Background: Community-acquired pneumonia (CAP) can lead to pulmonary and systemic inflammation, resulting in impaired gas exchange, sepsis, organ failure, and an increased risk of death. Corticosteroids have excellent anti-inflammatory and immunomodulatory effects that could mitigate some of the inflammation caused by pneumonia. There have been several randomized trials that have shown glucocorticoids have positive effects in patients with CAP. However, except for one trial, none showed a between-group difference regarding mortality.1-6 We now have the Community-Acquired Pneumonia: Evaluation of Corticosteroids (CAPE COD) Trial.
Clinical Question: In adult patients with severe community-acquired pneumonia admitted to the ICU, does early hydrocortisone treatment, compared to standard therapy, reduce mortality at 28 days?
What they did:
- Investigators conducted a phase 3, multicenter, double-blind, randomized, controlled, superiority trial.
- Patients enrolled from 31 Intensive Care Units in France.
- Patients admitted to the ICU with severe community-acquired pneumonia were randomized in a 1:1 ratio to receive hydrocortisone or a placebo.
- The study was funded by the French Ministry of Health.
- Adult patients (≥18 years old).
- Admitted to participating ICUs with severe community-acquired pneumonia.
- Pneumonia diagnosis supported by clinical and radiologic criteria.
- Severity of pneumonia was defined by at least one of the following criteria:
- Invasive or noninvasive mechanical ventilation with a PEEP of 5 cm of water
- High-flow nasal cannula oxygen therapy with a Pao2:Fio2 ratio of less than 300; With FiO2 ≥50%
- Non-rebreather with PaO2:FiO2 ratio <300
- Pao2:Fio2 ratio less than 300
- Pulmonary Severity Index score > 130 (Category V, which is associated with the highest mortality)
- Pneumonia caused by influenza
- Pneumonia caused by septic shock
- Patient treated by vasopressors for septic shock at the time of inclusion
- Clinical history suggesting aspiration of gastric content
- Patient treated by invasive mechanical ventilation within 14 days before current hospital admission
- Patients treated by antibiotics for a respiratory infection for more than seven days at the admission to the hospital (except if a pathogen resistant to these antibiotics is isolated)
- History of cystic fibrosis
- Post-obstructive pneumonia
- Patients in which rapid PCR test is positive for flu
- Active tuberculosis or fungal infection
- Active viral hepatitis or active infection with herpes viruses
- The decision to withhold mechanical ventilation or endotracheal intubation
- Hypersensitivity to corticosteroids
- Patients needing anti-inflammatory corticosteroids or substitutive hydrocortisone for any reason
- Patients under treatment with more than 15 mg/d of prednisone (or equivalent) for more than 30 days
- Patients already enrolled in another drug trial with mortality as an end-point. If the patient already participates in another therapeutic trial with a different endpoint, the investigator must verify that inclusion in CAPE COD can not prejudice it.
- Pregnant or breastfeeding woman
- Patient on judicial protection
- Hydrocortisone 200mg daily (continuous infusion) for 4 days. On the 4th day, the medical team used predefined criteria to decide whether to administer hydrocortisone for a total of 8 or 14 days, depending on whether the patient’s condition had improved.
- Normal Saline via continuous infusion according to the same regimen used in the hydrocortisone group.
- Death from any cause by day 28
- Death from any cause by day 90
- The length of ICU stay
- Noninvasive ventilation or endotracheal intubation among patients who were not receiving any type of ventilation at baseline
- Endotracheal intubation among patients who were receiving noninvasive ventilation at baseline
- The initiation of vasopressor therapy by day 28
- Number of ventilator-free days and vasopressor-free days by day 28
- Change in the PaO2:FIO2 ratio by day 7
- Change in SOFA score by day 7
- Secondary infections or GI bleeding by day 28
- The daily amount of insulin administered by day 7
- Weight gain by day 7
- 5948 patients were assessed for eligibility.
- 5148 excluded
- 800 enrolled and randomized in the trial.
- 401 in the hydrocortisone group
- 399 in the control group.
