We evaluated the Japanese surveillance system, NESID, for surveillance of MDRP infections quantitatively and qualitatively mainly based on the CDC guidelines [25]. Although current surveillance fairly captured the trend of MDRP infections in Japan, the surveillance should be designed based on actions and goals for the control of drug-resistant P. aeruginosa infections in the country.
The number of accurate MDRP infections occurring at TDSSs decreased over time, as did the number of cases reported as MDRP infections. Moreover, NESID showed good representativeness because TDSSs uniformly existed in each prefecture according to the population [16, 17], and there was no geographical bias was detected from our findings (Figure 1). Thus, we considered that the decreasing trend in surveillance data in TDSSs reflected the actual trend of MDRP infections across Japan. Additionally, data from JANIS, another national-level AMR surveillance system in Japan, also showed a decreasing trend of MDRP detection in inpatient settings from 0.05% in 2017 to 0.03% in 2022, which may also reflect the actual domestic situation due to its high coverage of hospitals (close to one-fourth of all hospitals participated in 2022) in Japan [15, 19]. Consequently, the original objective of NESID to assess the trend of MDRP infections has been achieved, and we confirmed from the surveillance data that the number of MDRP infections in Japan has been decreasing.
We learned through the interviews that some of the interviewees were of the opinion that the current surveillance system does not meet public health needs. Thus, the current situation should be reviewed to improve the surveillance system to meet new objectives in the control of drug-resistant P. aeruginosa infections nationwide. Drug-resistant P. aeruginosa was classified into CRPA, MDRP, extensively drug-resistant P. aeruginosa (XDRP), and pan drug-resistant P. aeruginosa (PDRP) according to the standardized international terminology published in 2012 [26]. Furthermore, difficult-to-treat resistance P. aeruginosa (DTRP) was defined in 2018, with a focus on the lack of effective first-line drugs [27-29]. As drug-resistant P. aeruginosa has been classified in several ways, each country focuses on different types [30]. In the AMR NAPs published in 132 countries [31], 51 countries (38.6%) mentioned P. aeruginosa, and 10 countries (7.6%) included it in their numerical benchmarks; eight countries targeted CRPA, while two countries targeted MDRP. No NAPs were targeting XDRP, PDRP, or DTRP at the end of 2023 (Supplementary Table 2).
The important points to be discussed in designing a surveillance system for drug-resistant P. aeruginosa are the definition of drug resistance including antimicrobial susceptibility testing and whether to conduct notifiable diseases surveillance or sentinel surveillance (Table 4).Regarding the definition of drug resistance, we focused on MDRP and CRPA, which were targeted by NAPs published in several countries.
Surveillance for drug-resistant P. aeruginosa to eliminate MDRP
MDRP is defined as P. aeruginosa non-susceptible to at least one agent in three and over antimicrobial categories such as carbapenems [26]. MDRP has quite limited treatment options and often leads to high inpatient mortality and prolonged hospital stays [32]. Thus, early detection is essential for an effective response to MDRP outbreaks. However, not all MDRP outbreaks can be identified through the sentinel surveillance unless the outbreak is epidemiologically or genetically unusual [33]. To eliminate MDRP outbreaks promptly, notifiable disease surveillance is preferable to sentinel surveillance because notifiable disease surveillance can detect MDRP infections from all hospitals. One challenge of the notifiable disease surveillance of MDRP is the difficulty in confirming drug resistance. Our results revealed an increase in the number of cases reported as MDRP infections despite not meeting the reporting criteria from 2018 to 2022. This issue would persist even if the surveillance system were changed to notifiable disease surveillance. Thus, quality assurance of antimicrobial susceptibility testing is the key to the successful implementation of the notification system. Another challenge is that we need to note whether the definition of MDRP is consistent when we compare the proportions of MDRP with other countries and regions.
Surveillance for drug-resistant P. aeruginosa to control carbapenem resistance
Carbapenems are essential broad-spectrum antibiotics that lead many developed countries to monitor CRPA infections. Such monitoring enables these countries to compare the carbapenem resistance rates of P. aeruginosa with each other [34]. For instance, EARS-Net determined the proportions of CRPA in each country within the WHO European Region and showed large differences [12]. A previous article indicated that appropriate use of carbapenems was inversely correlated with the prevalence of CRPA [35]; thus, the proportions of CRPA could serve as an indicator of the rational use of carbapenems. In addition, as carbapenemase-producing isolates of P. aeruginosa can cause severe outbreaks [3, 36], laboratory testing for carbapenemase is crucial. In the U. S., the AR Lab Network reported detecting carbapenemase genes in 2.2% of 68,712 CRPA isolates analyzed between 2017 and 2022, with 58.3% of them identified as VIM [14]. In contrast, only 4.2% of 382 meropenem-resistant CRPA isolates collected from 78 hospitals in Japan possessed acquired carbapenemases, with the IMP type being the most frequent, and no VIM-producing CRPA isolates were detected. Multilocus sequence typing of the 382 CRPA isolates revealed that ST274 was predominant. Notably, ST274 CRPA isolates rarely carried carbapenemase [37]. We consider it important to integrate the reporting system with laboratory-based surveillance to comprehensively evaluate the trend of CRPA infections, including the results of genomic sequencing. However, if CRPA infections are classified as notifiable diseases, excessive numbers of cases would be reported, potentially leading to economic costs and burdens for medical and laboratory institutions because P. aeruginosa has resistance to carbapenems through various mechanisms other than carbapenemase production [4]. Thus, sentinel surveillance integrated with laboratory-based surveillance would be indispensable to controlling carbapenem resistance by assessment of the trend of the number of CRPA infections and evaluation of whether CRPA isolates produce any of the carbapenemases.
Our study has some limitations. First, as we interviewed staff members working at hospitals and one local infectious disease surveillance center located in a specific prefecture, it is possible that the situation might differ somewhat from those in other regions. However, we endeavored to generalize the findings by conducting interviews with national academic authorities from various prefectures. Second, the interview results could be influenced both by social desirability bias that interviewees’ facilities have fewer underreported or misreported cases and by recall bias. Thus, we performed not only qualitative but also quantitative analysis for a comprehensive evaluation. Third, the reports of MDRP infections in NESID and our questionnaire survey were not confirmed by laboratory testing.