Sorry, you need to enable JavaScript to visit this website.
Skip to main content

Empiric Steroids: the Good, the Bad, and the Ugly

Edward D. Harris, Jr., MD | September 1, 2008
View more articles from the same authors.

Case Objectives

  • Describe appropriate indications for use of steroids.
  • Appreciate contraindications for use of empiric steroids.
  • Understand which chronic illnesses and systemic diseases can benefit from glucocorticoid use.

Case A: Part 1


A 55-year-old male immigrant with no significant past medical history came to the emergency department for evaluation of a 3-day history of dyspnea, nonproductive cough, and fever. The physical examination revealed tachypnea, bibasilar crepitations, and diffuse, end expiratory wheezes. The patient was started empirically on a cephalosporin and a macrolide for community-acquired pneumonia (CAP). Given the diffuse wheezing on exam, he was also treated with intravenous corticosteroids.

Comment: The diagnosis of CAP in patients hospitalized for severe symptoms is often made without evidence for a causative agent. Antibiotic coverage is initiated, and, because many of these patients have chronic obstructive pulmonary disease (COPD), glucocorticoids are often added. In a prospective, randomized study of giving 40 mg prednisolone intravenously for 3 days in addition to antibiotics to one cohort of moderate–severe CAP patients, and antibiotics alone to a second cohort, the glucocorticoids were reported to promote resolution of symptoms and reduce the duration of intravenous antibiotic therapy.(1) Similarly, a retrospective, observational study of 546 patients with CAP found that mortality decreased in the patients with severe CAP who were given both systemic steroids and antibiotics.(2) Other rationales for glucocorticoid therapy in severely ill patients with pneumonia are that many of them have "critical illness–related corticosteroid insufficiency," and in vitro studies indicate that alveolar cell apoptosis and activation of proteases induced by Legionella pneumophila are inhibited by methylprednisolone.(3)

Among the many questions raised by this brief case report is a crucial one: How aggressively should physicians search for a specific etiologic agent in presumed CAP before instituting therapy? One point to support a search is that there are more and more relatively specific therapies for parasitic diseases often found in immigrant populations, in addition to tuberculosis. Strongyloides stercoralis, for example, responds to treatment with thiabendazole and ivermectin, and various studies indicate that glucocorticoid therapy is not helpful in pulmonary S. stercoralis. Another answer is that when there are risk factors for "atypical" causes, such as the patient being an immigrant, a closer look is indicated. In patients such as this one, the use of sustained glucocorticoids for "expiratory wheezing" is questionable without a better sense of the causative agent.

Case A: Part 2

Over the subsequent days, the patient deteriorated, developing respiratory failure and requiring transfer to the intensive care unit. The diffuse wheezing persisted. Subsequent chest radiographs and computed tomography (CT) scans revealed progressive bilateral diffuse granular opacities. A more extensive work-up pursued to evaluate for atypical and fungal pneumonia revealed no culprit organism. Bronchoscopy revealed thick yellow mucus, and cultures were sent. Ultimately, bronchial washings demonstrated the larvae of S. stercoralis.

Let us hope that the proper therapeutic regimen (including antiinfectives and discontinuing the glucocorticoids) helped this man. The principal message for this and other cases is that glucocorticoids (don't use the term "corticosteroids" because that word includes the sex hormones) should be thought of as being essentially for acute use in whatever dose is needed, with plans to taper or replace with other medications. The exceptions must be diseases that are chronic and prolonged for which there are no more therapeutic options.

Whenever you are initiating glucocorticoid therapy empirically, and you aren't completely certain of the underlying illness, ask yourself, "Are there any conditions remaining on the differential diagnosis that could get much worse on glucocorticoids?" If the answer is yes, consider withholding the steroids, empirically treating the other condition, or pursuing more aggressive diagnostic testing to rule out the condition.

Case B: Part 1

A 49-year-old woman had the onset over 3 weeks of morning stiffness, pain, and swelling bilaterally in her proximal interphalangeal (PIP) joints, metacarpophalangeal (MCP) joints, wrists, and knees. She is postmenopausal, has an unremarkable past medical history and no allergies, and her only medications are a statin drug for high low-density lipoprotein (LDL) and gabapentin for sleep. She has not left her home for the past 6 months. She is fatigued and finds that she cries more easily. She sees you, her primary care physician, 3.5 weeks after symptoms began.