- 795 patients were included in the primary analysis
- 1 patient from the hydrocortisone group died before receiving treatment
- 2 patients in the control group withdrew their consent
- 2 did not have the legal capacity to give consent
- Type of respiratory support
- IMV ≈22%
- NIV ≈22%
- HFNC ≈42%
- NRB ≈14%
- Median Interval from Hospital Admission to ICU Admission ≈5hrs
- Median Interval from ICU Admission to Initiation of Trial Agent ≈15hrs
- 28-day mortality was 6.2% in the hydrocortisone group vs. 11.9% in the placebo group (p = 0.006).
- By day 90, mortality was 9.3% in the hydrocortisone group and 14.7% in the placebo group (absolute difference, −5.4 percentage points; 95% CI, −9.9 to −0.8).
- Among 442 patients who had not received any mechanical ventilation at baseline:
- Endotracheal intubation was performed in 18.0% of the hydrocortisone group.
- Endotracheal intubation was performed in 29.5% in the placebo group.
- Hazard ratio: 0.59 (95% CI, 0.40 to 0.86).
- Among 618 patients who had received no invasive ventilation at baseline:
- Cumulative incidence of invasive mechanical ventilation before day 28:
- 19.5% in the hydrocortisone group.
- 27.7% in the placebo group.
- Hazard ratio: 0.69 (95% CI, 0.50 to 0.94).
- Cumulative incidence of invasive mechanical ventilation before day 28:
- Among the 703 patients who had not received vasopressors at baseline:
- Cumulative incidence of vasopressor initiation:
- 15.3% in the hydrocortisone group.
- 25.0% in the placebo group.
- Hazard ratio: 0.59 (95% CI, 0.43 to 0.82).
- Cumulative incidence of vasopressor initiation:
- During the first 28 days after randomization, 169 serious adverse events occurred in 151 of 795 patients (19.0%).
- 70 events in the hydrocortisone group.
- 99 events in the placebo group.
- ICU-acquired infections were observed in:
- 9.8% of patients in the hydrocortisone group.
- 11.1% of patients in the placebo group.
- Hazard ratio: 0.87 (95% CI, 0.57 to 1.34).
- Gastrointestinal bleeding was rare in both groups.
- By day 7, patients receiving insulin therapy during the first week of the trial received a median of:
- In the hydrocortisone group, 35.5 IU per day (interquartile range, 15.0 to 57.5).
- In the placebo group, 20.5 IU per day (interquartile range, 9.4 to 48.5).
- The investigators framed a research question centered on patient outcomes, ensuring the study’s relevance and practicality.
- The researchers employed a double-blind, placebo-controlled, randomized controlled trial design to minimize potential biases, enhancing the reliability of the results.
- Multiple centers were involved, increasing the study’s external validity and making the findings more applicable and generalizable to diverse patient populations.
- The trial’s registration with clinicaltrials.gov enhanced transparency and allowed other researchers to replicate the study, promoting scientific rigor.
- The study included a substantial number of participants, increasing the statistical power and precision of the results.
- Researchers used concealed, centralized electronic web-based randomization to ensure the equitable allocation of participants to treatment arms, reducing selection bias.
- The research team assessed patients for eligibility consecutively, minimizing selection bias and ensuring a more representative study population.
- The treating physicians and research personnel were all unaware of the treatment arm, reducing the potential for bias in treatment administration and outcome assessment.
- The researchers labeled placebo and hydrocortisone bottles identically, decreasing the likelihood of unblinding bias and ensuring that neither patients nor healthcare providers could differentiate between the two.
- Researchers conservatively assumed that patients with missing data for the primary outcome had met the primary outcome (death), addressing potential attrition bias in the analysis.
- Enrolled patients represented a critically ill cohort; nearly 45% were on mechanical ventilation.
- The study was conducted exclusively in one country, limiting its external validity.
- The researchers applied strict exclusion criteria, likely introducing some selection bias.
- The researchers initially estimated a mortality rate of 27% for the placebo group and 20% for the hydrocortisone group but observed a lower mortality rate of 11.9% for placebo and 6.2% for hydrocortisone, which may impact the study’s validity.
- Although the investigators calculated a necessary sample size of 1200, they enrolled only 800 participants, potentially affecting the study’s statistical power.
- 45% of the enrolled patients came from just 4 ICUs, limiting the generalizability of the results.
- Approximately 70% of the enrolled cohort consisted of men, limiting the generalizability of the findings to women.
- The median age of the participants was 67, potentially limiting the study’s generalizability to younger age groups.
