Sink Contamination And Carbapenem-Resistant Gram-Negative Bacteria Infection In The Intensive Care Unit:A Prospective Multicenter Study

Objective:To understand the characteristics of sinks in the Intensive Care Unit（ICU）at tertiary hospitals in Chengdu city,a field investigation about the setup,cleaning and disinfection compliance of the sink was conducted in all of the selected ICU.To detect the splashing distance during the clinical use of the sink by the fluorescent labeling method.To understand the extent to which sinks are contaminated by carbapenem-resistant Gram-negative bacteria（CR-GNB）and to what degree contaminated sinks are associated with CR-GNB hospital acquired infections.Materials and Methods:A prospective multicenter study was addressed and carried out in 16 ICUs at 11tertiary hospitals in Chengdu city from March 2019 to January 2020,including 9general ICU（GICU）and 7 neonatal ICU（NICU）.This study was approved by the Ethical Committee of the West China Hospital with informed consent being waived.1.An on-site survey form was designed and a field survey was conducted in all of the ICUs included in the study.The fluorescent powder trace method was used to track the splashed location and distance of the sink during clinical use.All the data were entered to the database and checked by another person.SPSS 22.0 software was used to for the statistic work.Chi-square test for continuity correction of related variables was conducted to analyze the risk factors of sink splashing to find out the reason of spatter from the sink.2.All the sinks in the ICUs were sampled once,including faucets,surfaces,bowls,overflows,drains and drain holes.Swabs were immediately placed into sterile tubes and the streaked onto the special selected-plate containing meropenem to screen CR-GNB quickly.Isolates from sinks were subjected to preliminary species identification based on matrix-assisted laser desorption/ionization–time of flight mass spectrometry（MALDI-TOF-MS,Bruker,Billerica,Massachusetts）.A mathematical model was established by R4.0.2.The‘oversampling’method to expand the number of positive samples was used,the sink was contaminated by CR-GNB or not was regarded as an independent variable and medical institutions was regarded as a random effect to establish a multivariate mixed effect Logistic regression model,to analyze the risk factors that cause CR-GNB contamination in the sinks at ICU.3.All CR-GNB clinical isolates which were recovered from the patients in ICUs with one or more sinks positive for CR-GNB or those form other ICUs at the same hospital,within the period between two weeks before and three months after the sink sampling data.All the isolates from sinks and clinical samples were subjected to whole-genome sequencing using the Illumina Hi Seq X10 platform.The genome sequence was analyzed by single nucleotide polymorphism（SNP）,the difference of the SNP between the CR-GNB isolated from the contaminated sink and the clinical isolates was compared,and the genetic relationship was analyzed,and the phylogenetic tree was established.4.The results of sink contamination were feedback to the infection control team of the hospital.The cleaning and disinfection policy were reviewed.After 30 days of that,re-sampling of the sink to identify the contamination of CR-GNB was conducted to verify the effect of the interventions.Results:1.In total,16 ICUs were invited to conduct this study,including 9 GICU and 7 NICU at 11 hospitals in Chengdu City,Sichuan Province,southwest China,covering 561intensive care beds.A total of 158 sinks were investigated and sampled.The bed ratio of the sink was 3.6:1,and the rate of routine cleaning and disinfection was 75.95%（95%CI[confidence interval],69.29%～82.61%）and 87.97%（95%CI,82.90%～93.04%）,respectively.Only 53.16%（95%CI,45.48%-60.94%）of the sinks were cleaned and disinfected twice a day.The splashing rate of the sink during the clinical use was 58.86%（95%CI,51.19%～66.53%）,and the most frequent splashing location was the surfaces of the sink（94.62%,95%CI[91.10%～98.14%]）.The risk factors of splashing were sink material（χ²=6.17,P=0.013）,aerators（χ²=4.56,P=0.033）and install in GICU（χ²=10.12,P=0.001）.2.Before intervention,a total of 789 sink samples were collected and 82 isolates of CR-GNB were cultured from 43 sinks.The CR-GNB contamination rate of sink was27.22%（95%CI,24.14%～30.30%）,of which carbapenem-resistant Acinetobacter baumannii（CRAB）,carbapenem-resistant Pseudomonas aeruginosa（CRPA）and carbapenem-resistant Klebsiella pneumoniae（CRKP）contamination rate was 10.76%（95%CI,8.61%～12.91%）,7.59%（95%CI,5.76%～9.42%）and 3.80%（95%CI,2.48%～5.12%）,respectively.None sink was contaminated by carbapenem-resistant Escherichia coli（CREC）,and the other carbapenem-resistant Enterobacteriaceae（CRE）contamination