Glucocorticoids for Pneumonia in Africa — Old Therapy, New Context
Arthur Kwizera, Martin W. Dünser
Published October 29, 2025
DOI: 10.1056/NEJMe2514533
The African continent is home to 1.55 billion people, one fifth of the world’s population, spread over an area larger than the United States, China, and India combined. The disease burden is enormous,1 and health care resources are severely constrained. In one multinational survey, one in eight hospitalized patients was critically ill, and 20.7% died within 7 days after admission.2 Lacking sufficient resources, most hospitals are forced to care for patients in general wards. Lower respiratory tract infections, including community-acquired pneumonia (CAP), frequently cause critical illness3 with high mortality. Effective and low-cost interventions applicable in this setting are desperately needed.
In a report now published in the Journal, Lucinde and colleagues describe the results of the Steroids in Pneumonia (SONIA) trial,4 a pragmatic, randomized study conducted in 18 Kenyan hospitals with limited or no intensive care capacity. The trial, which enrolled 2180 adults with CAP who had been admitted to general medical wards, compared standard care for CAP with standard care plus a 10-day course of glucocorticoids (50 mg prednisolone equivalent, preferably oral). At 30 days, mortality was 22.6% with glucocorticoids as compared with 26.0% with standard care (hazard ratio for death, 0.84; 95% confidence interval, 0.73 to 0.97; P=0.02). Although modest, this absolute difference of 3.4 percentage points represents a potentially large effect on public health in a region where CAP mortality remains high.
Because of resource limitations, the severity of CAP was not assessed in the current trial, which makes comparison with other studies difficult. However, more than one third of the patients had hypoxemia at baseline and almost 10% had hypotension, which suggests that a substantial number had severe disease.
The primary safety concern was hyperglycemia (which occurred in 16.6% of the patients who received glucocorticoids). Although hyperglycemia is mostly controllable, blood glucose monitoring is difficult in overburdened sub-Saharan hospitals. As the prevalence of diabetes is increasing across Africa, introducing glucocorticoid therapy without strengthened laboratory and nursing capacity may put patients at risk and requires a nuanced balance of the survival benefits and metabolic complications.
The results of the SONIA trial are in line with a long arc of research on the role of glucocorticoids in the management of CAP. The initial excitement about steroid use for CAP in the mid-20th century was subsequently muted by contradictory results and fears of harm. More recently, large-scale randomized clinical trials conducted in high-income countries have reinvigorated interest, in which adjunctive steroids in severe CAP improved survival and accelerated recovery, as shown in the CAPE COD and REMAP-CAP trials.5,6 The findings of the SONIA trial align with this evidence.
Yet, the SONIA trial is more than just a confirmatory study. For African clinicians, it offers locally generated, context-relevant data, a rarity in global critical care research. Historically, African practitioners have had to rely on evidence from settings with vastly different resources and patient profiles, often with uncertain or even harmful results when applied locally, as shown in previous studies of sepsis and fluid resuscitation.7 Similar caution is necessary before extrapolating the SONIA results to non-African contexts where the available resources and mortality of patients with CAP who are not treated in an intensive care unit (ICU) are quite different. In such cases, the glucocorticoid-associated reduction in mortality observed in the SONIA trial may not be seen.
Particularly relevant to sub-Saharan Africa is the frequent coexistence of hypoxemia in CAP-related critical illness.8 The SONIA trial was not powered to assess a mortality benefit in the hypoxemia subgroup specifically. Future work should explore whether combining judicious noninvasive oxygen therapy with glucocorticoids could further reduce mortality. However, designing and executing such trials in Africa would present unique challenges, because limited infrastructure (including oxygen therapies) and ICU resources, workforce shortages, and competing clinical demands make conventional randomized trials difficult to sustain.9 A registry-based trial platform across first-referral hospitals could offer a cost-efficient, scalable method to test such strategies while building sustainable surveillance capacity.10Embedding safety monitoring, cost-effectiveness analysis, and review of implementation outcomes would generate actionable evidence tailored to real-world conditions.
The SONIA trial underscores that progress in global health does not always hinge on new types of technology but on reevaluating existing therapies within the contexts of where they are most needed. For clinicians, it offers hope that glucocorticoids, an inexpensive, accessible intervention, can save the lives of African patients with pneumonia. For policymakers, it highlights the necessity of ensuring safe, equitable implementation. And for researchers, it reaffirms that with good mentorship and leadership, Africa can and should take a lead role in the generation of evidence to meet its own health care challenges.
