Mycoplasma pneumoniae is a ubiquitous organism with evidence of past exposure in up to 80% of the population by young adulthood.1M. pneumoniae infections causes a wide variety of respiratory and extrapulmonary manifestations, the most common being atypical community-acquired pneumonia. Other presentations include status asthmaticus and peripheral or central nervous system pathology.1–3M. pneumoniae infections are responsible for up to 40% of community-acquired pneumonia episodes in school-age children.1,3 Current empiric treatment guidelines, therefore, recommend macrolide antibiotics for children older than 5 years of age.4 Neurologic disease is present in up to 11.5% of patients hospitalized because of M. pneumoniae infections and can be severe leading to pediatric intensive care unit (PICU) admission.
There is a paucity of data on the role of M. pneumoniae infections in critically ill children requiring PICU admission. The lack of population-based studies describing the burden of severe M. pneumoniae infections in children poses a challenge in defining optimal empiric treatment in children presenting with community-acquired life-threatening pneumonia.
We sought to describe the incidence, characteristics and outcomes of critically ill children diagnosed with confirmed M. pneumoniae infection in 2 PICUs providing statewide pediatric intensive care services in Queensland, Australia.
MATERIALS AND METHODS
Statewide multicenter retrospective linkage study of all children younger than 16 years requiring PICU admission in Queensland, Australia, from January 1, 2008, until December 31, 2013.5 During this 6-year period, 2 PICUs provided pediatric intensive care services for the state of Queensland and northern New South Wales, capturing a pediatric population of 987,481 (2015 census). The institutional review board approved the protocol including waiver of informed consent. Further details are provided in the Online Supplementary Methods (http://links.lww.com/INF/D139).
Patients with severe M. pneumoniae infection were defined as children admitted to the PICU with both positive laboratory results and a clinical presentation consistent with M. pneumoniae infection. During the study period, it was routine practice at both tertiary institutions to request testing for Mycoplasma if the patients had symptoms suggestive of Mycoplasma infection. Cases were reviewed by 2 clinicians (K.M.M. and L.J.S.). We searched the institutional microbiology databases for test results in all patients admitted to PICU, within a time window ranging from 72 hours before PICU admission to day of PICU discharge.
In addition, the prospective institutional PICU databases were searched for M. pneumoniae infections using the Australia and New Zealand Pediatric Intensive Care Registry–specific diagnostic coding.5 In patients coded as positive for M. pneumoniae infection who had no laboratory evidence of infection in the initial result search, we expanded the search for microbiologic results outside the defined time window to confirm evidence of M. pneumoniae infection by either polymerase chain reaction (PCR) or serologic testing (requiring titers >1:160).
Length of PICU stay was defined as the primary outcome. M. pneumoniae cases were compared with PICU patients admitted non-electively with an infection. Subgroups were compared using the Mann–Whitney U test, and proportions were compared using Fisher exact or χ2 test. Logistic and linear regressions were used to explore associations between binary and log-transformed continuous outcomes. For multivariate analyses, we adjusted for factors that showed a trend for difference between groups (P < 0.10). All analyses were conducted using SPSS, version 22, IBM statistics.
During the study period, 11,526 children were admitted to PICU, including 3005 nonelective admissions with infection as leading cause. A total of 708 M. pneumoniae PCR tests were performed in 650 PICU patients, including 428 patients admitted with any infection. Three patients coded as having M. pneumoniae infection in the PICU database had negative microbiologic results and were excluded. A total of 30 M. pneumoniae cases meeting both clinical and laboratory criteria were included, representing 1.0% of all nonelective infection-related admissions. Most cases were confirmed via PCR (n = 22, 73%), and 8 (27%) had serology titers greater than the defined threshold of 1:160.
Of the confirmed M. pneumoniae cases, median age was 7 years, with 23% presenting younger than 5 years. Twenty-four (80%) presented with respiratory symptoms including 22 (73%) with pneumonia, 6 with asthma and 2 diagnosed with sepsis in addition to respiratory pathology. The remaining 6 (20%) patients had neurologic diagnoses, including 5 with encephalitis and 1 with Guillain–Barré syndrome. Two patients with M. pneumoniae infection (6.7%) died. The deaths were unlikely related to Mycoplasma infection (leukemia with septic shock in 1 patient; trisomy 21 and congenital cardiac disease with cardiac arrest in 1 patient).
Characteristics and outcomes of children with confirmed M. pneumoniae infection requiring PICU admission were compared with both children with negative results on testing for M. pneumoniae and all PICU admissions with an infectious etiology (Table 1). Children with M. pneumoniae infection were older but presented with similar illness severity. Outcomes, including mortality, PICU length of stay, and length of respiratory support were comparable to children requiring PICU admission caused by other infectious etiologies (P > 0.10; Table 1).
While M. pneumoniae represents a common cause of community-acquired pneumonia, the epidemiology of severe infections causing life-threatening disease in children has received less attention. This retrospective study including over 3000 infection-related PICU admissions identified a low incidence of microbiologically confirmed M. pneumoniae infections representing only 1% of infection-related PICU admissions and 0.3% of all PICU admissions. Greater than 95% of PCR tests performed were negative. Severity and PICU outcomes of M. pneumoniae infections were comparable to other infectious etiologies. In our population, M. pneumoniae infections in critically ill children are uncommon, which should be considered when reviewing recommendations for empiric antimicrobial treatment.
