The benefit of corticosteroids in community-acquired pneumonia (CAP) remains controversial. We did a meta-analysis to include all the randomized controlled trials (RCTs) which used corticosteroids as adjunctive therapy, to examine the benefits and risks of corticosteroids in the treatment of CAP in adults.
Databases including Pubmed, EMBASE, the Cochrane controlled trials register, and Google Scholar were searched to find relevant trials. Randomized and quasi-randomized trials of corticosteroids treatment in adult patients with CAP were included. Effects on primary outcome (mortality) and secondary outcomes (adverse events) were accessed in this meta-analysis.
Nine trials involving 1001 patients were included. Use of corticosteroids did not significantly reduce mortality (Peto odds ratio [OR] 0.62, 95% confidence interval [CI] 0.37–1.04; P = 0.07). In the subgroup analysis by the severity, a survival benefit was found among severe CAP patients (Peto OR 0.26, 95% CI 0.11–0.64; P = 0.003). In subgroup analysis by duration of corticosteroids treatment, significant reduced mortality was found among patients with prolonged corticosteroids treatment (Peto OR 0.51, 95% CI 0.26–0.97; P = 0.04; I2 = 37%). Corticosteroids increased the risk of hyperglycemia (Peto OR 2.64, 95% CI 1.68–4.15; P<0.0001), but without increasing the risk of gastroduodenal bleeding (Peto OR 1.67, 95% CI 0.41–6.80; P = 0.47) and superinfection (Peto OR 1.36, 95% CI 0.65–2.84; P = 0.41).
Results from this meta-analysis did not suggest a benefit for corticosteroids treatment in patients with CAP. However, the use of corticosteroids was associated with improved mortality in severe CAP. In addition, prolonged corticosteroids therapy suggested a beneficial effect on mortality. These results should be confirmed by future adequately powered randomized trials.
Citation: Nie W, Zhang Y, Cheng J, Xiu Q (2012) Corticosteroids in the Treatment of Community-Acquired Pneumonia in Adults: A Meta-Analysis. PLoS ONE 7(10): e47926. https://doi.org/10.1371/journal.pone.0047926
Editor: Miguel Santin, Barcelona University Hospital, Spain
Received: May 31, 2012; Accepted: September 18, 2012; Published: October 24, 2012
Copyright: © 2012 Nie et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This study was supported by grant no. 81170025 from the National Natural Science Foundation of China and projects of "Major New Drugs Innovation and Development" from the National Ministry of Science and Technology (no. 2011ZX09302-003-001). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Community-acquired pneumonia (CAP) is a common and serious infectious disease associated with high morbidity and mortality. It is the sixth leading cause of death and the most common infectious cause of death worldwide . Despite effective antibiotic therapy, about 12–36% patients admitted to the intensive care unit (ICU) with severe CAP die within a short time . Therefore, the development of an efficacious treatment has important implications for reducing the high mortality.
During infectious pneumonia, inflammatory cytokines, such as interleukin (IL)-6, IL-8 and IL-10 acted as acute phase proteins . A recent study showed that the excess of IL-6 and IL-10 was associated with a high mortality rate in CAP . Corticosteroids are the most effective and widely used anti-inflammatory drugs. An early study demonstrated the association of glucocorticoids with antibiotics attenuated local inflammatory response and decreased bacterial burden in the experimental model of severe pneumonia . In a mouse pneumonia model, Li et al.  found that hydrocortisone decreased inflammatory response significantly. In addition, Salluh et al.  reported that relative adrenal insufficiency occurred in most of the patients with severe CAP, suggesting underlying benefits of corticosteroids treatment in these patients. Taken together, these facts indicated a potential beneficial effect of corticosteroids in pneumonia.