Examination reveals bilateral tenderness and soft tissue swelling in the finger joints, except for the distal interphalangeal (DIP) joints and the palmar aspect of the wrists. Both knees are swollen, with a particularly large effusion on the right. A right popliteal cyst is present, and full extension lacks 20 degrees. MCP joints 1–3 are tender on the right foot. Full extension of the left elbow is limited. No subcutaneous nodules or skin rash is found. Eye exam is normal, as is examination of the cardiorespiratory system.

While you await the results of laboratory tests, would it be appropriate to begin treatment with 15 mg prednisone each morning because of concern about the right knee?

Comment: All aspects of this case suggest a diagnosis of early rheumatoid arthritis (RA). The morning stiffness is a sine qua non for joint inflammation. The reasons for the bilaterality of RA are not definitively clear, but the leading hypothesis links this to inflammatory neuropeptides. Her joint distribution is typical: DIP joints are rarely affected in RA.

While awaiting lab tests and arriving at a definitive diagnosis, it is inappropriate to start therapy with prednisone, 15 mg daily. The initiation of this low-dose therapy is likely to lead to an improvement in the patient's symptoms but may lead to a conflict later: the relatively low dose is "forgotten" by the physician, but the patient cannot bear stopping it. Since the 1950s, when cortisone was first used to treat RA, there has been a tendency among many physicians, once the diagnosis of RA is definite, to use oral, daily glucocorticoids. This is now recognized as inappropriate, especially since the efficacy of methotrexate, with or without biologic agents, has clearly been demonstrated. If the daily dose were to be continued for more than a month, this postmenopausal woman would be at a high risk for glucocorticoid-induced osteoporosis.(4) Early intervention could include nonsteroidal antiinflammatory drugs, education about joint protection, and consideration of medical treatment for reactive depression. In addition, tetracycline derivatives (e.g., minocycline and doxycycline) have been found to have small but definite efficacy in early RA.(5)

Although oral daily glucocorticoids are contraindicated at this point, it would be appropriate to aspirate as much fluid as possible from the patient's right knee and inject 40 mg triamcinolone (an injectable glucocorticoid) or an equivalent. This route of administration is approximately twice as effective as intramuscular injection of similar doses (6), and the aspiration of fluid itself alleviates symptoms. The fluid should be sent for cell count and culture.

Case B: Part 2

Let's assume that glucocorticoids were not initiated. The joint fluid contained 23,000 polymorphonuclear leukocytes (PMNs)/mm3 and culture was negative. The patient's hemoglobin was 10.2 g/dL, plasma white blood cell (WBC) count was 9500/µL with 3% eosinophils and 80% PMNs. Urinalysis and a comprehensive serum screen were normal. The C-reactive protein was 6.8 mg/dL, and both rheumatoid factor and anticitrullinated antibody (highly specific for RA) were present in serum. Now, almost 5 weeks after symptoms began, the patient is not improved and probably is worse.

Should oral glucocorticoids be used empirically now to suppress the presumed RA?

While it is appropriate to make a diagnosis of RA in this woman, it is not appropriate to begin oral glucocorticoids unless there is a definitive plan to discontinue them relatively soon. Such a plan, used by many rheumatologists, has received credence from the BeSt study, a randomized trial designed to compare targeted treatment strategies.(7) The results—reaching a low disease activity score—were obtained with two protocols: (a) initial combination therapy with methotrexate, sulfasalazine, and rapidly tapered high-dose prednisone therapy, and (b) initial combination therapy with methotrexate and infliximab (a monoclonal antibody directed against tumor necrosis factor [TNF] alpha). Patients on either of these two regimens were more likely to achieve a clinical remission of disease and to have significantly less joint damage progression after 2 years than those on either sequential monotherapy or step-up to combination therapy. Those on the protocol that included initial prednisone did not have more toxicity than those who were not given prednisone.(7)

In my judgment, the empiric use of glucocorticoids in RA should be limited to one of three scenarios:

  • Administration of a single high dose of intramuscular glucocorticoids very early in the disease process before other therapies are given (i.e., for acute symptom relief).
  • Use of intraarticular glucocorticoids at 2- to 3-month intervals (i.e., to relieve a single painful, inflamed joint).
  • Use of oral glucocorticoids in protocols in which they are tapered to zero dose within a month (i.e., as in the first protocol above, where the glucocorticoid is combined with other disease-modifying agents such as methotrexate).

