Proportions of Pseudomonas aeruginosa and Antimicrobial-Resistant P aeruginosa Among Patients With Surgical Site Infections in China: A Systematic Review and Meta-analysis

Abstract Background Pseudomonas aeruginosa is one of the most common pathogens in surgical site infections (SSIs). However, comprehensive epidemiological and antibiotic resistance details for P aeruginosa in Chinese SSIs are lacking. We evaluated the proportions and antimicrobial resistance of P aeruginosa among patients with SSIs in China. Methods Relevant papers from January 2010 to August 2022 were searched in databases including PubMed, Embase, Web of Science, China Biomedical Literature Database, China National Knowledge Infrastructure, Wanfang, and Weipu. A meta-analysis was performed to analyze the proportions and 95% confidence interval (CIs) of P aeruginosa among patients with SSIs. Meta-regression analysis was used to investigate the proportion difference among different subgroups and antimicrobial resistance. Results A total of 72 studies met inclusion criteria, involving 33 050 isolated strains. The overall proportion of P aeruginosa among patients with SSIs was 16.0% (95% CI, 13.9%–18.2%). Subgroup analysis showed higher proportions in orthopedic (18.3% [95% CI, 15.6%–21.0%]) and abdominal surgery (17.3% [95% CI, 9.9%–26.2%]). The proportion in the central region (18.6% [95% CI, 15.3%–22.1%]) was slightly higher than that in other regions. Antibiotic resistance rates significantly increased after 2015: cefoperazone (36.2%), ceftriaxone (38.9%), levofloxacin (20.5%), and aztreonam (24.0%). Notably, P aeruginosa resistance to ampicillin and cefazolin exceeded 90.0%. Conclusions The proportion of P aeruginosa infection among patients with SSIs was higher than the data reported by the Chinese Antimicrobial Resistance Surveillance System, indicating rising antimicrobial resistance. The existing antimicrobial drug management plan should be strengthened to prevent a hospital epidemic of drug-resistant P aeruginosa strains.

