalized for any reason or treated for VTE within 30 days, or had a hypercoagulability disorder.
The researchers’ definition of VTE prophylaxis included graduated compression stockings and intermittent pneumatic compression devices, as well as prophylactic doses of low molecular weight heparin, unfractionated heparin, fondaprinux and warfarin at any dose. Patients were judged to have received VTE prophylaxis if they received any of these treatments within 48 hours of admission.
The study population comprised 49,948 patients with a mean age of 67 years. Contraindications to VTE prophylaxis were present in 7 percent of the patients. Upon index admission, more than 68 percent of patients had at least one risk factor for VTE and 54 percent had an ACCP indication for VTE prophylaxis.
Only 18 percent of at-risk patients received VTE prophylaxis within 48 hours of admission. The rate for patients with ACCP indications for prophylaxis also was 18 percent. Among patients without specified risk factors for developing VTE, 13.7 percent received prophylaxis.
Patients being admitted from nursing homes had the highest rate of VTE prophylaxis, at 31.6 percent, followed by patients with a history of VTE, at approximately 25 percent. Cancer patients received prophylaxis at a rate of 17.4 percent. Less than 2 percent of patients who received VTE prophylaxis during admission received post-discharge prophylaxis.
The researchers noted that other studies of VTE prophylaxis have contributed a range of results. The Multi-Centre Chart Audit of the Utilization of Risk Assessment and of Prophylaxis of Venous Thromboembolism in Acutely Ill Medical Patients in Canada (CHART) found only 23 percent of medically ill hospitalized patients in Canada received VTE prophylaxis and only 16 percent received appropriate prophylaxis.
Several other studies have reported higher rates; however, Oster et al pointed out that these studies were based on medical records and therefore researchers could assess immobility. In addition, studies finding higher levels of VTE prophylaxis were multinational and included many non-U.S. and Canadian hospitals, and a much higher percentage of studied admissions were at teaching hospitals.
The authors also noted that several studies have called into question the 2008 ACCP guidelines, suggesting that while VTE prophylaxis may be useful in reducing mortality and incidence of VTE in certain subgroups, it is associated with greater risk of bleeding and may bestow no net benefit. The ACCP revised those guidelines in 2012, and limited the recommendation for prophylaxis to patients at high risk of developing VTE according to the Padua Prediction Score, which considers a patient with a cumulative score of 4 or more to be at high risk. Active cancer, previous VTE, reduced mobility and known thrombolytic condition each confer 3 points. Recent trauma or surgery receives 2 points, and each of the following receives 1 point: 70 years of age or older; heart and respiratory failure; acute MI or ischemic stroke; acute infection and rheumatological disorder; obesity; and ongoing hormonal treatment. The new guidelines were issued after the current study was completed.
“[I]n defense of clinicians," the researchers concluded, "the nonspecific nature of the (2008) ACCP recommendations may be one factor inhibiting greater uptake of prophylaxis… The 2012 ACCP guidelines reflect a more cautious approach, distinguishing increased risk and low risk patients, and recommending anticoagulant prophylaxis only for the former group.”
The authors noted as limitations of their study that use of administrative databases can lead to errors, and that reliance on ICD-9 codes to assess the patients’ conditions renders the researchers unable to determine whether a condition was present on admission or developed during admission, leading to possible mischaracterization of a patient’s risk profile.
The study was sponsored by Pfizer and Bristol-Myers Squibb.