To the best of our knowledge, this is the largest population-based study investigating the incidence and outcomes of M. pneumoniae infections in children admitted to PICU. The 2 PICUs included in this study admitted >90% of all critically ill children younger than 16 years of age in Queensland, capturing a pediatric population of approximately 1 million children.5 Previous studies reported ICU admission rates for pediatric patients with M. pneumoniae infections ranging from 4.6%–16%.2 The low rate of confirmed M. pneumoniae infection among critically ill children with respiratory infections requiring PICU admission contrasts with much higher rates reported in mild disease. Notably, disease severity and outcomes of children with M. pneumoniae infection in our PICU cohort were comparable with common infections in PICU.
Accurate determination of the impact of M. pneumoniae infections on patient outcomes is challenging because of the high rates of asymptomatic infections and technical limitations of currently available microbiologic diagnostic methods.2,6 Single serologic tests have low sensitivity and specificity and seroconversion or a 4-fold titer increase is superior for diagnosis. While increased IgM levels are observed in acute infections, elevated levels may persist for months. Over recent years, PCR tests have, therefore, largely replaced antibody-based M. pneumoniae testing in most institutions, including ours. PCR assays are highly sensitive but may overestimate rates of infection by detecting asymptomatic carriage in healthy children or postinfectious shedding as this organism can persist for variable periods after the acute infection.6 Our findings show that even in children where clinicians suspected Mycoplasma, only 7% were found to be positive. While prospective studies testing all PICU admissions would be required to provide data on prevalence, our findings indicate that Mycoplasma represents a rare cause of life-threatening infection leading to PICU admission.
The case definition employed in the present study required the presence of both positive microbiologic test results and a clinical presentation consistent with M. pneumoniae infection. Using linkage of individual patient data from the prospective PICU databases with the prospective microbiologic databases, we were able to confirm episodes coded as M. pneumoniae infection with the corresponding laboratory test result. However, a number of limitations have to be considered. Universal testing for M. pneumoniae infection was not performed, and given that only 12.5% of patients admitted with any infection were considered by clinicians to warrant Mycoplasma testing, this may result in an underestimation of the rate of mycoplasma infections in critically ill children. In addition, the findings are specific to the population studied, and a higher incidence of M. pneumoniae infections may be found in other geographical settings.
Our findings have potential implications for reviewing antimicrobial stewardship. Most guidelines recommend adding empiric macrolide antibiotics for children older than 5 years of age presenting with severe community-acquired pneumonia, as atypical organisms are frequently thought to be commonly causative.4 The routine use of empiric macrolide therapy is increasingly being questioned, with systematic reviews on community-acquired pneumonia revealing insufficient evidence to support specific treatment of mycoplasma infections.7,8 In addition to exposing children to risks caused by side effects of macrolide antibiotics, unnecessary macrolide use may contribute to increasing macrolide resistance in M. pneumoniae and selection of macrolide-resistant Streptococci.9,10 For life-threatening severe respiratory infections, other microbial agents offer better tissue penetration and bactericidal activity against common pathogens such as streptococci and staphylococci.
Given the low incidence of M. pneumoniae infection in critically ill children, our results question whether empiric macrolide treatment for children with life-threatening respiratory infections is justified. With the availability of contemporary rapid PCR testing, antimicrobial treatment directed specifically at M. pneumoniae may best be initiated on receipt of positive test results, unless the clinical presentation is strongly suggestive of M. pneumoniae infection.
1. Atkinson TP, Balish MF, Waites KB. Epidemiology, clinical manifestations, pathogenesis and laboratory detection of Mycoplasma pneumoniae
infections. FEMS Microbiol Rev. 2008;32:956–973.
2. Hon KL, Leung AS, Cheung KL, et al. Typical or atypical pneumonia and severe acute respiratory symptoms in PICU. Clin Respir J. 2015;9:366–371.
3. Jain S, Williams DJ, Arnold SR, et al; CDC EPIC Study Team. Community-acquired pneumonia requiring hospitalization among U.S. children. N Engl J Med. 2015;372:835–845.
4. Bradley JS, Byington CL, Shah SS, et al; Pediatric Infectious Diseases Society and the Infectious Diseases Society of America. The management of community-acquired pneumonia in infants and children older than 3 months of age: clinical practice guidelines by the Pediatric Infectious Diseases Society and the Infectious Diseases Society of America. Clin Infect Dis. 2011;53:e25–e76.
5. Moynihan K, Barlow A, Alphonso N, et al. Impact of viral respiratory pathogens on outcomes after pediatric cardiac surgery. Pediatr Crit Care Med. 2017;18:219–227.
6. Zhang L, Zong ZY, Liu YB, et al. PCR versus serology for diagnosing Mycoplasma pneumoniae infection
: a systematic review & meta-analysis. Indian J Med Res. 2011;134:270–280.
7. Gardiner SJ, Gavranich JB, Chang AB. Antibiotics for community-acquired lower respiratory tract infections secondary to Mycoplasma pneumoniae
in children. Cochrane Database Syst Rev. 2015;1:CD004875.
8. Williams DJ, Edwards KM, Self WH, et al. Effectiveness of β-lactam monotherapy vs macrolide combination therapy for children hospitalized with pneumonia. JAMA Pediatr. 2017;171:1184–1191.
9. Serisier DJ, Martin ML, McGuckin MA, et al. Effect of long-term, low-dose erythromycin on pulmonary exacerbations among patients with non-cystic fibrosis bronchiectasis: the BLESS randomized controlled trial. JAMA. 2013;309:1260–1267.
10. Yamada M, Buller R, Bledsoe S, et al. Rising rates of macrolide-resistant Mycoplasma pneumoniae
in the central United States. Pediatr Infect Dis J. 2012;31:409–400.