Surgical site infections (SSIs) are the most common and challenging nosocomial infections in surgical patients worldwide.Approximately 230 million surgeries are performed each year worldwide, of which approximately 31% result in patients with varying degrees of SSIs; importantly, one-third of postoperative deaths are related to SSIs [1][2][3][4].The incidence of SSIs varies among countries with different economic levels.In Europe, the SSI rate is 2%-5%; in Africa, it is as high as 51.1% [5].It is estimated that in the United States, 2.5% of surgical patients will have varying degrees of SSIs each year.Compared with that for patients without infections, the average length of hospital stay for patients with SSIs is 7-11 days longer and the risk of death is 2-11 times higher than those for patients without infection.The annual total cost associated with SSIs exceeds $10 billion [6][7][8].In short, SSIs not only bring a heavy financial burden to patients but are also life-threatening.
Although the pathogenic microorganisms that cause infection are diverse and the epidemiology of SSIs is also different among different environments and different countries, it has been reported that the proportions of gram-negative bacilli increases over time and that Pseudomonas aeruginosa is one of the most common pathogens [9].In a hospital in Egypt, 70 strains of P aeruginosa were detected in 200 SSI samples (the detection rate was 35.0%) [10].In a tertiary referral hospital in Amman, Jordan, a large number of gram-negative bacteria were detected in SSI samples, of which P aeruginosa was one of the main pathogenic bacteria; the detection rate was 14.8% [11].
The increase in antibiotic resistance of pathogens has become a major public health problem.Among them, multidrug-resistant Pseudomonas aeruginosa in Chinese SSIs • OFID • 1 Open Forum Infectious Diseases M A J O R A R T I C L E (MDR) P aeruginosa has become increasingly serious in nosocomial infections and is closely related to the increase in mortality and length of hospital stay [12].In 2017, the World Health Organization (WHO) published the first list of antibiotic-resistant "key pathogens."There are 3 extremely important pathogens, including carbapenem-resistant P aeruginosa [13].The report of the National Notifiable Disease Surveillance System showed that the antimicrobial resistance rates for P aeruginosa to imipenem, quinolones, and third-generation cephalosporins were 15%, 9%, and 20%, respectively [8].According to WHO data, the antimicrobial resistance rates of P aeruginosa to fluoroquinolones, ceftazidime, and aminoglycosides in 2020 were 46.4%, 41.0%, and 37.1%, respectively [14,15].The antimicrobial resistance mechanisms of P aeruginosa are complex and diverse, and some strains can develop resistance to multiple antibiotics at the same time, posing a great challenge to the selection of appropriate antibiotics.Previous studies have found that the most suitable antibiotics for MDR P aeruginosa are carbapenems.However, P aeruginosa produces large amounts of metallo-β-lactamases, leading to resistance to β-lactams, including carbapenems, greatly limiting the choice of the correct treatment plan [16].
In China, data from the National Nosocomial Infection Surveillance System showed that SSIs accounted for 10.4% of all nosocomial infections and that the rate of detection of P aeruginosa in SSIs was 8.9% [17].In 2022, data from the Chinese Antimicrobial Resistance Surveillance System indicated that the antimicrobial resistance rates of P aeruginosa to imipenem and meropenem steadily declined in the past 5 years, fluctuating between 18.9% and 30.7%; the antimicrobial resistance rate to polymyxin B was low, fluctuating between 0.5% and 1.2%, consistent with the data reported in 2020; and the antimicrobial resistance rates to gentamycin, ciprofloxacin, ceftazidime, cefepime, and piperacillin were below 20.0%.Chinese clinicians have gradually realized that the antimicrobial resistance of P aeruginosa is becoming increasingly serious, especially the increasing number of P aeruginosa strains capable of producing metallo-β-lactamases, leading to the ineffectiveness of multiple antibiotics in the treatment of SSIs and increasing the physical and economic burden on patients [18].
Understanding the epidemiological characteristics and antibiotic resistance of P aeruginosa that cause SSIs is critical for clinicians to develop prevention and treatment measures for SSIs.Although previous studies have described the epidemiological characteristics and antimicrobial resistance of P aeruginosa in nosocomial infections, substantial uncertainty remains in the proportions of P aeruginosa among SSIs.Therefore, the purpose of this systematic review was to summarize and assess the proportions and antimicrobial resistance rate of P aeruginosa among SSIs to provide further guidance for the prevention of SSIs and to promote the best empirical antimicrobial treatment.

METHODS
The study was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [19] and registered on the PROSPERO database (CRD42022363001).

Search Strategy
The study used free words combined with subject terms to comprehensively search the PubMed, Embase, Web of Science, China Biomedical Literature Database, China National Knowledge Infrastructure, Wanfang, and Weipu databases (Supplementary Table 1).The focus of this study was to analyze the epidemiological and antimicrobial resistance characteristics of P aeruginosa in the past decade.Therefore, the search time was limited to January 2010 to August 2022.The search terms included surgery, postoperative, surgical wound infection, site infection, P aeruginosa, and China.The retrieved papers were managed using EndNote (version 20), and duplicates were eliminated.Relevant conference papers were manually searched in the journal database of the Army Medical University Library, and all references included in the study were reviewed.

Inclusion and Exclusion Criteria
The literature included in the meta-analysis met the following criteria: (1) patients were clinically diagnosed with SSIs; (2) the sample collection began in 2010; (3) the study subjects were Chinese; and (4) there were sufficient data to calculate the proportions or antimicrobial resistance of P aeruginosa.The following exclusion criteria were applied: (1) abstracts, reviews, or communication papers; (2) studies with a small number of detected bacterial strains (requiring at least 197 isolated bacterial strains [20] to estimate the expected proportion of P aeruginosa at 8.9% [17] and an error rate of <4%); (3) a lack of sufficient information, including incomplete or unavailable research data; and (4) research based on data from the National Nosocomial Infection Surveillance System.