Recently, a multicenter randomized controlled trial (RCT) performed by Confalonieri et al.  demonstrated that hydrocortisone treatment in severe CAP was associated with a significant reduction in mortality. A retrospective study conducted by Garcia-Vidal et al.  found that mortality decreased in the patients who received systemic steroids along with antibiotic treatment for severe CAP. Moreover, results from a systematic review showed that administration of corticosteroids in patients with CAP was associated with a lower mortality . However, these findings were not confirmed in the subsequent larger RCTs –. Another recent retrospective study also showed that adjunctive therapy with corticosteroids did not influence the mortality rate . Furthermore, results from a meta-analysis found that participants receiving corticosteroids displayed no significant differences in mortality compared with placebo . Consequently, the benefits of corticosteroids treatment in CAP are still uncertain.
The aim of this meta-analysis was to evaluate the efficacy and safety of corticosteroids adjunctive therapy in the treatment of CAP in adults.
Materials and Methods
This meta-analysis was performed according to a predetermined protocol described in the following paragraph, using standard systematic review techniques, as outlined by the Cochrane Handbook for Systematic Reviews of Interventions and PRISMA Statement , .
Eligibility Criteria and Outcome Measures
Studies fulfilling the following selection criteria were included in this meta-analysis: (1) study design: RCTs, including quasi-RCTs; (2) population: adult patients with CAP; (3) intervention: corticosteroids adjunctive therapy in CAP; (4) comparison intervention: placebo or standard treatment; and (5) outcome variables: mortality. Studies were excluded if: studies enrolled pediatric patients or nosocomial pneumonia patients.
The primary outcome measure was mortality. Secondary outcome measures included: hyperglycemia, gastroduodenal bleeding, and superinfection.
Information Sources and Search
A Pubmed, EMBASE and the Cochrane controlled trials register search identified RCTs of corticosteroids adjunctive therapy versus control participants published up to May 2012 in any language literature. The electronic search strategy included the terms “steroids”, “glucocorticoids”, “corticosteroids”, “hydrocortisone”, “prednisone”, “methylprednisolone”, “dexamethasone” and “community-acquired pneumonia”. We also searched the reference lists of original reports and systematic review of studies involving CAP to identify studies not yet included in the computerized databases. In addition, we reviewed the cited lists of eligible trials by Google Scholar to ensure that all appropriate studies were included.
A total of 772 studies were retrieved, and the process of identifying relevant trials is shown in Figure 1. Among the 772 initially potentially relevant studies, one study was searched from Google Scholar through reviewing the cited lists of eligible trials , and three studies were identified from reviews by hand searched –. 748 studies were excluded because of irrelevant, review or commentary articles. After full-text articles assessed for eligibility, fifteen studies were excluded because four studies enrolled pediatric patients and the other eleven studies were non-RCTs. Subsequently, nine RCTs were included in qualitative synthesis and quantitative synthesis.
Data Collection Process
Two authors independently reviewed full manuscripts of eligible trials, and the relevant data were extracted into predesigned data collection forms. We verified accuracy of data by comparing collection forms from each reviewer. Any discrepancy was resolved by discussion or a third author would assess these articles. The following data were collected from each study: first author, year of publication, study design, location, sample size, participant demographics, corticosteroids doses, duration of corticosteroids treatment and outcome variables. Authors of the included studies were contacted via E-mail if further study details were needed.
Methodological quality assessment was independently performed by two of the authors, and any disagreement was resolved by consensus. Risk of bias was evaluated as high, low, or unclear using the Cochrane Risk of Bias Tool for RCTs .
The outcome measure was assessed in an intention-to-treat (ITT) manner. Peto method works well when intervention effects are small, events are not particularly common and the study has similar numbers in two groups . Therefore, mortality and adverse events were analyzed using the Peto method to calculate odds ratio (OR) and 95% confidence interval (CI). Heterogeneity was assessed using the I2 test statistic and classified as low (≤25%), moderate (25–50%), and high (>50%). Subgroup analysis was performed by the severity of CAP and the duration of corticosteroids treatment. We defined the prolonged course as corticosteroids treatment for more than 5 days . Prespecified sensitivity analyses were conducted to determine the influence of statistical models (the fixed-effects model and the random-effects model) on effect size. Graphic exploration with funnel plot was used to evaluate the publication bias visually. The Egger’s test  was used to assess publication bias statistically. A P value <0.05 was considered statistically significant. All statistical tests were performed by using the RevMan 5.1 software (Nordic Cochrane Center, Copenhagen, Denmark) and STATA 11.0 software (Stata Corporation, College Station, TX).
