Usefulness of sputum gram stain for etiologic diagnosis in community-acquired pneumonia: a systematic review and meta-analysis.

BACKGROUND: Implementation of sputum Gram stain in the initial assessment of community-acquired pneumonia (CAP) patients is still controversial. We performed a systematic review and meta-analysis to investigate the usefulness of sputum Gram stain for defining the etiologic diagnosis of CAP in adult patients. METHODS: We systematically searched the Medline, Embase, Science Direct, Scopus and LILACS databases for full-text articles. Relevant studies were reviewed by at least three investigators who extracted the data, pooled them using a random effects model, and carried out quality assessment. For each bacterium (Streptococcus pneumoniae, Haemophilus influenzae, Staphylococcus aureus, and Gram-negative bacilli), pooled sensitivity, specificity, positive and negative likelihood ratios were reported. RESULTS: After a review of 3539 abstracts, 20 articles were included in the present meta-analysis. The studies included yielded 5619 patients with CAP. Pooled sensitivity and pooled specificity of sputum Gram stain were 0.59 (95% CI, 0.56-0.62) and 0.87 (95% CI, 0.86-0.89) respectively for S. pneumoniae, 0.78 (95% CI, 0.72-0.84) and 0.96 (95% CI, 0.94-0.97) for H. influenzae, 0.72 (95% CI, 0.53-0.87) and 0.97 (95% CI, 0.95-0.99) for S. aureus, and 0.64 (95% CI, 0.49-0.77) and 0.99 (95% CI, 0.97-0.99) for Gram-negative bacilli. CONCLUSION: Sputum Gram stain test is sensitive and highly specific for identifying the main causative pathogens in adult patients with CAP. TRIAL REGISTRATION: This study has been registered at PROSPERO International prospective register of systematic reviews under registration no. CRD42015015337 .

Biomarkers for predicting short-term mortality in community-acquired pneumonia: A systematic review and meta-analysis.

OBJECTIVES: The pneumonia severity index and CURB-65 are risk assessment tools widely used in community-acquired pneumonia (CAP). However, limitations in these prognostic scores have led to increasing interest in finding biomarkers that might provide additional information. To date, the role of these biomarkers has not been fully elucidated. METHODS: We systematically searched the Medline, Web of Knowledge, Science Direct, and LILACS databases. We included studies that assessed the accuracy of biomarkers for the prediction of in-hospital or ≤30-day mortality, in hospitalized adults with CAP. Two independent investigators extracted patient and study characteristics, which were thereafter pooled using a random effects model. Relationships between sensitivity and specificity of biomarkers and prognostic scores were plotter using the area under the receiver operator characteristic curve (AUC). RESULTS: We included 24 articles and 2 databases from 1069 reviewed abstracts, which provided 10,319 patients for analysis. Reported mortality rates varied from 2.4% to 34.6%. The highest AUC values for predicting mortality were associated with pro-adrenomedullin (0.80) and prohormone forms of atrial natriuretic peptide (0.79), followed by cortisol (0.78), procalcitonin (0.75), copeptin (0.71), and C-reactive protein (0.62). There were no statistically significant differences between the AUCs of the studied biomarkers, other than for copeptin and C-reactive protein, which performed comparatively poorly. When compared with the CAP-specific scores, the AUCs were not significantly different from those of most biomarkers. CONCLUSIONS: The identified biomarkers are able to predict mortality with moderate to good accuracy in CAP. However, biomarkers have no clear advantage over CAP-specific scores for predicting mortality.