- There were important differences in baseline demographics.
- Participants in the placebo arm had a median PSI score of 130 compared to 127 in the hydrocortisone arm.
- No pathogen was identified in 47% of patients in the hydrocortisone group and 42% in the placebo group.
- The trial was stopped early due to observed benefits, which could potentially overestimate the study’s results.
- While patients received respiratory support, supportive care, and antibiotics at the discretion of the medical team, the specific interventions provided outside the study were not documented, potentially leading to co-intervention bias.
- The search for pathogens was left to the discretion of the medical team, introducing variability in the diagnostic process.
- More than 75% of patients did not complete the treatment course.
- The potential benefits of corticosteroids in severe community-acquired pneumonia patients likely depend on several key factors. Firstly, corticosteroids may benefit pneumonia but are most likely effective in the sickest patients requiring ICU admission without contraindications. This represents a niche population where the benefits outweigh the risks.
- Secondly, while corticosteroids can be beneficial in severe pneumonia, the magnitude of this benefit may not be substantial. Demonstrating a significant reduction in mortality with a single medication in the critical care setting is challenging.
- Thirdly, the potential benefit of corticosteroids in severe pneumonia is most likely associated with early treatment, ideally administered within 24 hours from patient presentation.
- Lastly, although C-reactive protein (CRP) is not typically used in the emergency department to diagnose bacterial pneumonia, patients with CRP levels exceeding 15mg/dL (or > 150 mg/L) may derive benefits from corticosteroid therapy. CRP is a marker of inflammation, and corticosteroids are known to attenuate this inflammatory response.
- More research is needed to identify specific subgroups that may benefit from early corticosteroid therapy.
Inside The Numbers
- Early hydrocortisone treatment significantly reduced 28-day mortality in severe community-acquired pneumonia ICU patients (6.2% vs. 11.9% in the placebo group, p = 0.006), with consistent benefits across secondary outcomes.
- The Relative Risk reduction, which represents the decrease in risk relative to the placebo group, was 47%. This corresponds to an absolute risk difference of 5.7%, meaning there was a 5.7% lower chance of mortality at 28 days in the hydrocortisone group compared to the placebo group. This translates to a Number Needed to Treat (NNT) of 17.5, indicating that for nearly every 18 severe CAP ICU patients treated with hydrocortisone, one additional patient’s life was saved in 28 days compared to the placebo group.
- However, the fragility index of 6 warns that these benefits hinge on relatively few events, emphasizing the need for cautious interpretation. Furthermore, truncated RCTs can overestimate treatment effects due to random error. This effect is neutralized as the number of enrolled patients increases toward the planned sample sizes.
- Investigators excluded 86% of initially screened patients based on the stringent selection criteria to eliminate those who might be harmed or benefited from hydrocortisone treatment. However, some criteria were rather subjective. For instance, more than 10% of excluded patients were suspected of having aspiration pneumonia, a diagnosis often relying on a suggestive medical history and classic X-ray pattern that may not apply to all.
- About 25% of the excluded patients had septic shock, but the classification of hypotension as septic shock was at the discretion of the treating team. Notably, 45% of the enrolled patients required positive pressure mechanical ventilation, which often includes sedation, both of which could lead to hypotension, necessitating vasopressor therapy.
- Influenza patients were also excluded due to potential harm from corticosteroids, and although no significant differences in post-enrollment positive influenza tests were observed between groups (3.8% in the hydrocortisone group vs. 4% in the placebo group), the treatment teams made testing decisions, and specific testing patterns for each cohort were not disclosed.
- The placebo group had a higher percentage of patients in category V PSI compared to the hydrocortisone group. (49.2% vs 45.7), category V PSI carries a 30% mortality rate, whereas the mortality rate is 10% for category IV PSI. Additionally, 47% of the patients in the hydrocortisone group and 42% in the placebo group had no identified pathogens. Subgroup analysis of patients with PSI > 130 and those with isolated pathogens showed confidence intervals that crossed 0, indicating no clear benefit.
- Considering that the placebo group had 12 more patients with more severe baseline illness who were more likely to die and 16 more patients with an identified bacterial infection, given the trial’s fragility index of only 6, these disparities could have greatly impacted the observed outcome differences.
