By Richard R. Watkins, MD, MS, FACP

Associate Professor of Internal Medicine, Northeast Ohio Medical University; Division of Infectious Diseases, Cleveland Clinic Akron General Medical Center, Akron, OH

Dr. Watkins reports that he has received research support from Actavis.

SYNOPSIS: A case-control study observed an increased risk for developing community-acquired Staphylococcus aureus bacteremia with the use of systemic glucocorticoids. A distinct dose-response relationship was found.

SOURCE: Smit J, Kaasch AJ, Sogaard M, et al. Use of glucocorticoids and risk of community-acquired Staphylococcus aureus bacteremia: A population-based case-control study. Mayo Clin Proc 2016;91:873-880.

Staphylococcus aureus bacteremia (SAB) continues to cause significant morbidity and mortality. Previous studies produced conflicting results about whether steroids increase the risk of SAB and may have been biased by confounding variables. Therefore, Smit and colleagues sought to more clearly define the risk of SAB associated with the use of steroids.

The investigators used data collected from a population-based registry from northern Denmark. They chose to limit the study to patients with community-acquired SAB (CA-SAB), defined as no previous diagnosis of SAB within five years, because patients with previous SAB are at increased risk for recurrences compared to the general population. Current steroid users were defined as patients whose most recent prescription redemption was within 90 days before the index date. This group was further characterized into new users, defined as those who redeemed their first-ever prescription within 90 days of the index date, and long-term users who redeemed their prescription within 180 days. Former users were defined as those patients whose last prescription was 90 to 180 days before the index date, and nonusers were persons who filled no prescriptions for steroids within 180 days of the index date. Prescriptions for inhaled and topical steroids were excluded. Risk for CA-SAB was further assessed among current steroid users according to prednisolone-equivalent cumulative doses, from 150 mg or less up to greater than 1,000 mg. Conditional logistic regression was used to calculate crude and adjusted odds ratios (ORs).

A total of 2,638 patients with CA-SAB were identified, of whom 379 were current users of steroids, along with 26,379 randomly selected controls (1:10 ratio), which included 827 current steroid users. Only 5% of the isolates were methicillin-resistant Staphylococcus aureus (MRSA). Compared to controls, patients with CA-SAB had more co-morbidities, including diabetes (18% vs. 5%), chronic pulmonary diseases (14% vs. 6%), and cancer (25% vs. 7%). The adjusted OR for CA-SAB among new steroid users was 2.73 (95% CI, 2.17-3.45), 2.31 (95% CI, 1.90-2.82) for long-term users, and 1.33 (95% CI, 0.98-1.81) for former users. Also, the risk of CA-SAB increased with escalating doses of steroids. Compared to nonusers, the adjusted OR ranged from 1.32 (95% CI, 1.01-1.72) for those taking a cumulative dose ≤ 150 mg, up to 6.25 (95% CI, 4.74-8.23) for those with a cumulative dose > 1,000 mg. Finally, there were no significant differences in the risk of CA-SAB based on sex or age.


The results of this study showed there was a considerable risk of developing CA-SAB for patients in the cohort who were taking steroids. Furthermore, the risk increased with higher steroid doses. Smit and colleagues were astute for excluding cases of hospital-acquired SAB, which likely would have introduced confounding variables, such as post-surgical wound infections and SAB caused by intravenous catheters. Although the prevalence of MRSA bacteremia was very low compared to that of the United States, there is no logical reason to suspect that a higher rate of MRSA would have led to any different outcomes.

As the authors mention, there are several pathophysiologic mechanisms by which steroid use might predispose to CA-SAB. For example, steroids inhibit toll-like receptor signaling, a key component in the innate immune response to S. aureus infections. Furthermore, steroids reduce phagocytosis, cytokine production, and leukocyte adhesion. Previous studies have demonstrated that the adverse effects of steroids on immunity are dependent on the length of therapy and the dosage used. Wound healing also is impaired by steroids, and this loss of skin barrier protection can allow S. aureus to enter deeper tissues and, ultimately, the bloodstream.

Despite the interesting findings of the study, a few words of caution are necessary. First, there were no data on steroid adherence among the patients or on infective foci, such as vascular access devices. Second, patients who use steroids tend to be sicker than the general patient population and, thus, more likely to be susceptible to CA-SAB. Third, the study was conducted in northern Denmark, and the results might not be generalizable to other populations and geographic areas.

What is the take-home message from this study? Clinicians should maintain a high level of suspicion for CA-SAB in patients taking steroids and focus accordingly, i.e., blood cultures and a thorough history and physical examination, with particular attention to joint or back pain and new heart murmurs that might indicate an infected joint, discitis, or endocarditis.