Data Extraction and Bias Risk Assessment
Data were extracted independently by 2 researchers (Y.Y. and G. L.) using a unified data table.The main data extracted included the first author, publication year, study area, study design, hospital level, and data that were used to calculate the proportions of P aeruginosa and drug-resistant P aeruginosa.For conflicts regarding data extraction, a consensus was reached through consultation with a third party.For duplicated publications, only data with the highest quality and highest number of detected bacterial strains or the most complete information were extracted.
The quality of the included literature was evaluated using predetermined criteria extracted and modified from a previous case series scale consisting of 8 items [21].By answering low, high, or unclear to questions, bias can be identified in selected literature.The total score ranges from 0 to 8 points, and the higher the score is, the higher the quality.In this study, R version 4.1.3software was used to summarize the risk of bias.A score <4 points was defined as high risk, and a score ≥4 points was defined as low risk.

Statistical Analysis
All statistical analyses were performed using R version 4.1.3software.Statistical tests were all 2-tailed.Unless stated, P < .05 was considered statistically significant.The proportions of P aeruginosa and antimicrobial-resistant P aeruginosa isolates among SSI patients in each study were calculated using the following formula: The proportion in each study and its 95% confidence interval (CI) were calculated using Freeman-Tukey double arcsine transformation [22].A DerSimonian-Laird random effects model was used to estimate the combined proportions of the included literature and its corresponding 95% CI [23].The Cochrane Q test was used to analyze the heterogeneity among the studies.The Q statistic approximately follows the χ 2 distribution with k -1 (k is the number of studies).When P is <.10, it can be considered that there is heterogeneity between studies.The magnitude of heterogeneity was quantitatively evaluated based on the Higgins I 2 value, which ranges from 0% to 100%, and significant heterogeneity is generally considered to exist for I 2 values >50% [24].A funnel plot was used to analyze whether there was publication bias in the included literature, and Egger linear regression was used to test the asymmetry of the funnel plot [25].
The differences in the proportions of P aeruginosa and the antimicrobial resistance rate for P aeruginosa were explored through subgroup analysis and univariate meta-regression.In the meta-regression analysis, the dependent variable was the proportion of P aeruginosa or the antibiotic resistance data for P aeruginosa isolates.The independent variables were surgical type (dummy variable: Orthopedic), region (dummy variable: Eastern region), hospital level (dummy variable: Tertiary), risk of bias (dummy variable: High), study design (dummy variable: Retrospective), sample size (dummy variable: <500 isolates), and study time (dummy variable: 2015 or after).In the meta-regression analysis, the restricted maximum likelihood method was used to estimate the variance between studies, and the proportion of variance explained by any meta-regression model was estimated using the R 2 statistic [26,27].

Literature Search
The initial search obtained 3758 relevant studies.After excluding duplicates, we reviewed the titles and abstracts of 1916 papers (preliminary screening) and excluded 1323 studies.After further reading the abstracts and full texts, we excluded 521 papers.Among them, there were 283 studies with a total number of isolated bacterial strains <197, 172 studies with nonsurgical wound infections, 33 studies without valid data, 12 studies with nonhuman subjects, 12 duplicated publications, 7 case reports, and 2 studies based on the National Nosocomial Infection Surveillance System.Ultimately, 72 studies that met the inclusion criteria were included in the analysis (Figure 1).

Study Characteristics and Bias Risk Assessment
The characteristics of the 72 included studies are shown in Table 1.Among them, 69 studies reported the proportion of P aeruginosa among SSIs.The number of bacterial strains isolated per study ranged from 201 to 2162, with a total of 33 050 strains isolated.Another 3 studies reported antimicrobial resistance data for P aeruginosa.The results of the risk of bias assessment are shown in Figure 2, and the details are shown in Supplementary Table 2.The highest score for a single study was 7, and the lowest score was 1.There were 48 high-quality studies (≥4).All studies were conducted between 2010 and 2023, and 24 of the 34 provinces in China were represented (Figure 3).Among them, 28 studies were from the eastern coastal region, 33 studies were from the central region, and 11 studies were from the western region.Most studies were retrospective (62 of 72).

