Extensively Drug-Resistant Carbapenemase-Producing Pseudomonas aeruginosa and Medical Tourism from the United States to Mexico, 2018–2019

Discussion

We describe a large, prolonged outbreak of XDR VIM-CRPA among US medical tourists who underwent bariatric surgery in Tijuana, Mexico. Most isolates were clonal and linked to a surgeon who operated at multiple healthcare facilities; we also identified isolates genetically distinct from this outbreak strain and associated with other healthcare facilities and surgeons. Although serious complications from laparoscopic sleeve gastrectomy in the United States are uncommon (≈1%–2%),^[31,32] >40% of case-patients in our investigation required postoperative hospitalization in the United States, highlighting the severity of infections. Active outreach to exposed persons accounted for one third of case-patients identified. Our investigation underscores the potential for medical tourism to introduce highly concerning pathogens into the US healthcare system.

Several lines of epidemiologic and laboratory findings in this investigation support a point source outbreak linked to surgeon 1, a surgeon specializing in bariatric surgery who operated primarily at facility A, although the exact source of pathogen was not identified. VIM-CRPA infections appeared to increase starting in September 2018 and decreased in March 2019 after travel agency A notified exposed clients and stopped referrals to facility A. Clonal strains isolated from case-patients after surgery performed by surgeon 1 across multiple facilities, and infection control lapses at facility A led us to hypothesize a persistently contaminated mobile medical device; a laparoscope transported between facilities with surgeon 1 was a plausible outbreak source. P. aeruginosa is known to persistently colonize medical devices, including flexible endoscopes;^[33,34] in Brazil, surgeons transporting their own laparoscopic equipment between different hospitals were the suspected source of a multifacility Mycobacterium spp. Outbreak.^[35] Alternative explanations include a widespread persistently contaminated environmental or water source at facility A or a persistently colonized healthcare worker, such as surgeon 1.

We also identified case-patients with VIM-CRPA who were not epidemiologically linked to facility A or surgeon 1 and isolates that were genetically distinct from the outbreak cluster. These infections appeared to be sporadic. Although 2 of these case-patients underwent surgery at the same facility, their isolates were not more closely related to each other than to those from case-patients who underwent surgery at other facilities, decreasing suspicion for a second outbreak. Similar to most isolates linked to procedures performed by surgeon 1, these sporadic cases belonged to ST111, which has been associated with epidemic spread of carbapenemases in P. aeruginosa globally.^[36] Since July 2020, CDC has received 6 additional reports of VIM-CRPA cases among US residents who had undergone elective invasive medical procedures in Tijuana, none of whom were reported to have a common procedure, facility or surgeon; however, 1 case-patient was operated on by surgeon 1. These recent infections underscore the potential for US residents to acquire highly resistant bacteria when receiving medical care abroad, even in the absence of a recognized outbreak. In some countries, MDR organisms rarely identified in the United States may be more common, increasing the potential for acquiring resistant organisms, regardless of quality of care. Persons considering medical tourism and US healthcare providers caring for prospective or returned medical tourists should be aware that standards for infection control, as well as regulations and enforcement practices, vary by country and facility.^[37] US public health authorities and healthcare providers might have limited access to information to inform recommendations for follow-up care or testing for medical tourists.

In the United States, carbapenemases are rarely the cause of carbapenem resistance in P. aeruginosa, and few clinical laboratories perform carbapenemase testing for CRPA. Despite increased carbapenemase testing for CRPA through the Antibiotic Resistance Laboratory Network, our investigation shows CP-CRPA continues to be underdetected. Nearly 1 in 3 cases in this investigation represent CRPA clinical isolates that were not tested for carbapenemases, despite being highly resistant and identified from patients who had medical histories of concern during a well-publicized outbreak.^[22] CP-CRPA are overwhelmingly MDR and often XDR.^[38] Identification of these antimicrobial susceptibility testing phenotypes, especially in patients with a history of healthcare outside the United States, should increase suspicion for CP-CRPA.

Despite warnings from US public health agencies, medical tourists continued to undergo surgery at facility A during January 1–March 1, 2019. Nearly 30% of interviewed persons who had surgery were aware of the outbreak or negative news stories associated with facility A before their surgery; however, interviewed persons might have made travel or surgery reservations and deposits before issuance of travel health notices, which could have influenced their decisions to proceed with the procedure. Consistent with a 2015 survey of bariatric medical tourists from Canada, we found primary motivations for bariatric medical tourism among interviewees included shorter wait times and lower cost.^[39] A qualitative study from Canada also showed that bariatric medical tourists identified the Internet as a primary source of information for identifying providers and validating decisions to engage in medical tourism.^[40] Difficulty reconciling conflicting information sources might have delayed the effect of the CDC travel warnings.

Our investigation had several limitations. Because of limited data from the outbreak facility, its international setting, and lack of environmental cultures, we could not determine the outbreak source, although several hypotheses were considered. Additional cases might have gone undetected for 2 reasons. First, CP-CRPA is underdetected because of low suspicion of the potential for CRPA to harbor carbapenemases and limited availability of testing for carbapenemases. Second, active outreach was limited in several ways: only referrals from travel agency A, rather than all surgical patients at facility A, were available to US public health authorities; we focused efforts on persons who were at greatest risk for having current or new-onset infections, but <50% of targeted persons were reached. Because of high nonresponse rates and underdetection of CP-CRPA, our calculated attack rate during January–March 2019 is probably a lower bound; however, additional patients could have undergone surgery who were not included on our list, thereby overestimating the attack rate. Third, although transmission to household contacts of case-patients was not identified, this transmission was not routinely assessed for all case-patients. Fourth, persons interviewed might have been more likely to have infections compared with other facility A patients and possibly differed in their motivations for medical tourism and awareness of public health notifications, and might not be representative of all facility A patients or Tijuana bariatric medical tourists.

In this investigation, epidemiologic and molecular data link a single surgeon, performing surgeries at multiple facilities, to a prolonged outbreak among medical tourists. US patients and providers should be aware of the risk for colonization and infection with highly resistant pathogens not commonly encountered in the United States after medical tourism.

Emerging Infectious Diseases. 2022;28(1):51-61. © 2022 Centers for Disease Control and Prevention (CDC)