Prospective Study of the Detection of Bacterial Pathogens in Pediatric Clinical Specimens Using the Melting Temperature Mapping Method

The melting temperature (T(m)) mapping method is a novel technique that uses seven primer sets without sequencing to detect dominant bacteria. This method can identify pathogenic bacteria in adults within 3 h of blood collection without using conventional culture methods. However, no studies have examined whether pathogenic bacteria can be detected in clinical specimens from pediatric patients with bacterial infections. Here, we designed a new primer set for commercial use, constructed a database with more bacterial species, and examined the agreement rate of bacterial species in vitro. Moreover, we investigated whether our system could detect pathogenic bacteria from pediatric patients using the T(m) mapping method and compared the detection rates of the T(m) mapping and culture methods. A total of 256 pediatric clinical specimens from 156 patients (94 males and 62 females; median age, 2 years [<18 years of age]) were used. The observed concordance rates between the T(m) mapping method and the culture method for both positive and negative samples were 76.4% (126/165) in blood samples and 79.1% (72/91) in other clinical specimens. The T(m) mapping detection rate was higher than that of culture using both blood and other clinical specimens. In addition, using the T(m) mapping method, we identified causative bacteria in pediatric clinical specimens quicker than when using blood cultures. Hence, the T(m) mapping method could be a useful adjunct for diagnosing bacterial infections in pediatric patients and may be valuable in antimicrobial stewardship for patients with bacterial infections, especially in culture-negative cases. IMPORTANCE This study provides novel insights regarding the use of the melting temperature (T(m)) mapping method to identify the dominant bacteria in samples collected from pediatric patients. We designed a new set of primers for commercial use and developed a database of different bacteria that can be identified using these primers. We show that the T(m) mapping method could identify bacteria from blood samples and other clinical specimens. Moreover, we provide evidence that the T(m) mapping method has a higher detection rate than that of the culture-based methods and can achieve a relatively high agreement rate. We believe that our study makes a significant contribution to this field because rapid identification of the source of bacterial infections can drastically improve patient outcomes and impede the development of antibiotic-resistant bacteria.