Two secondary and 5 tertiary care hospitals provided 16,548 microbiology results on
bacteria. Together, they represented a total of 246,266 admissions accumulating 2,218,861 patient-days in 2017.
Among gram-negative bacteria, non-susceptibility to 3rd generation cephalosporins (3GC) in E. coli and K. pneumoniae were 43.9% and 30.2%, respectively. Non-susceptibility to carbapenems in E. coli and K. pneumoniae were 0.9% and 4.2%, respectively. On average, 29.5% and 19.7% of P. aeruginosa were non-susceptible to carbapenems and ceftazidime, respectively; 50.9% and 53.6% of A. baumannii were non-susceptible to carbapenems and cefepime, respectively.
Among gram-positive bacteria, 26.3% of Staphylococcus aureus were non-susceptible to oxacillin; 0.7% of Enterococcus faecalis were non-non-susceptible to vancomycin; but no E. faecium was found to be vancomycin-resistant.
Figure 6 compares AMR data of Dongguan city to CHINET. E. coli, K.
pneumoniae, and A. baumannii were consistently and significantly more susceptible in Dongguan city compared to the national data. However, P. aeruginosa were significantly less susceptible to piperacillin-tazobactam (29.9% vs. 13.4%), imipenem (29.5% vs. 23.6%), gentamicin (17.5% vs. 10.7%), and ciprofloxacin (22.2% vs.
14.8%) in Dongguan city compared to CHINET data. With the exception of S. aureus, non-susceptibility was consistently higher in tertiary care hospitals compared to secondary care hospitals.
Figure 6 Comparison of non-susceptibility of indicator pathogens to antimicrobial resistance markers between Dongguan city and the China Antimicrobial Surveillance Network
A. B.
C. D.
E. F.
G.
A total of 1508 and 4090 pathogens were allocated to HAI and CAI, respectively.
Non-susceptibility was consistently higher in HAI compared to CAI, although statistically significant differences were identified only for P. aeruginosa, and for E.
coli and K. pneumoniae non-susceptible to both ceftriaxone and ciprofloxacin (Fig 7).
Figure 7 Comparison of non-susceptibility of indicator pathogens to antimicrobial resistance markers between healthcare-associated and community-acquired infections
A. B.
C. D.
E.
Table 3 Incidence density of healthcare-associated and community-acquired infections due to indicator pathogens, non-susceptible to one or more antimicrobial resistance markers
Note: AMG: aminoglycoside; CAI: community-associated infection; CAR: carbapenem; FQ:
fluoroquinolone; HAI: healthcare-associated infection; OXA: oxacillin; RMP: rifampicin; 3GCR: 3rd generation cephalosporin
Table 3 summarises incidence densities of HAI and CAI due to indicator
pathogens non-susceptible to one or more antimicrobial combinations. No HAI due to carbapenem non-susceptible E. coli was reported. The incidence density of HAI due to E. coli non-susceptible to 3GC and fluoroquinolones combined was 0.09 (95%CI:
0.07-0.11) per 1000 patient-days. The incidence proportion of CAI due to E. coli non-susceptible to 3GC and fluoroquinolones combined was 0.24 (95%CI: 0.22-0.27) per
100 admissions. The incidence density of HAI due to K. pneumoniae non-susceptible to 3GC, fluoroquinolones and carbapenem combined was 0.008 (0.004-0.014) per 1000 patient-days.
Incidence densities of HAI due to E. coli and K. pneumoniae non-susceptible to 3GC and fluoroquinolones combined were significantly lower compared to the 2014/2015 multi-centre surveillance study in Germany (Figure 8). No significant differences in incidence densities of HAI due to MDR K. pneumoniae were identified between Dongguan city and Germany.
Figure 8 Incidence densities of healthcare-associated infections due to resistant Escherichia coli and Klebsiella pneumoniae – Differences between Dongguan City, 2017 and Germany, 2014 and 2015
CAR: carbapenem; ESCCOL: Escherichia coli; FQ: fluoroquinolone; KLEPNE: Klebsiella pneumoniae;
IRR: Incidence rate ratio; 3GC: 3rd generation cephalosporin
Note: No healthcare-associated infection due to E. coli non-susceptible to 3rd generation cephalosporin, fluoroquinolones and carbapenem combined were reported.
