Identification of tachypnea and subcostal retractions as clinical signs for the diagnosis of community-acquired pneumonia among children: systematic review and meta-analysis

Determine whether tachypnea and subcostal retractions can be efficient predictors for the diagnosis of Community-Acquired Pneumonia (CAP) among children. These were the databases used: PubMed, LILACS, The African Journal Database and The Cochrane Central Library. Original studies were included which assessed the diagnostic performance of the clinical criteria for tachypnea or subcostal retraction defined in accordance with the criteria of the World Health Organization (OMS) for CAP diagnosis in children ≤ 5 with cough and fever. The likelihood ratio (LR), the diagnosis odds ratio (DOR), and their respective confidence intervals at 95% (IC95%) were estimated for each clinical test. 975 studies were found, eight were included in the review. 4740 patients were enrolled and 3584 (75%) were analyzed, 916 (19%) of which had a CAP diagnosis. When data were combined, tachypnea had a positive LR of 3.16, (95% CI 2.11-4.73) and a negative LR of 0.36 (95% CI 0.23-0.57). The DOR was 10.63 (95% CI 4.4-25.66, I2=93%). When subcostal retractions were analyzed, a positive LR of 2.49 (95%CI 1.41-4.37) and a negative LR of 0.59 (95%CI 0.4-0.87) were obtained. The DOR was 5.32 (95%CI 1.88-15.05, I2=89%). The presence or absence of tachypnea and subcostal retractions can be used in CAP diagnosis; it is worth considering the relative uncertainty in its diagnostic power and relatively modest LR. The confidence of these results is low due to the inadequate quality of the related evidence.