Proportions of Antimicrobial Resistance
We further analyzed the resistance of P aeruginosa to different antibiotics, and 54 of the 72 papers reported P aeruginosa resistance to 69 antibiotics.Among them, 18 antibiotics were reported in 10 or more studies (Table 3, Figure 5).We performed a meta-analysis of these antibiotics and compared the antimicrobial resistance rates for P aeruginosa before and after 2015.The results indicated that compared with that reported in the studies conducted before 2015 (31.5% [95% CI, 20.2%-43.8%]), the rate of antimicrobial resistance of P aeruginosa to cefoperazone (67.7% [95% CI, 54.6%-79.6%])was significantly higher in the studies conducted after 2015 (R 2 = 42.75).Similarly, the rate of resistance to ceftriaxone was significantly higher after 2015 (R 2 = 42.27)-thatis, 39.7% (95% CI, 27.6%-52.5%)before 2015 and 78.6% (95% CI, 61.5%-91.9%) in 2015 and after.After 2015, there was also a notable increase in the resistance rate to levofloxacin (R 2 = 12.49) and aztreonam (R 2 = 10.88).In addition, the resistance of P aeruginosa to cefazolin and ceftazidime showed increasing trends, but there was no significant difference between the subgroups.Notably, the rate of resistance of P aeruginosa to ampicillin and cefazolin exceeded 90.0%.The resistance rates of P aeruginosa in different regions and types of hospitals are detailed in Supplementary Tables 3 and 4.