Nine clinical trials including 1001 adult patients qualified for inclusion , –, , , , . 468 patients were allocated to the corticosteroids groups, and 533 patients were allocated to the control groups. Among these trials, five were conducted in Europe, two in North America, one in Africa, and one in Asia. The mean age of the patients ranged from 34 to 72 years. Four trials specifically included severe CAP patients , , , . The remaining five trials included those with mild to severe CAP patients , , , , . Four trials were multicenter RCTs , , , . The selected trials used various corticosteroids, including hydrocortisone, prednisolone, dexamethasone, and methyl-prednisolone. The durations of corticosteroids treatment ranged from 1 to 9 days. The characteristics of the trials are shown in Table 1.
We assigned an unclear risk of bias to one study  due to insufficient information regarding randomization and allocation. Despite two studies were open-label RCTs , , we assigned a low risk of bias to them, as the lack of blinding would be unlikely to affect mortality. Double-blinded RCTs were assigned to a low risk of bias. Table 2 summarizes the risk of bias.
Data on mortality was available from eight trials (n = 970). Mortality was not significantly reduced by the use of corticosteroids (Peto OR 0.62, 95% CI 0.37–1.04; P = 0.07) (Figure 2). There was low heterogeneity (I2 = 13%).
Four trials (n = 488) reported gastroduodenal bleeding events , , , . No significant difference was detected (Peto OR 1.67, 95% CI 0.41–6.80; P = 0.47; I2 = 0%). Three trials (n = 565) reported superinfection events , , . There was also no difference (Peto OR 1.36, 95% CI 0.65–2.84; P = 0.41; I2 = 71%). Data on hyperglycemia events was available for three trials (n = 573) –. Corticosteroids was associated with more hyperglycemia events (Peto OR 2.64, 95% CI 1.68–4.15; P<0.0001; I2 = 0%).
In the subgroup analysis by the severity (Figure 3), significant association was found among severe CAP patients and mortality (Peto OR 0.26, 95% CI 0.11–0.64; P = 0.003; I2 = 0%), but was not found among mild to severe CAP patients (Peto OR 0.95, 95% CI 0.50–1.78; P = 0.86; I2 = 0%). Subgroup analysis was also performed by duration of corticosteroids treatment. As shown in Figure 4, significant reduced mortality was found among patients with prolonged treatment (Peto OR 0.51, 95% CI 0.26–0.97; P = 0.04; I2 = 37%). However, the subgroup analysis of trials with shorter course failed to support the beneficial effect of corticosteroids in CAP (Peto OR 0.87, 95% CI 0.37–2.05; P = 0.75; I2 = 0%).
The sensitivity analysis using the fixed-effects model yielded estimates similar to those of the Peto odds ratio for the risk of mortality (RR 0.63, 95% CI 0.38–1.04; P = 0.07; I2 = 0%). The sensitivity analysis using a random-effects model yielded estimates similar to those of the Peto odds ratio for the mortality risk (RR 0.73, 95% CI 0.43–1.23; P = 0.23; I2 = 0%).