Case C: Part 1

A 43-year-old man developed a skin eruption believed to be an allergy to a diuretic (his only medication) that he was taking for hypertension. Within several weeks, he became profoundly tired, had spiking fevers during the day, and began to lose weight. He noted stiffness in knees and hands, especially in the morning. His wife noted that he seemed "pale."

On examination, he appeared ill and was mildly tachypneic. Oral temperature was 39.4°C. Blood pressure was 117/75, pulse was 112 (regular). He had a faint serpiginous skin eruption on his trunk. Cardiac examination showed tachycardia and prominent heart sounds. Both his liver and spleen were palpable, the latter extending 6 cm below the left costal margin. Mild pitting edema of the lower extremities was found, but he had no joint effusions. Neurologic examination was reported as "nonfocal." Stool was negative for occult blood.

Initial laboratory tests revealed a WBC count of 2100/µL with 40% band forms and 10% lymphocytes; platelets were low. Smear exam showed multiple fragmented red blood cells (RBCs). Hemoglobin was 8.3 g/dL. Urinalysis showed numerous WBCs. The erythrocyte sedimentation rate (ESR) was 120 mm/h. The C-reactive protein was 28.8 mg/dL.

Further laboratory tests showed: blood urea nitrogen (BUN) was 56 mg/dL; bilirubin was 4.2 mg/dL; aspartate aminotransferase (AST) and gamma-glutamyltransferase (GGT) were both elevated; blood glucose was normal; creatine kinase showed slight elevation; electrolyte tests showed a sodium level of 132 mEq/L and a potassium level of 5.4 mEq/L. Protein results were albumin of 2.8 g/dL and globulin of 2.4 g/dL. Tests for tick-borne diseases were negative.

While awaiting other tests, would it be appropriate to begin 80 mg prednisone/day in divided doses?

Comment: The physicians should consider but then reject this course until most possibilities of infection have been ruled out. If blood and urine cultures are negative, systemic vasculitis or malignancy is high on the list of differential diagnoses. The patient is very ill with a systemic process that could be generating an excess inflammatory cytokine release, so-called cytokine storm.

If infection seems unlikely, malignancy (most often lymphoma) and systemic vasculitis are the most probable diagnoses. In a patient such as this, who is continuing to worsen, empiric steroids, 80100 mg prednisolone/day, intravenously, are appropriate therapy while the work-up continues. However, steroids should not be administered unless infection has been excluded by appropriate studies.

Case C: Part 2

The blood smear showing fragmented RBCs led to other tests: fibrinogen was low; fibrin degradation products were high; and serum ferritin was 1120 µg/L.

As the patient was stabilized, but not improved, with glucocorticoid administration 3 days later, a subsequent bone marrow examination revealed phagocytosis of hematopoietic cells by invasive macrophages. A protocol for treatment of aggressive lymphoma was added to the prednisolone, and the patient improved.

The extremely high ferritin levels, in association with fever, splenomegaly, cytopenias, and hypofibrinogenemia suggest a diagnosis of hemophagocytic lymphohistiocytosis.(8)

The lesson about empiric glucocorticoid therapy from this case is "don't use steroids unless you can't help it!" Although the diagnosis was not clear, the patient was worsening and most causes of infection (with exception of miliary tuberculosis) had been ruled out. Severe multisystem disease without a clear diagnosis is more often caused by infection, malignancy, or systemic vasculitis. Indeed, in the absence of infection, it often is appropriate to test the hypothesis that the problem is a "steroid-responsive disease." The drug can be stopped if the patient does not improve, or if a cause for which there are specific therapies (e.g., the Strongyloides in case A) emerges.

The cases presented above demonstrate important concepts about when to use and when to avoid use of empiric steroids. In addition to RA, other forms of arthritis that have partial or greater response to glucocorticoids, usually given intraarticularly, include acute gouty arthritis, juvenile arthritis, osteoarthritis, pseudogout, psoriatic arthritis, and rheumatic fever. Glucocorticoids are often used in oral doses for systemic vasculitis and other connective tissue diseases, including dermatomyositis/polymyositis, mixed connective tissue disease, polymyalgia rheumatica, and systemic lupus erythematosus.

