Discussion

We observed a high incidence of preoperative ESBL-producing E. coli colonization (32.3%) and a high rate of SSIs (7.3%) among patients undergoing elective orthopedic procedures. Preoperative ESBL-producing E. coli colonization was independently associated with the development of SSIs. These findings highlighted the potential clinical relevance of MDROs colonization in surgical populations in endemic regions. ^{Reference Apisarnthanarak, Kiratisin and Mundy14,Reference Apisarnthanarak, Kiratisin, Saifon, Kitphati, Dejsirilert and Mundy15} In our study, only one patient had a history of hospital admission within the prior year, suggesting that traditional healthcare-associated risk factors may not adequately capture the current epidemiology of MDROs colonization. In the settings with a high prevalence of antimicrobial resistance, community-acquired colonization with MDROs might be an important reservoir for subsequent infection. Increasing reports of community-based ESBL-producing GNB colonization among healthy individuals in endemic regions support the hypothesis that asymptomatic colonization may influence perioperative infectious risk, even in patients without established healthcare exposures. ^{Reference Karanika, Karantanos, Arvanitis, Grigoras and Mylonakis16,Reference Manyahi, Matee, Majigo, Moyo, Mshana and Lyamuya19,Reference Boonyasiri, Tangkoskul, Seenama, Saiyarin, Tiengrim and Thamlikitkul22–Reference Dayan, Dabbah, Weissman, Aga, Even and Glikman25}

A previous meta-analysis by Karanika et al. reported a global prevalence of fecal ESBL-producing GNB colonization of approximately 14% among healthy adults, with substantially higher rates in parts of the West Pacific and Southeast Asia, Africa, and the eastern Mediterranean (22%–46%). ^{Reference Karanika, Karantanos, Arvanitis, Grigoras and Mylonakis16} The colonization rate observed in our study was consistent with these regional estimates. Colonization with MDROs especially in multidrug-resistant GNB have been associated with subsequent infection in several clinical settings. ^{Reference Karanika, Kinamon, Grigoras and Milonakis26–Reference Apisarnthanarak, Kondo and Mingmalairak31} Data regarding the clinical impact of gastrointestinal colonization with multidrug-resistant GNB on postoperative infectious outcomes remain limited, particularly in patients undergoing elective surgical procedures. ^{Reference Tischendorf, de Avila and Safdar28–Reference Cheikh, Belefquih, Chajai, Cheikhaoui, El Hassani and Benouda30} However, few studies have been reported the risk for fecal multidrug-resistant GNB colonization and infection, particularly in patients with underlying hematologic malignancies. ^{Reference Tischendorf, de Avila and Safdar28,Reference Cornejo-Juarez, Suarez-Cuenca and Volkow-Fernandez29} Cheikh et al. performed a retrospective cohort study of pediatric cardiac surgery in Morocco. ESBL-producing GNB colonization was around 15% (17/111). Among the 17 patients, 23.5% (4/17) developed a postoperative infection due to ESBL-producing GNB. ^{Reference Cheikh, Belefquih, Chajai, Cheikhaoui, El Hassani and Benouda30} In Thailand, preoperative screening for ESBL-producing GNB colonization was conducted in patients underwent abdominal surgery. ESBL-producing GNB colonization and dirty wound classification were associated with SSIs. However, there was no association between carbapenem prophylaxis and reduction in SSIs. ^{Reference Apisarnthanarak, Kondo and Mingmalairak31}

Our study is the first study to assess the incidence of SSIs after elective orthopedic procedures in Thailand. We found a high rate of SSIs compared to CDC’s NHSH report (<3%). ^{Reference Rosenthal, Richtmann and Singh21} Our study supports the data from the International Nosocomial Infection Control Consortium (INICC) (resource-limited settings) that there are higher rates of SSIs postsurgery. The data from INICC has been collected from a multinational, multicenter, collaborative on health care associated control program that used a surveillance system. This system is in 30 countries in Latin America, Asia, Africa, and Europe which are MDROs endemic regions. For example, the rate of SSIs after hip prosthesis was 2.6% (INICC data) versus 1.3% (CDC’s NHSN data). ^{Reference Rosenthal, Richtmann and Singh21} Elective orthopedic procedures are considered as a clean procedure with low rate of SSIs. The rates of SSIs were higher in patients with certain risk factors based on the patient’s underlying disease, immune status, and procedure types. ^{Reference Edwards, Peterson and Mu17} High rate of SSIs in MDROs endemic region may be related to the virulence of the pathogen, antimicrobial resistance and lack of infection prevention and control program.

In our study, the association between ESBL-producing E. coli colonization and development of SSIs was consistent with previous studies. ^{Reference Cheikh, Belefquih, Chajai, Cheikhaoui, El Hassani and Benouda30,Reference Apisarnthanarak, Kondo and Mingmalairak31} However, we had only 7 patients with SSIs and limited microbiological culture confirmation. Three patients with SSIs who had re-hospitalization were response with empirical of board spectrum antibiotic treatment. Thus, we had insufficient evidence to clarify the process of SSIs development. Nevertheless, ESBL-producing GNB colonization may be enhanced or be the causative pathogen by hematological route from gut translocation, or direct contamination at the skin surface at surgical sites. Active surveillance screening of ESBL-producing GNB colonization for patients undergoing elective orthopedic surgery may have a significant contribution on the antibiotic perioperative prophylaxis and infection control implementation. However, the cost-effective analysis of the active surveillance screening and outcomes required further studies.

In our study, MDROs surveillance cultures recovered only ESBL-producing E. coli. CREs, CRAB, CRPA, MRSA, and VREs isolates were not recovered. ESBL-producing E. coli is the major community associated MDROs pathogen in Thailand which is consistent with previous studies. ^{Reference Suwantarat and Carroll11,Reference Kiratisin, Chongthaleong and Tan13–Reference Apisarnthanarak, Kiratisin, Saifon, Kitphati, Dejsirilert and Mundy15,Reference Rosenthal, Richtmann and Singh21,Reference Yungyuen, Chatsuwan and Plongla32} Although prevalence of CREs infections has been increasing in Thailand. It may be possible that it is still related to nosocomial pathogen. ^{Reference Hsu, Apisarnthanarak, Khan, Suwantarat, Ghafur and Tambyah10,Reference Suwantarat and Carroll11,Reference Yungyuen, Chatsuwan and Plongla32} The recent national surveillance molecular characterization of multidrug-resistant GNB from the Research University Network Thailand study reported the high prevalence of extended-spectrum cephalosporinase E. coli (14.5%) and the low prevalence of carbapenem-resistant E.coli (1.3%). ^{Reference Yungyuen, Chatsuwan and Plongla32} In our study, the patients with elective orthopedic surgery had no traditional risk of MDROs. Thus, the positive ESBL-E. coli colonization could be reflected the ESBL-producing E. coli colonization in the community setting which were higher than CRE colonization.

The global strains of E. coli ST 131, that carry the CTX-M gene may be the major strain of ESBL-producing E. coli isolate from our study. From our study, all ESBL-producing E. coli strains were highly susceptible to carbapenems. However, the control of carbapenems use should be considered. Thus, piperacillin/ tazobactam (in noncritical infections), amikacin, and Cefoperazone/sulbactam are considered as alternative regimen. The rectal swab culture revealed a great performance for detecting ESBL-producing E. coli colonization. Our study was also confirmed the low prevalence of community-acquired MRSA in Thailand. Active surveillance screening for MRSA in patients without risk factors for HAIs and anti-MRSA perioperative prophylaxis are not recommended.