In this current meta-analysis of 9 RCTs including 1001 patients hospitalized for CAP, mortality was not significantly reduced by the use of corticosteroids. In the severe CAP subgroup, however, we observed a statistically significant decrease in mortality associated with the use of corticosteroids. There could be two potential reasons for the observed survival benefit. First, the immunomodulatory and anti-inflammatory effects of corticosteroids may damp or attenuate the inflammatory response due to severe CAP. Corticosteroids are the most powerful inhibitors of inflammation. They can switch on genes that encode anti-inflammatory cytokines and switch off genes that encode pro-inflammatory cytokines , . Second, in addition to the inflammatory response, the critical illness-related corticosteroid insufficiency (CIRCI) may play an important role in severe CAP. According to IDSA/ATS consensus guideline, CIRCI should be screened in all patients at risk for severe CAP . In a systematic review, Salluh et al.  showed that the prevalence of CIRCI in severe CAP ranged from 0% to 48%. Cortisol is a major regulator in the immune system and inflammation. Therefore, corticosteroid replacement therapy might be effective in critical illness, including severe CAP. In subgroup analysis by the duration of corticosteroids treatment, we found that prolonged corticosteroids treatment (>5 days) was associated with a greater benefit compared with treatment courses less than 5 days. Recently, Annane et al.  assessed the use of corticosteroids for severe sepsis and septic shock in a systematic review. They showed that hydrocortisone for a prolonged duration (>5 days) may improve survival in severe sepsis and septic shock . Moreover, a study on acute respiratory distress syndrome (ARDS) suggested that more than 7 days corticosteroids strategy led to reduction in markers of inflammation, duration of mechanical ventilation, and intensive care unit stay . Therefore, the beneficial effect of prolonged corticosteroids treatment course on decreasing mortality in CAP cannot be excluded.
The potential side effects from corticosteroids in CAP should be clarified. Theoretically, corticosteroids could favor the onset of metabolic disorders, gastroduodenal bleeding, muscle weakness, and superinfection. Previous study found that corticosteroids increased the risk of hyperglycemia and hypernatremia . In addition, there was no evidence for an increased risk of bleeding, superinfection, or neuromuscular weakness . In our meta-analysis, treatment with corticosteroids in CAP was associated with an increased risk of hyperglycemia, but was not associated with gastroduodenal bleeding and superinfection. However, we could not address the association between treatment with corticosteroids and risk of hypernatremia or neuromuscular weakness. It was due to insufficient information can be extracted from primary publications. Further studies should be designed to analyze these issues. Hyperglycemia occurred frequently in corticosteroids treated patients. Berghe et al.  indicated that intensive blood glucose control reduced morbidity and mortality among critically ill patients. Therefore, the latest IDSA/ATS guideline suggested that close attention to tight glucose control was required, if CAP patients received corticosteroids . A strict surveillance of control of blood glucose levels must be systematically conducted by physicians.
The results from meta-analyses could be influenced by publication bias. Egger’s test did not show significant publication bias. The asymmetric funnel plot, however, showed potential publication bias. Therefore, more studies are still needed to confirm the findings from this meta-analysis. Furthermore, there was no significant statistical heterogeneity in most of the comparisons. Thus, heterogeneity did not seem to have influenced the results. We also carried out a sensitivity analysis. The change of statistical models did not alter our conclusion of mortality, suggesting the reliability of this result.
Some limitations of this meta-analysis should be considered. First, the number of available studies that could be included in this meta-analysis was moderate. Therefore, the results could be influenced by the factors like random error. Second, this study was a study-level meta-analysis but not an individual patient-level meta-analysis. It is known that study-level analyses can lead to biased assessments and use of aggregated summary values has some limitations for explaining the heterogeneity , . Third, because only studies that were indexed by the selected databases were included for data analysis, some relevant published studies or unpublished studies were missed, which may have biased our results.
In summary, although our overall results did not suggest a benefit for corticosteroids treatment in patients with CAP, analyses restricted to severe CAP patients or prolonged corticosteroids treatment showed a survival benefit. However, the number of studies only including severe CAP patients was small (4 studies, n = 214). Thus, large-scale, double-blind, placebo-controlled trials are still needed to evaluate the effects of corticosteroids in adults with severe CAP. A sensitive and validated formula to identify those who need corticosteroids in CAP patients is warranted. In addition, type of corticosteroids, the dosage, the duration, and tapering of the treatment, should be evaluated in rigorously designed and adequately powered RCTs in the future.