Ten percent (562/5598) of the pathogens in the combined dataset were isolated from blood. Among these, 41.4% (65/157) of E. coli were non-susceptible to 3GC, but all E. coli were susceptible to carbapenems; 19.6% (10/51) and 2.0% (1/51) of K.
pneumoniae were non-susceptible to 3GC and carbapenems, respectively; all of 15 isolated P. aeruginosa were susceptible to carbapenems; 28.6% (4/14) of A.
baumannii susceptible to carbapenems; 21.3% (10/47) of S. aureus were non-susceptible to oxacillin; but all 14 isolated E. faecalis and 4 isolated E. faecium were susceptible to vancomycin.
Incidence proportions of BSI due to MDR E. coli were significantly higher
compared to the 2017 EARS-Net report (14.9% vs. 4.6%, P<0.001) (Figure 9). There were no significant differences for MDR K. pneumoniae, or MDR A. baumannii, or MDR S. aureus, respectively.
Figure 9 Incidence proportions of bloodstream infections due to multidrug resistant indicator pathogens between Dongguan city, 2017 and the 2017 EARS-Net report
ACIBAU: Acinetobacter baumannii; AMG: aminoglycoside; BSI: bloodstream infections; CAR:
carbapenem; EARS-Net: European Antimicrobial Resistance Surveillance Network; ESCCOL:
Escherichia coli; FQ: fluoroquinolone; KLEPNE: Klebsiella pneumoniae; MDR: multidrug resistant; OXA:
oxacillin; PEN: penicillin; PSEAER: Pseudomonas aeruginosa; RMP: rifampicin; STAAUR:
Staphylococcus aureus; 3GCR: 3rd generation cephalosporin
3.4 Article 4 – Implementation of IPC in China: a systematic review
In total, 56 articles were eligible for data extraction and analysis: 27 survey reports on structure, organisation and management of IPC; 17 observational studies
measuring outcome and process indicators; 5 interventional studies applying education and training; and 7 interventional studies testing the effectiveness of IPC strategies (Figure 10). They embraced three broad areas in IPC, issued by the
“National Health Commission of the People’s Republic of China (NHCPRC)”: 1) structure, organization and management of IPC; 2) education and training of IPC;
and 3) surveillance of process and outcome indicators relevant to IPC).
Figure 10 Systematic review profile – Implementation of IPC in China: a systematic review
NHCPRC area “structure, organisation and management of IPC”
The search terms addressing the NHCPRC area on “structure, organisation and management of IPC” identified 27 survey reports (Table 4):
- Element of “structure, organisation and management, guideline provision”
o Most primary care hospitals had an IPC committee (71%), a formal IPC programme (61%), and provided IPC guidelines (57%). Most
secondary/tertiary care hospitals had an IPC committee (98%), performed feedback on IPC indicators (93%), and provided IPC guidelines (85%).
o No information on feedback, allocated IPC funding/budget, and IPC research was identified for primary care hospitals.
o The frequencies of the elements were significantly different between hospital types, in favour for secondary/tertiary care hospitals.
- Element of “education and training”
o Significantly more secondary/tertiary care hospitals offered regular, postgraduate IPC training compared to primary care hospitals (75% vs.
53%, P<0.001).
o The survey reports did not describe details on target population, training content, or frequency of training activities.
- Element of “indicator and outcome surveillance, auditing”
o Surveillance of antimicrobial consumption (55%) was the most reported surveillance element in primary care hospitals, followed by HAI point prevalence surveys (39%), and incidence surveillance of surgical site infection (38%).
o Waste management (62%) was the most frequently audited element in primary care hospitals, followed by sterilization and medical device decontamination (58%), and environmental culturing (57%).
o Incidence surveillance of SSI (71%) was the most reported surveillance element in secondary/tertiary care hospitals, followed by HAI point prevalence surveys (67%), and surveillance of AMR (64%).
o Environmental culturing (92%) was the most frequently audited element in secondary/tertiary care hospitals, followed by waste management (57%), and sterilization and medical device decontamination (55%).
Table 4 NHCPRC areas and elements of infection prevention and control identified by 27 survey reports
NHCPRC area “education and training in IPC”
The search terms addressing the NHCPRC area on “education and training in IPC”
identified 5 single centre interventional studies.
- Five interventional studies on education and training in IPC:
o Education and training in IPC was delivered via lectures, problem-based learning, (focus) group discussion, video scenarios, and simulation training.
o The targets were new staff and nursing students.
o Training activities were associated with improvement of IPC knowledge, increase in hand hygiene compliance, and reduction of HAIs.
