It is one of the most reflexive co-prescriptions in hospital medicine: a patient is started on corticosteroids, and almost in the same pen-stroke, a proton pump inhibitor (PPI) is added “for gastric protection.” It feels prudent, harmless, and defensible. But is it actually evidence-based?
The short answer is no — corticosteroids alone are not an indication for a PPI. The reflex stems from a misreading of the evidence and conflates the modest, dose-dependent risk of steroids with the well-established risk of NSAIDs. This article unpacks the actual data, defines who genuinely needs protection, and explains why the “just in case” PPI is rarely as harmless as it seems.
The Origin of the Myth
The belief that steroids “cause ulcers” is deeply embedded in clinical folklore. It is largely a historical artefact from older, lower-quality studies and a degree of guilt by association — steroids and NSAIDs are often used for the same inflammatory conditions, and the GI risk of one was unfairly attributed to the other.
The reality is more nuanced. The absolute risk of clinically significant upper GI bleeding or perforation attributable to corticosteroid monotherapy in an otherwise low-risk outpatient is very small. The number needed to harm is high, and the number needed to treat with a PPI to prevent a single bleed in this group is correspondingly enormous.
What the Evidence Actually Shows
Steroids Alone: A Modest, Context-Dependent Signal
- Outpatients: In the community setting, corticosteroid use alone is associated with a negligible-to-small increase in absolute GI bleeding risk. For the vast majority of patients on a short course (e.g., a COPD exacerbation or an asthma flare), routine PPI co-prescription is not warranted.
- Hospitalised patients: The picture shifts. Acutely unwell, hospitalised patients are a different population — they have more comorbidity, physiological stress, and competing risk factors. The relative risk of GI bleeding with steroids appears higher here, but this largely reflects the company the steroid keeps (critical illness, coagulopathy, concurrent NSAIDs/anticoagulants) rather than the steroid acting in isolation.
The Real Culprit: Steroids PLUS NSAIDs
This is the single most important point. The combination of a corticosteroid and an NSAID is genuinely dangerous — the two act synergistically to dramatically increase the risk of peptic ulceration and bleeding (estimates of a fourfold or greater increase over either agent alone).
Much of the historical “steroids cause ulcers” signal almost certainly came from patients who were also taking NSAIDs. When the NSAID is the true driver, the lesson is not “add a PPI to the steroid” — it is question whether the NSAID is needed at all.
Who Genuinely Needs Gastroprotection?
Rather than blanket prescribing, the decision should be risk-stratified. A PPI is justified when a patient on steroids also has one or more of the following:
- Concurrent NSAID use (including low-dose aspirin in higher-risk patients). This is the strongest indication.
- Concurrent anticoagulant or antiplatelet therapy (e.g., DOACs, warfarin, clopidogrel).
- A history of peptic ulcer disease or prior GI bleeding.
- Advanced age combined with other risk factors.
- Critical illness / ICU-level stress ulcer prophylaxis criteria (e.g., mechanical ventilation, coagulopathy, shock) — a separate indication driven by the illness, not the steroid.
A Practical Risk-Based Approach
| Clinical Scenario | Routine PPI Needed? | Reasoning |
| Short steroid course, no other risk factors (e.g., asthma/COPD flare) | No | Absolute risk negligible; NNT to prevent harm is very high. |
| Steroid + NSAID | Yes | Synergistic ulcer risk; PPI clearly protective. Better still: stop the NSAID. |
| Steroid + anticoagulant/antiplatelet | Yes | Elevated bleeding risk; co-prescription justified. |
| Steroid + prior PUD or GI bleed | Yes | Established high-risk history. |
| Critically ill / meets stress ulcer prophylaxis criteria | Yes | Indication is the critical illness, not the steroid itself. |
| Long-term low-dose steroid, otherwise low risk | Generally No | Reassess if risk factors accrue; avoid indefinite “default” PPI. |
The Cost of the “Harmless” PPI
The reflex PPI is rarely scrutinised because it is perceived as benign. But unnecessary, often indefinite, PPI therapy carries its own well-documented downsides:
- Infections: Increased risk of Clostridioides difficile infection and community-acquired pneumonia.
- Micronutrient deficiencies: Hypomagnesaemia and reduced vitamin B12 absorption with prolonged use.
- Bone health: Associations with increased fracture risk — particularly relevant given steroids themselves drive osteoporosis.
- Renal effects: Links to acute interstitial nephritis and chronic kidney disease.