rate was 8.23%.The sink contaminated rate of CR-GNB in GICU was significantly higher than that in NICU（34.95%vs 12.73%,χ²=8.94,P=0.003）.Multivariate mixed effects logistic regression model shows that the main risk factors that cause sink CR-GNB contamination were the overflow（OR[odds ratio]=6.30,95%CI[1.76～22.51],P=0.004）,aerator（OR=9.21,95%CI[1.88～45.04],P=0.006）,the use of other disinfectants but not chlorine-releasing agents or disinfectant wipes（OR=19.89,95%CI[1.82～217.28],P=0.010）,cleaned by other persons but not cleaners or nurse（OR=206.44,95%CI[16.79～2537.16],P<0.001）.The disinfection work conducted by nurse was a protective factor（OR=0.01,95%CI[0.01～0.11],P<0.001）.3.The main strain of CRKP that causes sink contamination was ST11 type（83.33%,95%CI[79.43%～87.23%]）,carrying bla _KPC-2 carbapenemase gene.A total of 90.91%（95%CI,81.10%～100.00%）CRPA isolates from sink was ST253 type and harbor bla _VIM-2carbapenem enzyme genes.The CRAB isolates from the sink was mainly ST2type（95.45%,95%CI[86.74%～100.00%]）,all of them contain bla _OXA-23 carbapenem-ase genes.CRKP clinical isolates（n=28）were recovered from only 3 of the 7 ICUs with CRK-contaminated sinks and 3 other participating ICUs during the study period in the same hospital containing an ICU with CRK-contaminated sinks.The isolation of CRKP from the sink was related to 2 outbreaks.The 6 isolates（5 clinical isolates and 1 sink isolate）of ST789 type shared 0～2 SNPs,suggesting that they belonged to a common clone,which caused an infection outbreak and sink contamination.The 7 ST11 KL47CRKP isolates（4 sink isolates and 3 clinical isolates）shared 3～21 SNPs.The detailed analysis suggested that sink might be the potential source of an infection cluster and cross-transmission between sinks in the same ICU and between different hospitals.ST11 KL64 CRKP was transmitted in ICU in 2 different hospitals.There were 28 isolates of CRAB clinical isolates recovered from CRAB-contaminated sinks during the study period in the same hospital.The 28 clinical isolates 22 sink isolates can divided into 6 dominant clones according the difference of SNPs.Clone2a was transmitted between ICUs in two different hospitals and between two different sinks in the same hospital.Clone 2b and clone 2d caused an infection outbreak and sink contamination respectively,but there was no subsequent sink-to-patient transmission happened.As environmental reservoir,clone 2c might be the potential source of 4 infections.Clone 2e was isolated from two sinks,suggesting that the isolates spread between sinks,but did not cause infections.Clone 2f caused multiple contamination of seven sinks in an ICU,including the sinks in the treatment room,the disposal room,and wards.Two patients were infected by this clone.There were 41 CRPA isolates from sinks and clinical samples,including 33 sink source and 8 clinical source.The 25 isolates（including 1 clinical source and 24 sink source）of CRPA were ST253 type and they shared 0～9 SNPs,indicating that the infection was related to the contaminated sinks and the sinks in the ICU were contaminated extensively.Due to the COVID-19 epidemic,only 6 ICUs at 5 hospital had complete the intervention and did the subsequent secondary sampling.The contamination rate of the sink by CR-GNB reduced from 42.31%to 21.15%（χ²=4.44,P=0.035）,and the point contamination rate dropped down from 15.36%to 5.88%（χ²=11.81,P<0.001）.The contamination rate of CR-GNB in GICU decreased from 13.30%to 4.14%（χ²=10.31,P=0.001）,and that in NICU decreased from 24.49%to 12.73%（χ²=1.68,P=0.195）.The residual water of the overflow（30.00%vs 0.00%,P=0.020）,the surface of the overflow（28.21%vs 10.00%,P=0.076）and the aerator（13.46%vs1.89%,P=0.031）were the top three points site that decreased.Conclusion:The number of sinks in the ICU at tertiary hospitals in Chengdu is insufficient.The sinks are prone to splashing during use,and the incidence of CR-GNB contamination is relatively high.The overflows,aerators,disinfectants,and cleaning personnel are important high-risk factors that lead to the contamination.The CR-GNB in the contaminated sink as an environmental reservoir can cause an outbreak of infections.The measures such as enhanced disinfection can significantly reduce the CR-GNB pollution rate of the sink.Health care facilities should pay attention to the setting of sinks in the ICU.Easy to clean,free of splash,and less contamination are the main factors that must be considered.That meet the actual needs of medical work,establish relevant systems and procedures for sink management,and implement measures such as cleaning and disinfection.Cleaning and disinfection policy should be reviewed by the infection control department to avoid the occurrence of hospital infections caused by sink contamination.