Conceived and designed the experiments: WN QYX. Performed the experiments: WN YZ JWC. Analyzed the data: WN YZ JWC. Contributed reagents/materials/analysis tools: WN YZ JWC. Wrote the paper: WN YZ JWC QYX.
- 1. Bartlett JG, Mundy LM (1995) Community-acquired pneumonia. N Engl J Med 333: 1618–1624.
- 2. Fine MJ, Smith MA, Carson CA, Mutha SS, Sankey SS, et al. (1996) Prognosis and outcomes of patients with community-acquired pneumonia. JAMA 275: 134–141.
- 3. Endeman H, Meijvis S, Rijkers G, van Velzen-Blad H, van Moorsel C, et al. (2011) Systemic cytokine response in patients with community-acquired pneumonia. Eur Respir J 37: 1431–1438.
- 4. Martínez R, Menéndez R, Reyes S, Polverino E, Cillóniz C, et al. (2011) Factors associated with inflammatory cytokine patterns in community-acquired pneumonia. Eur Respir J 37: 393–399.
- 5. Sibila O, Luna CM, Agustí C, Baquero S, Gando S, et al. (2008) Effects of glucocorticoids in ventilated piglets with severe pneumonia. Eur Respir J 32: 1037–1046.
- 6. Li Y, Cui X, Li X, Solomon SB, Danne RL, et al. (2008) Risk of death does not alter the efficacy of hydrocortisone therapy in a mouse E. coli pneumonia model. Intens Care Med 34: 568–577.
- 7. Salluh JI, Verdeal JC, Mello GW, Araújo LV, Martins GAR, et al. (2006) Cortisol levels in patients with severe community-acquired pneumonia. Intens Care Med 32: 595–598.
- 8. Confalonieri M, Urbino R, Potena A, Piattella M, Parigi P, et al. (2005) Hydrocortisone infusion for severe community-acquired pneumonia: a preliminary randomized study. Am J Respir Crit Care Med 171: 242–248.
- 9. Garcia-Vidal C, Calbo E, Pascual V, Ferrer C, Quintana S, et al. (2007) Effects of systemic steroids in patients with severe community-acquired pneumonia. Eur Respir J 30: 951–956.
- 10. Siempos II, Vardakas KZ, Kopterides P, Falagas ME (2008) Adjunctive therapies for community-acquired pneumonia: a systematic review. J Antimicrob Chemoth 62: 661–668.
- 11. Fernandez-Serrano S, Dorca J, Garcia-Vidal C, Fernandez-Sabe N, Carratala J, et al. (2011) Effect of corticosteroids on the clinical course of community-acquired pneumonia: a randomized controlled trial. Crit Care 15: 1–9.
- 12. Meijvis SCA, Hardeman H, Remmelts HH, Heijligenberg R, Rijkers GT, et al. (2011) Dexamethasone and length of hospital stay in patients with community-acquired pneumonia: a randomised, double-blind, placebo-controlled trial. Lancet 377: 2023–2030.
- 13. Snijders D, Daniels JMA, De Graaff CS, Van Der Werf TS, Boersma WG (2010) Efficacy of corticosteroids in community-acquired pneumonia: a randomized double-blinded clinical trial. Am J Respir Crit Care Med 181: 975–982.
- 14. Sabry NA, Omar EED (2011) Corticosteroids and ICU Course of Community Acquired Pneumonia in Egyptian Settings. Pharmacology & Pharmacy 2: 73–81.
- 15. Salluh JI, Soares M, Coelho LM, Bozza FA, Verdeal JCR, et al. (2011) Impact of systemic corticosteroids on the clinical course and outcomes of patients with severe community-acquired pneumonia: a cohort study. J Crit Care 26: 193–200.