DISCUSSION
In this study, we conducted a meta-analysis of the epidemiological characteristics and antimicrobial resistance of P aeruginosa in SSIs in China.Compared with the data reported by the National Nosocomial Infection Surveillance System [17], the overall proportion of P aeruginosa among SSIs in this study  was higher.In addition, our results showed that the rate of resistance of P aeruginosa to a variety of antibiotics (cefoperazone, ceftriaxone, levofloxacin, and aztreonam) was significantly increased.The total proportion of P aeruginosa among SSIs (16.0%) was lower than the findings reported in Malaysia (26.25%) but significantly higher than the data reported by the National Nosocomial Infection Surveillance System (8.9%) [17,100].We analyzed the reasons for the discrepancy between our results and the data reported by the National Nosocomial Infection Surveillance System.We believe that this discrepancy may be caused by the following reasons.First, the data reported by the National Nosocomial Infection Surveillance System include all nosocomial infections.Among them, the number of bacteria isolated from SSIs was 1638, far less than the 33 050 strains isolated from SSIs in our study.In addition, the 69 studies included in our study involved >69 hospitals distributed in 24 different provinces, including tertiary hospitals and nontertiary hospitals.Second, the studies included in this study were all conducted after 2010, and some of the studies (28) were even completed after 2015.However, the data reported by the National Nosocomial Infection Surveillance System were collected in 2014.Over time, the proportions of P aeruginosa among SSI samples may fluctuate, leading to a difference between the results of this study and the data collected by surveillance systems.
In addition, the results of our study provide more detailed information on SSIs caused by P aeruginosa.Groupings were made by type of surgery, hospital level, and different regions and provinces, and then, the proportions of P aeruginosa among patients with SSIs in the different subgroups were analyzed.Subgroup analysis showed that the proportions of P aeruginosa varied between different types of surgery, with higher rates in orthopedic surgery (18.3%) and abdominal surgery (17.3%) and lower rates in other types of surgery (11.3%).The results indicated that compared with other types of surgery, patients undergoing orthopedic and abdominal surgery were more likely to have SSIs caused by P aeruginosa.This finding is consistent with the results of previous studies, suggesting that P aeruginosa is one of the main pathogenic causes of SSIs after orthopedic surgery and abdominal surgery and that P aeruginosa should be suspected if a SSI occurs after orthopedic surgery and abdominal surgery [101][102][103].The proportions of P aeruginosa in SSI samples in different regions were different.Among them, the proportion of P aeruginosa in SSI samples was highest for the central region (18.6%);importantly, most of the studies from the central region were orthopedic surgery-related or abdominal surgeryrelated studies.
Pseudomonas aeruginosa is resistant to a large number of antimicrobial drugs.It is particularly difficult to treat infections caused by this microorganism, and a delay in treatment can lead to increased mortality.Empirical treatment is dependent on microbiological test results.However, the high resistance of pathogenic bacteria to antibiotics may increase the possibility of inappropriate empirical treatment, resulting in poor clinical treatment outcomes and increasing the financial burden on patients.Therefore, in addition to taking appropriate antimicrobial treatment measures in a timely manner and following the guidelines for the use of antibiotics, it is also necessary to design an empirical treatment plan based on the results of antimicrobial susceptibility tests [104,105].In addition to intrinsic antimicrobial resistance, P aeruginosa can also acquire antimicrobial resistance through other known antimicrobial resistance mechanisms and then develop into MDR bacteria or pandrug-resistant bacteria, leading to life-threatening serious infections [106][107][108].In this study, variations exist in the antimicrobial resistance rates of P aeruginosa across different regions.In particular, the central region exhibits generally higher resistance rates compared to the other 2 regions, notably for antimicrobial agents such as amikacin, ciprofloxacin, and levofloxacin.These findings are consistent with the higher detection proportion of P aeruginosa in the central region.The rate of antimicrobial resistance of P aeruginosa to the carbapenem antibiotic meropenem (12.2%) is comparable to the rate of antimicrobial resistance reported by the Chinese Antimicrobial Resistance Surveillance System in 2020 (14.4%), but the overall trend of resistance is decreasing.The antimicrobial resistance rate before 2015 was 15.0%, and the antimicrobial resistance rate after 2015 was 7.7%.In this study, the resistance of P aeruginosa to meropenem was comparable to that reported in Uganda (14.0%) but significantly lower than that reported in  southwestern Iran (49.5%) and Taiwan (73.2%), a finding that may be related to whether the application of antibiotics was reasonable [109][110][111].
Pseudomonas aeruginosa produces β-lactamase, which is the main cause of its resistance to carbapenem antibiotics.Such strains of P aeruginosa are not only resistant to carbapenem antibiotics but may also develop significant resistance to other β-lactam antibiotics [112,113].In this study, the rates of resistance of P aeruginosa to cefoperazone (31.5% before 2015 and 67.7% after 2015), ceftriaxone (39.7% before 2015 and 78.6% after 2015), and aztreonam (20.4% before 2015 and 40.4% after 2015) significantly increased after 2015.The possible reason is that a large amount of these 3 antibiotics were used in clinical treatments, resulting in an increase in the resistance of P aeruginosa to these 3 antibiotics.According to data from the National Nosocomial Infection Surveillance System, the average rate of use of antimicrobial drugs in hospitalized patients in China was 50.5%, which is comparable to that of India (57%) [114].However, this rate is considerably higher than that reported in the Netherlands (30.9%), the United Kingdom (34.7%), and Canada (36.3%) [18].Clinical practitioners in China demonstrate a proclivity toward the use of broad-spectrum antimicrobial agents or combination therapy when treating infections [115].The above results suggest that clinicians need to avoid the abovementioned antibiotics, which produce high antimicrobial resistance in clinical practice.In the follow-up prevention and treatment of infections, the use of antimicrobial drugs should be more strictly monitored, and antimicrobial drug sensitivity tests should be used to guide treatment [18].Notably, when analyzing the antimicrobial resistance of P aeruginosa in this study, it was found to be resistant to cefazolin (70.1% before 2015 and 93.3% after 2015) and ampicillin (97.8% before 2015 and 94.1% after 2015), a finding that is consistent with the results of a study in Ethiopia [116].Despite not being recommended by the Clinical and Laboratory Standards Institute guidelines, research has continued to utilize them for susceptibility testing over the past decade.Our study confirms their ineffectiveness for susceptibility testing of P aeruginosa, offering valuable guidance for future clinical microbiological testing.
Aminoglycoside antibiotics are also commonly used antimicrobial drugs for the treatment of P aeruginosa infection.In this study, the resistance of P aeruginosa to aminoglycoside antibiotics, represented by amikacin, tobramycin, and gentamycin, was low.There was also a decreasing trend in the resistance to gentamycin (38.3% before 2015 and 29.7% after 2015).Quinolone antibiotics are the most recently discovered class of drugs and are active against gram-negative bacteria [117].In this study, the antimicrobial resistance of P aeruginosa to quinolone antibiotics, represented by levofloxacin and This study has the following limitations.First, the heterogeneity among the included studies was considerable, and the subgroup analysis and meta-regression analysis could not fully explain the source of the heterogeneity.Second, because smallsample studies are prone to producing accidental results, the analysis excluded studies with a small number of bacterial strains isolated from SSIs, resulting in a lack of data on the proportions of P aeruginosa in some provinces.In addition, the exclusion of this type of study meant that most of the surgical procedures included in the study were orthopedic and abdominal surgeries.The proportions of P aeruginosa infection were high in these 2 types of surgeries, potentially leading to a high estimate.Third, the majority of the regions included in the study were the central and eastern coastal areas, and the monitoring of surgical wound infections was more common in these areas.Therefore, the estimated proportions of P aeruginosa cannot represent the overall situation of SSIs in China.