NHCPRC area “outcome and process indicator surveillance”
The search terms addressing the NHCPRC area on “outcome and process indicator surveillance” identified 17 observational studies and 7 interventional studies:
- Seventeen observational studies on outcome- and process indicator surveillance:
o Five studies measured device-associated HAIs, three all cause HAIs, three Clostridium difficile infections (CDI), two ventilator-associated pneumonias (VAPs), two central line-associated bloodstream infections (CLABSIs), one surgical site infection (SSI), and one hospital-acquired pneumonia.
o Twelve studies applied the standard Chinese surveillance protocol, four applied a network protocol other than the official document, and one applied a research protocol.
- Seven interventional studies on outcome- and process indicator surveillance:
o Five studies used a multimodal strategy addressing hand hygiene improvement, CLABSI prevention, and VAP prevention.
o One study successfully tested pre-operative chlorhexidine mouthwash on VAP reduction.
o One study reported MRSA reduction in the environment by improved cleaning practices.
Mapping to ECDC key/WHO core components in IPC
- Identification of 7 of the 10 ECDC key components and 7 of the 8 WHO core components (Table 5):
o All three NHCPRC areas, which are directly linked to the three ECDC key components on “structure and organisation of IPC programmes”, “education and training in IPC”, and “performing surveillance (in a network) with timely feedback”;
o Provision and appropriate promotion of (locally adapted) guidelines, and performing audits;
o Materials and ergonomics, which were elements linked with “improving the provision of alcohol-based hand rub at the point of care”, or “using new catheter insertion kits and trolleys”;
o Multimodal strategy
Table 5 Comparison with ECDC key components and WHO core components
GENERAL DISCUSSION
To our best knowledge, this is the first research project comprehensively assessing both outcome indicator surveillance (i.e. HAI and AMR) and implementation of infection control in acute care hospitals in Mainland China. The HAI prevalence in Dongguan city was lower than anticipated. Published HAI prevalence data from China, as identified by the systematic review, were higher compared to the data reported by the 2014 Chinese National PPS report. The incidence proportions of non-susceptible indicator pathogens in Dongguan city was lower compared to China as a whole. Most of ECDC key/WHO core components were implemented in Chinese acute care hospitals, but gaps in effective IPC activities/implementation were identified.
4.1 Article 1 – Dongguan city HAI PPS study
The HAI prevalence of 2.9% in Dongguan is slightly higher compared to the findings of a recent large PPS of the Beijing region (2.1%) (55). However, it still ranges in the lower field compared to other Chinese surveys (2.7-3.9%) (17, 35, 46, 47, 56). On the other hand, compared to other HAI PPS reports from low-and middle-income and high-income countries, the prevalence of this survey appears quite low (7, 57).
Reasons for the relatively low HAI prevalence are not clear. Definition issues of the Chinese HAI diagnosis criteria (47, 55) and the absence of external data
validation, may partially be responsible. Finally, no information is available about case-mix or severity-of-illness of the patients in the survey, neither in this PPS nor in reports from countries other than China.
LRTI, UTI, SSI, and BSI accounted for three-quarters of all HAIs. This finding is consistent with previous published studies with combined proportions of 66.4% to 87.5% (9, 14, 46, 47). LRTI (35.5%) was the most frequent type of HAI. This finding is similar to previous Chinese reports, the ECDC PPS, and the national US PPS (7, 9, 17, 46, 47, 56). The US PPS found unexpectedly high numbers of GI (19.0%), mainly due to C. difficile infections (9). In this survey, GIs contributed only 3.8% to the
overall number of HAIs. However, C. difficile testing is not routinely available in Chinese hospitals, and this may have underestimated true C. difficile infection ratios (55).
The distributions of pathogens between this survey, the ECDC PPS, and the US PPS are very different. The most common pathogens in this survey and the ECDC PPS are E. coli (7). The most common pathogen in the US PPS was C. difficile (9).
The proportion of Klebsiella spp. of this survey, the ECDC PPS, and the US PPS were quite similar. However, A. baumannii was very frequent in this survey compared to Europe and the US (7, 9). This is a finding similar to other reports from low- and middle-income countries (58, 59).