- 16. Chen Y, Li K, Pu H, Wu T (2011) Corticosteroids for pneumonia. Cochrane Database Syst Rev 3: CD007720 DOI: https://doi.org/10.1002/14651858.CD007720.pub2.
- 17. Higgins JPT, Green S (2011) Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0. [updated March 2011]. The Cochrane Collaboration. Available from www.cochrane-handbook.org.
- 18. Moher D, Liberati A, Tetzlaff J, Altman DG (2009) Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med 6: e1000097.
- 19. Wagner Jr HN, Bennett Jr IL, Lasagna L, Cluff LE, Rosenthal MB, et al. (1956) The effect of hydrocortisone upon the course of pneumococcal pneumonia treated with penicillin. Bull Johns Hopkins Hosp 98: 197–215.
- 20. McHardy VU, Schonell M (1972) Ampicillin dosage and use of prednisolone in treatment of pneumonia: co-operative controlled trial. BMJ 4: 569–573.
- 21. Kirby Jr JC, Polis G, Romansky MJ (1960) A controlled blind study of pneumococcal pneumonia treated with tetracycline and tetracycline plus 6-methyl prednisolone. Am J Med Sci 240: 64–69.
- 22. Bradburn MJ, Deeks JJ, Berlin JA, Russell Localio A (2007) Much ado about nothing: a comparison of the performance of meta-analytical methods with rare events. Stat Med 26: 53–77.
- 23. Annane D, Bellissant E, Bollaert PE, Briegel J, Confalonieri M, et al. (2009) Corticosteroids in the treatment of severe sepsis and septic shock in adults. JAMA 301: 2362–2375.
- 24. Egger M, Smith GD, Schneider M, Minder C (1997) Bias in meta-analysis detected by a simple, graphical test. BMJ 315: 629–634.
- 25. Marik P, Kraus P, Sribante J, Havlik I, Lipman J, et al. (1993) Hydrocortisone and tumor necrosis factor in severe community-acquired pneumonia. a randomized controlled study. Chest 104: 389–392.
- 26. Mikami K, Suzuki M, Kitagawa H, Kawakami M, Hirota N, et al. (2007) Efficacy of corticosteroids in the treatment of community-acquired pneumonia requiring hospitalization. Lung 185: 249–255.
- 27. Rhen T, Cidlowski JA (2005) Antiinflammatory action of glucocorticoids–new mechanisms for old drugs. N Engl J Med 353: 1711–1723.
- 28. Galon J, Franchimont D, Hiroi N, Frey G, Boettner A, et al. (2002) Gene profiling reveals unknown enhancing and suppressive actions of glucocorticoids on immune cells. FASEB J 16: 61–71.
- 29. Mandell LA, Wunderink RG, Anzueto A, Bartlett JG, Campbell D, et al. (2007) Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin Infect Dis 44 (Supplement 2)S27–S72.
- 30. Salluh JIF, Shinotsuka CR, Soares M, Bozza FA, Lapa e SilvaJR, et al. (2010) Cortisol levels and adrenal response in severe community acquired pneumonia: A systematic review of the literature. J Crit Care 25 541: e1–e8.
- 31. Meduri GU, Marik PE, Chrousos GP, Pastores SM, Arlt W, et al. (2008) Steroid treatment in ARDS: a critical appraisal of the ARDS network trial and the recent literature. Intens Care Med 34: 61–69.
- 32. Van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx F, et al. (2001) Intensive insulin therapy in the critically ill patients. N Engl J Med 345: 1359–1367.
- 33. Higgins J, Thompson SG (2004) Controlling the risk of spurious findings from meta-regression. Stat Med 23: 1663–1682.
- 34. Lambert PC, Sutton AJ, Abrams KR, Jones DR (2002) A comparison of summary patient-level covariates in meta-regression with individual patient data meta-analysis. J Clin Epidemiol 55: 86–94.