- Polypharmacy and prescribing inertia: A PPI started “temporarily” on admission is frequently never stopped, following the patient into the community indefinitely.
The “IM Doctor” Clinical Pearl
“Protect the patient who is actually at risk — not the prescription.”
Steroids alone, in an otherwise low-risk patient, do not justify a PPI. Before reaching for gastroprotection, ask the more important question: is this patient also on an NSAID? If so, the highest-value intervention is usually to stop the NSAID, not to layer on another drug.
And if you do start a PPI on admission for a genuine short-term indication, document the stop date. The most common harm from a PPI is not its side-effect profile — it is the one nobody ever discontinues.
Key Takeaways
- Steroids alone are NOT an indication for a PPI in low-risk patients. The absolute GI risk is small.
- The real danger is steroid + NSAID — a synergistic, clinically significant ulcer risk.
- Risk-stratify, don’t reflex-prescribe. Reserve PPIs for concurrent NSAIDs/antithrombotics, prior PUD/GI bleed, advanced age with risk factors, or critical illness.
- The “harmless” PPI isn’t. Consider C. difficile, pneumonia, hypomagnesaemia, B12 deficiency, fracture, and renal risks.
- If you start one, plan to stop it. Always document a review or stop date to prevent indefinite, unintended continuation.
Clinical Scenario: The Unindicated Gastroprotection
Setting: Acute Medical Unit (AMU) Post-Take Ward Round
Patient: A 68-year-old male admitted with an acute exacerbation of Chronic Obstructive Pulmonary Disease (COPD).
The Presentation
During the morning ward round, the FY1 doctor presents the management plan for the patient, who has no history of dyspepsia, acid reflux, or peptic ulcer disease. He is not taking any antiplatelets, anticoagulants, or NSAIDs.
The junior doctor reads out the proposed prescription chart:
“I’ve started him on nebulisers, a 5-day course of oral Prednisolone 40mg once daily, and I’ve also added Lansoprazole 30mg daily for gastroprotection while he’s on the steroids.”
The Teaching Moment (Why we cross it off)
Turning it into a quick bedside teaching point, I explained the evidence-base to the FY1:
- The Myth of Routine Co-Prescribing: Systemic corticosteroid monotherapy in ambulatory or stable AMU patients does not significantly increase the absolute risk of de novo peptic ulceration or major upper GI bleeding. The absolute baseline risk of a steroid-induced bleed on monotherapy is incredibly low (around 0.13%).
- Look for the Synergy: Gastroprotection with a PPI is only indicated if the steroids are paired with definitive synergistic risk factors—such as concurrent NSAIDs, aspirin, DOACs, or a documented prior history of peptic ulcer disease.
- The Harm of Polypharmacy: Adding an unindicated PPI exposes the patient to unnecessary long-term risks (such as Clostridioides difficile infection, hypomagnesemia, and micro-nutrient malabsorption) without providing any evidence-based clinical benefit.
References and Further Reading
- NICE Clinical Knowledge Summaries: Corticosteroids – oral. Guidance on adverse effects and gastroprotection considerations. https://cks.nice.org.uk/topics/corticosteroids-oral/
- NICE NG184: Acute upper gastrointestinal bleeding in over 16s – management. https://www.nice.org.uk/guidance/ng184
- Narum S, Westergren T, Klemp M. Corticosteroids and risk of gastrointestinal bleeding: a systematic review and meta-analysis. BMJ Open. 2014;4(5):e004587. https://bmjopen.bmj.com/content/4/5/e004587
- StatPearls: Proton Pump Inhibitors. Review of indications and adverse effects of long-term use. https://www.ncbi.nlm.nih.gov/books/NBK557385/
Dr. Kunal Varma Bheecarry, MD is a GMC registered substantive Consultant Physician in the UK, specializing in Internal Medicine. With more than 15 years of experience on the clinical frontlines in both General Internal Medicine and Acute Medicine, he is passionate about translating complex medical topics, common problems and clinical guidelines into clear, actionable knowledge for healthcare professionals and patients alike. Every article is written and rigorously reviewed by him and other registered medical professionals to ensure clinical accuracy, guideline compliance, evidence-based reliability, authoritativeness and trustworthiness.
Disclaimer: All content is purely educational and does not constitute personal medical advice. It is not a substitute for professional clinical advice and readers must consult their own primary team. All views and interpretations are mine and not related to the work place.