- In a recent study by Meduri et al. (2022) involving 586 ICU-admitted patients with severe community-acquired pneumonia (CAP), no benefit was observed from methylprednisolone administration.8 But, it’s crucial to distinguish between this study and the CAPE COD trial. Notably, the Meduri trial had 96% male participants, while CAPE COD had 67%. Subgroup analysis in CAPE COD revealed a more pronounced benefit of hydrocortisone in women. Additionally, hydrocortisone was administered within 24 hours in CAPE COD, differing from the Meduri trial’s 96-hour window.
- However, a recent meta-analysis (Pitre 2023) of 18 RCTs involving 4,661 patients similarly concluded that corticosteroids reduce mortality in severe CAP, lessen the need for invasive mechanical ventilation, and lower ICU admission rates.9
Author’s Conclusion: “Among patients with severe community-acquired pneumonia being treated in the ICU, those who received hydrocortisone had a lower risk of death by day 28 than those who received placebo.”
The CAPE COD trial findings suggest that administering corticosteroids within the first 24 hours can significantly reduce mortality in severe community-acquired pneumonia patients admitted to the ICU. However, it’s worth noting that more than three-quarters of enrolled patients didn’t complete the treatment protocol, and statistical variations among different subgroups require further investigation. Nevertheless, the best available evidence supports the conclusion that early corticosteroid administration can reduce mortality in severe community-acquired pneumonia patients in the ICU.
Clinical Bottom Line:
Early administration of corticosteroids can reduce mortality in severe community-acquired pneumonia patients in the ICU.
- Confalonieri M, Urbino R, Potena A, et al. Hydrocortisone infusion for severe community-acquired pneumonia: a preliminary randomized study. Am J Respir Crit Care Med 2005;171:242-8. PMID: 15557131
- Snijders D, Daniels JMA, de Graaff CS, van der Werf TS, Boersma WG. Effi- cacy of corticosteroids in community-acquired pneumonia: a randomized double-blinded clinical trial. Am J Respir Crit Care Med 2010;181:975-82. PMID: 20133929
- Meijvis SCA, Hardeman H, Remmelts HHF, et al. Dexamethasone and length of hospital stay in patients with community-acquired pneumonia: a randomised, double-blind, placebo-controlled trial. Lancet 2011;377:2023-30. PMID: 21636122
- Fernández-Serrano S, Dorca J, Garcia- Vidal C, et al. Effect of corticosteroids on the clinical course of community-acquired pneumonia: a randomized controlled trial. Crit Care 2011;15(2):R96. PMID: 21406101
- Blum CA, Nigro N, Briel M, et al. Adjunct prednisone therapy for patients with community-acquired pneumonia: a multicentre, double-blind, randomised, placebo-controlled trial. Lancet 2015;385:1511-8. 12. PMID: 25608756
- Torres A, Sibila O, Ferrer M, et al. Effect of corticosteroids on treatment failure among hospitalized patients with severe community-acquired pneumonia and high inflammatory response: a randomized clinical trial. JAMA 2015;313:677-86. PMID: 25688779
- Dequin P, Meziani F, Quenot J, et al. Hydrocortisone in severe community-acquired pneumonia. N Engl J Med. 2023;388(21):1931-1941. PMID: 36942789
- Meduri GU, Shih M-C, Bridges L, et al. Low-dose methylprednisolone treatment in critically ill patients with severe community-acquired pneumonia. Intensive Care Med 2022;48:1009-1023. PMID: 35723686
- Pitre T, Abdali D, Chaudhuri D, Pastores SM, Nei AM, Annane D, Rochwerg B, Zeraatkar D. Corticosteroids in Community-Acquired Bacterial Pneumonia: a Systematic Review, Pairwise and Dose-Response Meta-Analysis. J Gen Intern Med. 2023 Aug;38(11):2593-2606. Epub 2023 Apr 19. PMID: 37076606
Guest Post By:
Marco Propersi, DO FAAEM
Vice-Chair, Emergency Medicine
Assistant Emergency Medicine Program Director
Vassar Brothers Hospital, Poughkeepsie, New York
Chanel Thomas, MD
PGY-1, Emergency Medicine Resident
Vassar Brothers Hospital, Poughkeepsie, New York
Post Peer Reviewed By: Salim R. Rezaie, MD (Twitter/X: @srrezaie)