For fear of exacerbating the underlying disease processes, certain comorbidities should preclude the use of glucocorticoids unless specifically indicated. Physicians should be very cautious about prescribing glucocorticoids for periods longer than several weeks to patients with the following disorders:

  • Infection—all forms. The exception to this rule is when high doses of glucocorticoids are indicated to diminish severe inflammatory manifestations of the infections (e.g., Pneumocystis carinii lung disease and tuberculous pneumonia).
  • Diabetes mellitus.
  • Poorly controlled hypertension.
  • Postmenopausal women, especially those with other risk factors for osteoporosis.


Take-Home Points

  • For a patient with a glucocorticoid-responsive chronic illness that has flared, administration of glucocorticoids intramuscularly or, less frequently, intravenously, will often suppress the disease flare and spare the patient the long-term risks of continued oral therapy, even in small doses.
  • Infection with aggressive bacteria or mycobacteria is the primary contraindication for use of glucocorticoids, unless the principal manifestation of infection is the inflammatory response to the inciting agent (e.g., diffuse tuberculous pneumonia, severe cases of CAP).
  • In severely ill patients with an undiagnosed systemic disease in whom most forms of infection have been ruled out, a therapeutic trial of high-dose glucocorticoids gives useful information and may help the patient.
  • As advances are made in the treatment of multisystem chronic diseases (including RA, inflammatory bowel disease, and systemic vasculitis), addition of other medications to supplement glucocorticoids is almost always indicated.


Edward D. Harris, Jr., MD George DeForest Barnett Professor, Emeritus, Stanford University Academic Secretary to Stanford University, Emeritus

Faculty Disclosure: Dr. Harris has declared that neither he, nor any immediate member of his family, has a financial arrangement or other relationship with the manufacturers of any commercial products discussed in this continuing medical education activity. In addition, his commentary does not include information regarding investigational or off-label use of pharmaceutical products or medical devices.


1. Mikami K, Suzuki M, Kitagawa H, et al. Efficacy of corticosteroids in the treatment of community-acquired pneumonia requiring hospitalization. Lung. 2007;185:249-255. [go to PubMed]

2. Garcia-Vidal C, Calbo E, Pascual V, et al. Effects of systemic steroids in patients with severe community-acquired pneumonia. Eur Respir J. 2007;30:951-956. [go to PubMed]

3. Furugen M, Higa F, Hibiya K, et al. Legionella pneumophilia infection induces programmed cell death, capsase activation, and release of high-mobility group box 1 protein in A549 alveolar epithelial cells: inhibition by methyl prednisolone. Respir Res. 2008;9:39. [go to PubMed]

4. Lane NE, Lukert B. The science and therapy of glucocorticoid-induced bone loss. Endocrinol Metab Clin North Am. 1998;27:465-483. [go to PubMed]

5. O'Dell JR, Haire CE, Palmer W, et al. Treatment of early rheumatoid arthritis with minocycline or placebo: results of a randomized, double-blind, placebo-controlled trial. Arthritis Rheum. 1997;40:842-848. [go to PubMed]

6. Harris ED Jr. Glucocorticoid therapy in rheumatoid arthritis: intra-articular injections versus systemic administration. Nat Clin Pract Rheumatol. 2006;2:184-185. [go to PubMed]

7. Allaart CF, Breedveld FC, Dijkmans BA. Treatment of recent-onset rheumatoid arthritis: lessons from the BeSt study. J Rheumatol Suppl. 2007;80:25-33. [go to PubMed]

8. Allen CE, Yu X, Kozinetz CA, McClain KL. Highly elevated ferritin levels and the diagnosis of hemophagocytic lymphohistiocytosis. Pediatr Blood Cancer. 2008;50:1127-1235. [go to PubMed]

This project was funded under contract number 75Q80119C00004 from the Agency for Healthcare Research and Quality (AHRQ), U.S. Department of Health and Human Services. The authors are solely responsible for this report’s contents, findings, and conclusions, which do not necessarily represent the views of AHRQ. Readers should not interpret any statement in this report as an official position of AHRQ or of the U.S. Department of Health and Human Services. None of the authors has any affiliation or financial involvement that conflicts with the material presented in this report. View AHRQ Disclaimers
Related Resources From the Same Author(s)
Related Resources