Discussion
Tracheostomy and ventilator-dependent children are frequently prescribed broad spectrum empiric antibiotics for illnesses7; 11; 12. Frequent and prolonged use of antibiotics is associated with emergence of antibiotic resistance in numerous patient populations13-15, but there is a paucity of data specific to tracheostomy and ventilator-dependent children. Our study found that increased antibiotic usage and episodes of hospitalization put these patients at risk of MDROs. Other studies have found respiratory tract MDROs increase the risk of complications1; 8; 9; 16, but we did not find that our participants had more complications following MDRO detection.
We had a higher rate of MDROs in our patient population compared to other studies of tracheostomy and ventilation-dependent children15. 76% of our participants had at least one respiratory tract culture with a MDRO. Our participants were predominantly dependent on mechanical ventilation (93%) which could explain the higher rate of MDROs in our group15; 17; 18. MRSA was the most common MDRO within our patient population with 49% of participants positive. This was consistent with a recent publication by McCaleb et al. that found a similar rate of MRSA within their single center study of pediatric tracheostomy patients19. S. aureus and Pseudomonas aeruginosa are the most commonly cultured organisms from tracheostomies of pediatric patients8; 9; 18; 19. Russell et al. found that tracheostomies in pediatric patients increase risk of Gram-negative organisms such as P. aeruginosa but not S. aureus 8 . It is suspected that organisms found in the normal flora of the respiratory tract, like S. aureus , are protective against lower respiratory tract infections but organisms such as Pseudomonas disrupt the normal flora which can lead to more infections 20. Gram-negative organisms are clearly associated with an increased risk of lower respiratory tract infections, but it is less clear if MRSA cultured from the airway represents colonization versus evidence of a lower respiratory tract infection.
We found that frequent hospitalizations and antibiotic usage increased the risk of MDRO in the respiratory tract of technology-dependent children. It has been shown that hospitalizations and prior antibiotic usage were independent risk factors for lower respiratory tract MDROs in adults, but has not been well described in children15. We also found that both intermittent antibiotic usage and chronic antibiotics increased risk of MDRO. We did not find that there was an independent effect from type of chronic antibiotic used, inhaled or oral, on risk of MDRO positivity. The literature is unclear about emergence of antibiotic resistance with inhaled antibiotics. Long term use of inhaled antibiotics in cystic fibrosis patients increases the mean inhibitory concentration (MIC) over time but does not affect the clinical efficacy of antibiotic treatment21-23. Furthermore, inhaled antibiotics in cystic fibrosis have not been found to select for intrinsically resistant organisms 21-23. This differs from a review by Laska et al. which demonstrated that use of inhaled antibiotics increased the risk of resistant organisms in patients with bronchiectasis24. There remains little literature describing emergence of antibiotic resistance with inhaled antibiotics in the pediatric tracheostomy and ventilator-dependent population.
Our study did not find having an MDRO increased the risk of complications including number of hospitalizations, ED visits, or antibiotic usage compared to those without MDRO. This finding contrasts several papers that found having resistant Gram-negative bacilli orPseudomonas to be a risk factor for increased length of stay, increased number of lower respiratory tract infections, and mortality1; 8; 9; 16. In our analysis without MRSA, Gram-negative bacilli were the most common type of resistant organism cultured. We did not find an increased risk of complications following MDRO positivity with the exclusion of MRSA from our analysis. We found that participants received more outpatient antibiotics after MDRO positivity compared to prior to MDRO positivity using a pair-wise comparison of these groups. This increased use of antibiotics could be evidence of more frequent respiratory tract infections following MDRO detection. It has been shown that specific organisms such as P. aeruginosa , Enterobacteriaceae, or Acinetobacter baumannii in tracheal cultures are associated with an increased number of lower respiratory tract infections 7; 8. The increased use of antibiotics seen following MDRO detection in our group may also represent physicians having a lower threshold to treat with antibiotics when an MDRO is found. Even when an MDRO is cultured from the respiratory tract it is hard to determine if the organism represents colonization or a true infection. One of the challenges in treating tracheostomy-dependent children is defining true infection and treating appropriately.
There are several limitations to this study. Our study is a single center retrospective chart review, so the data are less generalizable. This study is also limited by a relatively small sample size which may lead to underpowered statistical analyses. There is considerable variability in care provided including when tracheal aspirate cultures were obtained, when antibiotics were used, and thresholds for hospital admission. Data were limited only to what could be accessed in our electronic health medical record. We aimed to maximize the data collected for each patient by manually extracting data instead of using automated data collection software.
Despite these limitations our study offers insight into a less-studied pediatric technology-dependent patient population. Tracheostomy and ventilator-dependent children represent a group of patients with complex medical needs and high health care utilization. Respiratory tract infections including tracheitis and pneumonia are the most common cause of hospitalization within this population3-5. There are no guidelines published for treatment and management of respiratory tract infections in this patient population. Broad spectrum empiric antibiotics are frequently used to treat respiratory tract infections in these children but could increase risk of MDRO acquisition. Our study suggests that more antibiotic usage and hospitalizations may increase risk of MDRO positivity. This specific group of patients could benefit from increased antibiotic stewardship focus with a tailored antibiotic strategy. Additional studies are needed to help develop guidelines for treatment of respiratory tract infections in tracheostomy and ventilator-dependent children with the goal of minimizing the emergence of MDROs.
Acknowledgments : We thank Hongjie Gu, MS and Michael Wallendorf, PhD who provided statistical analysis through Division of Biostatistics at Washington University School of Medicine.