A total of 34.8% of the patients had one or more antimicrobials on the day of survey, which was similar with the findings reported in ECDC PPS (35%) (7). The proportion of patients receiving antimicrobials for therapeutic use (23.6%) in this survey is lower than in previous reports from China (27-54%), which were performed before the 2011 new legislation to implement campaigns for rational use of
antimicrobials in China (17, 47, 60).
4.2 Article 2 – HAI prevalence in China: a systematic review
The weighted HAI prevalence of 3.12% in this review was higher than the frequency reported by the 2014 Chinese National PPS report (2.7%), but lower than the
2011/2012 ECDC PPS (5.7%), and the recent multistate US PPS (4.0%) (7, 9, 35).
The data were almost exclusively from tertiary-care or specialty hospitals. This low prevalence may be explained in several ways. First, the Chinese HAI definitions differ from CDC and ECDC definitions (14), and they have not been updated since their publication by the Ministry of Health in 2001 (14, 17, 35, 46, 47, 55). Furthermore, a recent multicentre PPS from China suggested that microbiological confirmation may be overvalued (i.e. >80% of LRTI were microbiologically confirmed) (14). Second, there is no validation of surveillance. In most hospitals, data collection is done by physicians, who have an interest in reporting as few HAIs as possible (14). Third, antibiotic prescription in Chinese hospitals is disturbingly high (61, 62). High
antimicrobial use, together with overvaluing microbiology for case definition, can result in underreporting. Fourth, hospital managers do not invest sufficient resources into infection prevention and control (63). Knowledge about infection prevention and control outside hand hygiene is low among both doctors and nurses (64). Together, this results in a lack of expertise when performing HAI surveillance.
LRTI was the most frequent HAI, which is consistent with the 2014 Chinese National PPS report, and the 2011/2012 ECDC PPS and the 2011 US PPS (7, 9, 35).
Interestingly, BSIs in children’s hospitals was unexpectedly low (5.6%) compared to the past ECDC PPSs, where the proportion of BSI among HAI in children was 45%
(65). SSIs were equally distributed among hospital settings but were low in children’s hospitals, which may be due to low surgical activities, and short length of stays.
The most frequent microorganism in general hospitals was P. aeruginosa (14.9%), which is similar to findings of the national PPS report in 2014 (15.5%) (35).
Furthermore, K. pneumonia (19.1%) was the most common microorganism isolated in the children hospitals, whereas, in Europe and the USA, the most common pathogens in children were coagulase-negative staphylococci (CoNS) (21% and 31%, respectively) (65, 66). The CHINET surveillance programme reported CoNS as the most common microorganisms isolated from blood cultures (67). Given that BSI accounted for a lower proportion of HAI in the population of children in China, it is not surprising that CoNS did not emerge as the most common pathogen.
4.3 Article 3 – AMR prospective surveillance in Dongguan city 2017
The incidence proportions of non-susceptible indicator pathogens in Dongguan city was lower than the average incidence proportion in Mainland China, and lower than anticipated. Third-generation cephalosporin (3GC)-resistant E. coli and K.
pneumoniae remain a major challenge in Dongguan city. It is lower compared to the 2017 CHINET report and other Chinese study (30, 68), but similar to Thailand (25).
The proportion of resistance to carbapenem in K. pneumoniae has emerged from 3.0% in 2005 to 20.9% in 2017 in China (69). With 4.2%, this proportion was much lower in Dongguan city, but similar to the average of invasive pathogens in Europe
(7.2%) (21). Also similar to the average of invasive MDR pathogens in Europe (3.9%), 5% of P. aeruginosa were multidrug-resistant in Dongguan city. The proportions of vancomycin-resistant E. faecalis (0.7%) in our study were lower compared to the US (8.5%), but similar to Canada (0.1%) and other Asia-Pacific regions (0.01%) (70, 71).
Compared to Germany (23), the incidence densities of HAI due to E. coli and K.
pneumoniae resistant to 3GC and fluoroquinolones combined were significantly lower. On the other hand, the incidence densities of CAI due to E. coli and K.
pneumoniae resistant to 3GC, or 3GC and fluoroquinolones combined, were
significantly higher. One reason for this may be that 50% outpatients in China receive antibiotics (72); particularly cephalosporins and fluoroquinolones (73). CAIs are treated without performing microbiological testing and far too many broad-spectrum antimicrobials are prescribed (74). The incidence density of infections due to
carbapenem-resistant E. coli and K. pneumoniae together in our study was 0.017 per 1,000 patient-days. This finding was lower compared to the 2015 Chinese National Carbapenem Resistant Enterobacteriaceae (CRE) surveillance (0.041 per 1,000 patient-days) (24), but was four-fold higher compared to the 2011/2012 French national CRE surveillance (0.0041 per 1,000 patient-days) (75).
