Summary: | <i>Mycoplasma pneumoniae (Mp)</i> is a common cause of lower respiratory tract infection (LRTI) in children that is difficult to distinguish from LRTI of other etiologies. We aimed to determine if a combination of clinical, laboratory, and chest radiographic features can help identify patients at higher risk of <i>Mp</i> LRTI. We reviewed medical charts of children referred to our tertiary hospital with suspected acute mycoplasmal LRTI. Pharyngeal swabs obtained from patients were tested by <i>Mp</i> PCR. We compared epidemiological and clinical data of children with positive and negative <i>Mp</i> PCR results. In addition, a multivariable logistic regression analysis was performed to predict <i>Mp</i> LRTI based on the patient’s age, duration of symptoms, presence of extrapulmonary manifestations, laboratory findings, and chest radiographic findings. We included 65 children with <i>Mp</i> PCR-negative and 49 with <i>Mp</i> PCR-positive LRTI and no viral co-detection. Children with <i>Mp</i> LRTI were older (median age 5.8 vs. 2.2 years, <i>p</i> < 0.001), had a longer duration of symptoms on referral (median 7 vs. 4 days, <i>p</i> < 0.001), and lower median WBC (9.9 vs. 12.7 × 10<sup>9</sup>/L, <i>p</i> < 0.001). On chest radiograph, unilateral infiltrates were more frequently observed in the <i>Mp</i> PCR-positive group (57.5% vs. 24.1%, <i>p</i> = 0.001). Age, duration of symptoms, and chest radiographic findings had the highest predictive value for <i>Mp</i> LRTI in a multivariable logistic regression model. Our analysis suggests that a combination of clinical, laboratory, and chest radiographic features can be used to assess the likelihood of <i>Mp</i> LRTI and assist in decision-making for which children need further tests or macrolide antibiotic treatment.
|