CONCLUSIONS
In summary, the proportions of P aeruginosa and the susceptibility to antimicrobial drugs vary with region and time and need to be monitored at all times.Compared with the data reported by the Chinese Antimicrobial Resistance Surveillance System, the proportion of P aeruginosa among SSIs obtained in this study was higher.Therefore, it is necessary to carry out long-term monitoring to understand the actual proportion and antimicrobial resistance of P aeruginosa among SSIs and to develop appropriate healthcare mechanisms.Pseudomonas aeruginosa isolated from surgical sites has high antimicrobial resistance to ampicillin and cefazolin, and antimicrobial resistance to cefoperazone, ceftriaxone, levofloxacin, and aztreonam has significantly increased.We therefore recommend initiating appropriate infection prevention measures, strengthening existing antimicrobial stewardship programs, and conducting regular antimicrobial surveillance to prevent antimicrobialresistant P aeruginosa in hospitals.

=
Number of Pseudomonasaeruginosa isolates Number of all the detected isolates × 100% Proportions of antimicrobial-resistant Pseudomonasaeruginosa = Number of detected Pseudomonasaeruginosa isolates resistant to a given antibiotic Number of Pseudomonasaeruginosa isolates detected × 100%

Figure 1 .
Figure 1.Literature inclusion and exclusion process.

Figure 2 .
Figure 2. Results of the risk of bias assessment.Abbreviation: SSI, surgical site infection.

Figure 3 .
Figure 3. Geographic distribution of Pseudomonas aeruginosa proportions among patients with surgical site infections in China.

Figure 4 .
Figure 4.The proportions of Pseudomonas aeruginosa among patients with surgical site infections in different surgical subgroups.Abbreviation: CI, confidence interval.

Table 1 . Continued
In this column, "multiple" refers to the different types of surgeries involved in the research and cannot be classified into a specific type of surgery.
Abbreviation: SSI, surgical site infection.a

Table 2 . Proportions of Pseudomonas aeruginosa Infection Among Different Subgroups of Patients With Surgical Site Infections
Abbreviation: CI, confidence interval.aOtherrefersto multiple surgeries and a specific type of surgery, other than orthopedic or abdominal surgeries; these were reported in a small number of studies.bStudytypecomprises prospective or surveillance.cThenumber of all Pseudomonas aeruginosa isolates.

Table 3 . Results of the Combined Proportions of Drug-Resistant Pseudomonas aeruginosa at Different Times
Abbreviation: CI, confidence interval.