Short-Course Glucocorticoids May Not Be Harmless

May 22, 2017

Short Attention Span Summary

Clay Smith, MD,  Twitter

Can this be right?
Using insurance data, >1.5 million US adults were studied to determine the effect of short term steroid use on adverse outcomes.  Authors found that risks of sepsis, venous thromboembolism, and fracture increased within 30 days of the short-term steroids.  I wonder about confounding.  Could it be that those who received short courses of steroids were more ill, already battling infection or injury, and were already at greater risk for sepsis, VTE, or fall?  What would have been the impact if steroids had been withheld – increased admissions, increased asthma mortality, for instance?  So…what to do with this?  I acknowledge the association but at this point plan to continue giving short course steroids when appropriate.  There is quantifiable good from short-course steroids, harm from withholding steroids in certain cases, and I can’t see the small risks in this study overturning that.

Spoon Feed
There was an association with short-term steroid use and several adverse outcomes.

Source/Another Spoonful
Short term use of oral corticosteroids and related harms among adults in the United States: population based cohort study. BMJ. 2017 Apr 12;357:j1415. doi: 10.1136/bmj.j1415.

Journal Watch offered their cautious take on this article.

---

Peer Reviewer Comments
Thomas Davis, MD,  Twitter

This is a very provocative article. They did a nice job using a modern technique (self controlled case study) to control for confounders. Additionally, the study selected out older patients as well as those with more serious comorbidities which reflects the patient population I am more likely to prescribe steroids. Though as pointed out, there may very well be remaining confounders that still cannot be controlled for. 

However, my take away from the study is a little less dismissive of the findings and a little more cautious than the one described. For me, the study reminds me to respect drugs (which was the take away from our journal club review of this article). It encourages me to pause before prescribing and think about the real side effects of steroids. In a young healthy child with croup or an asthma exacerbation where the indication for steroids is well established, I won’t hesitate to continue using steroids despite this article. Those children have a low baseline risk for complications. 

However, in regard to a recent patient with radicular back pain who has poorly controlled diabetes, I chose not to prescribe steroids even though our spinal surgeons advocate its use, because the evidence for steroids is weak and the risk is measurable. 

As this study points out, about half of the indications for steroids were for URI, back pain, and allergies. In many of these cases, there is little evidence that steroids offer benefit. It makes me think of the recent Ann Emerg Med article by Barniol that showed no additive benefit of steroids to antihistamines for urticaria. Or the studies showing no additive benefit to using steroids in patients with anaphylaxis to prevent rebound. 

At the end of the day:
If the data for steroids are good, this article doesn’t change a thing.
If the data for steroids are weak, this article makes me pause.
If the data for steroids show improvement in quality of life only, I talk to the patient.