Both E. coli and K. pneumoniae were the predominant pathogens isolated from blood. The incidence proportion of BSI due to MDR E. coli was significantly higher compared to the 2017 EARS-Net report (21); whereas the incidence proportion of BSI due to MDR K. pneumoniae was similar. The incidence proportion of BSIs due to MDR S. aureus in our study was higher compared to the 2017 EARS-Net report (21), although the difference did not reach statistical significance. Interestingly, the
incidence density of hospital-onset BSI due to MRSA in our study was very low (0.004 per 1,000 patient-days), which is even lower compared to Europe (0.026 per 1,000 patient-days) (22). Two major factors may explain finding: 1) the average LOS in our study hospitals was longer (6.2-10.6 days) compared to the US (4.5 days) or Europe (5.1 days), which inflates the denominator (1, 7, 39); and 2) blood culture sampling is low, with a rate of 10.4 performed blood culture sets per 1,000
patient-days. This is significantly lower than e.g. in Denmark, Finland, Sweden, France, or UK (>50 blood culture sets per 1,000 patient-days) (21).
4.4 Article 4 – Implementation of IPC in China: a systematic review
Structure, organisation and management of infection prevention and control Effective IPC in an acute care hospital needs an IPC programme with sufficient staffing and an allocated budget, support from the hospital management, and well-defined duties and targets. Only two-thirds of primary care hospitals have an IPC programme and an IPC committee. No information was found for any of the other elements of this NHCPRC area. The majority of secondary/tertiary care hospitals have an IPC programme, but only a third have an allocated budget. It is difficult to estimate the challenges concerning the proper functioning of IPC programmes, but it has been shown that competing resources may have a negative impact on their effectiveness (76). Staffing, as identified in several reports, is at minimal level (31, 32, 77). However, high staff turnover, particularly among IPC doctors (78-80), is of even more concern than understaffing. This is partially explained by low salaries and limited career options (79). Most IPC departments are managed by junior IPC doctors and IPC nurses. Due to the hierarchical gaps, IPC professionals face
structural/hierarchical challenges, and struggle in influencing behaviour change (81).
Almost all secondary/tertiary care hospitals have an IPC committee in place.
However, the importance of IPC is not always recognized by hospital management, and IPC committee members do not always participate actively in the committees (79, 80). Most secondary/tertiary care hospitals indicated to have a feedback mechanism in place. However, too often, information is conveyed to the hospital management only, and data reaches healthcare workers too late, if at all, to be meaningful and actionable (82, 83).
Education and training in infection prevention and control
Half of primary-care and three-quarters of secondary/tertiary care hospitals indicated to have IPC education and training in place. Education and training should be team-
and task-oriented, the content should follow local guidelines, and implementation should be multimodal (31). Unfortunately, survey reports lack details, and thus, it is difficult to assess the available resources for education and training and whether they are adequate. Professionals working in IPC need knowledge and skills on
management and implementation research (84). The European Society for Clinical Microbiology and Infectious Diseases has created a platform to offer comprehensive IPC training to doctors (84, 85). Implementation research and management are not part of IPC training in Mainland China. Basic and intermediate skill levels of
education and training focus on legal aspects, mandatory surveillance, definitions, diagnosis, HAI classification, HAI prevention, and hand hygiene (86). The contents of the advanced level were not sufficiently specified in the documents to allow
conclusions on delivering skills regarding on the topic of project management and implementation research.
Surveillance of outcome and process indicators
A range of surveillance activities were identified in the survey reports and in the observational studies, with significant differences between hospitals. Many hospitals perform regular prevalence surveys, but given that yearly prevalence surveys on HAI are mandatory in Mainland China, the proportion of primary- (40%) and
secondary/tertiary care hospitals (60%) is surprisingly low. Most prospective surveillance measures SSI, which is similar to a recent European survey (82).
secondary/tertiary care hospitals (60%) is surprisingly low. Most prospective surveillance measures SSI, which is similar to a recent European survey (82).