Interestingly, the 2019 IDSA guidelines recommended that adults with CAP, including those with severe pneumonia, not routinely receive corticosteroids in the absence of refractory septic shock.¹ What do we do with conflicting recommendations? Many clinicians, including myself, opted to wait for additional data–which we now have.

The tides have shifted towards administration of hydrocortisone to a select group of patients admitted with pneumonia. The CAPE COD trial, published early this year, demonstrated a 5.6 percent (95 percent CI, 1.7 to 9.6 percent) absolute reduction in mortality by day 28 in patients with severe CAP randomized to receive adjunct hydrocortisone 200 mg intravenously daily compared with those who received placebo.³ Additionally, fewer patients in the hydrocortisone arm were intubated, at 19.5 percent, compared with those in the placebo group, at 27.7 percent (Hazard Ratio, 0.69 [95 percent CI 0.60-0.94]). Although patients in the hydrocortisone group had more hyperglycemia, this did not translate into an increase in measurable adverse events. Crucially, to be included in the pivotal CAPE COD study, patients had to meet one of the following criteria: Pulmonary Severity Index score of >130, or ventilatory support (invasive mechanical ventilation, non-invasive ventilation (NIV), high-flow nasal cannula with fraction of inspired oxygen (FiO2) >0.5 and PaO2/FiO2 (P/F) ratio <300, or nonrebreather mask with a P/F ratio dependent on the oxygen flow rate).

The impressive results of the CAPE COD study should not be interpreted as “all hospitalized patients with CAP should receive steroids.” First, patient selection is critical. The patients in the CAPE COD trial were sick; nearly half of the cohort was intubated or on NIV. This is consistent with a 2018 meta-analysis of 11 trials that found steroids were associated with a reduced odds of death compared to control groups (relative risk, 0.66; 95 percent CI, 0.47-0.92). In subgroup analyses of severity of illness, the mortality benefit was only statistically significant in those with severe pneumonia. Further, the trial excluded those with influenza or aspiration pneumonia. Second, steroid choice is important. The CAPE COD trial used hydrocortisone, which has more mineralocorticoid activity than most other glucocorticoids. This may explain some differences between the results of CAPE COD and negative trials that used steroids with less mineralocorticoid activity.⁵ Third, hydrocortisone was early in the hospital course–an average of 20 hours from hospital admission. Future trials will provide additional insight, but in the interim, the data suggest we should administer hydrocortisone administration to patients who would meet the CAPE COD inclusion criteria.