Chest Pain and Dyspnea? Taking Steroids? Think PE.

Abstract & Commentary

By Barbara A. Phillips, MD, MSPH, Professor of Medicine, University of Kentucky; Director, Sleep Disorders Center, Samaritan Hospital, Lexington. Dr. Phillips serves on the speakers bureau for PotomaCME.

Synopsis: Corticosteroid use is associated with an increased risk of symptomatic pulmonary embolism. The greatest risk is in the first 30 days of use and increases with increasing steroid dose.

Source: Stuijver DJ, et al. Use of oral glucocorticoids and the risk of pulmonary embolism: A population-based case-control study. Chest 2013;143:1337-1342.

The purpose of this study was to quantify the risk of symptomatic pulmonary embolism (PE) in patients taking corticosteroids. To address this question, the authors used a large database containing demographic details and complete medication histories of more than 2 million people in The Netherlands. Because virtually all patients in The Netherlands are registered with a single pharmacy, records for prescription drug use are essentially complete. This report is the result of a population-based, case-control study of this database. For purposes of this analysis, cases were defined as adult patients with a first hospital admission for PE between 1998 and 2008. Diagnosis of PE was objectively confirmed in more than 95% of these cases. Each case had up to four age- and sex-matched controls from the same database of patients who had not been hospitalized for PE.

For each patient, the investigators identified all prescriptions for oral (systemic) glucocorticoids, including cortisone, hydrocortisone, prednisone, prednisolone, triamcinolone, methylprednisone, dexamethasone, and betamethasone. Patients were classified as never users, current users, or former users on the basis of the timing of the prescription and the PE event. For current users, duration of current glucocorticoid use was categorized as recent use (PE diagnosed in the first 30 days of steroid use), intermediate use (PE diagnosed between 31 and 365 days of steroid use), and long-term use (more than a year of steroid use at the time of PE). The authors also calculated cumulative steroid dose at the time of PE. Daily dose equivalents of prednisolone 10 mg were as follows: cortisone 37.5 mg, hydrocortisone 30 mg, prednisone 10 mg, triamcinolone 7.5 mg, dexamethasone 1.5 mg, and betamethasone 1.5 mg. The analysis controlled for other risk factors for PE (trauma and fractures, malignancy, pregnancy, cardiovascular disease, diabetes, and surgery) and for medications that could reduce the risk of PE (vitamin K antagonists, heparins, and antiplatelet agents [i.e., aspirin, clopidogrel bisulfate, and dipyridamole]).

The analysis included 4495 patients who had PE and 16,802 controls. Overall, the mean age was 60 years, and women comprised 57% of both groups. Current use of glucocorticoids was more frequent among those patients who had PE (13.1%) than controls (2.5%). After adjustment for confounders, the risk estimate (odds ratio [OR]) for PE among corticosteroid users was 4.4 (95% confidence interval [CI], 3.8-5.0). Former users had a much lower increased risk of PE (cases, 8.4%; control subjects, 6.7%) with an adjusted OR of 1.2 (95% CI, 1.1-1.3).

Among those currently using corticosteroids, the risk of PE was highest within the first 30 days of glucocorticoid use (adjusted OR, 5.9; 95% CI, 2.3-7.9) and gradually decreased with longer duration of use. The adjusted OR was 1.9 (95% CI, 1.3-2.9) for long-term use (> 1 year). The association between glucocorticoids and PE varied depending on the dose of glucocorticoids used. A clear dose-response relationship was observed, with low-dose glucocorticoids carrying a two-fold increased risk of PE (adjusted OR, 1.8; 95% CI, 1.3-2.4) up to a 10-fold increased risk for the highest daily dose of glucocorticoids (adjusted OR, 9.6; 95% CI, 4.3-20.5). Almost always, the risk of PE was highest within the first 30 days of glucocorticoid use, irrespective of the dose.


Glucocorticoids have become one of the most widely prescribed medications, and are currently used by about 1% of the adult population.1 Despite their undeniable benefits, corticosteroids are accompanied by numerous side effects, including increased cardiovascular morbidity and mortality.2,3 Endogenous glucocorticoid excess among patients with Cushing syndrome has been associated with an increased incidence of venous thromboembolism (VTE) both during active disease and after surgery.4 But the effect of exogenous glucocorticoids on VTE risk is less clear. For one thing, it is difficult to separate any potential effect of corticosteroids on VTE risk from the risk resulting from the indication for which they are given. However, several in vitro studies have revealed that exogenous glucocorticoids enhance both synthesis and secretion of von Willebrand factor and plasminogen activator inhibitor-1, suggesting a direct activation of coagulation and inhibition of fibrinolysis.5-7

This large, population-based, case-control study has demonstrated that use of oral glucocorticoids was associated with a four-fold increased risk of PE both in a time-and dose-dependent fashion. The greatest risk was within the first 30 days of glucocorticoid use and for the highest daily dose. However, even among low-dose glucocorticoid users, patients were at the highest risk of PE within the first month after treatment onset compared with long-term users.

The authors acknowledge the fact that glucocorticoids are frequently prescribed in conditions that are inflammatory (and therefore thrombogenic), such as chronic obstructive pulmonary disease, arthritides, and cancer, which makes it very difficult to parse the precise role in the increased risk of PE in those who take steroids. At the very least, however, the current study should heighten our index of suspicion for PE in patients who are taking corticosteroids, especially early in the course and at high doses.


1. van Staa TP, et al. QJM 2000;93:105-111.

2. Souverein PC, et al. Heart 2004;90:859-865.

3. Wei L, et al. Ann Intern Med 2004;141:764-770.

4. Stuijver DJ, et al. J Clin Endocrinol Metab 2011;96: 3525-3532.

5. Heaton JH, et al. Mol Endocrinol 1989;3:185-192.

6. Huang LQ, et al. Blood Coagul Fibrinolysis 1995;6:438-445.

7. Morange PE, et al. Diabetes 1999;48:890-895.